1 INFO-VAX	Wed, 25 Jun 2003	Volume 2003 : Issue 347       Contents:0 $brkthruw from a symbiont returns SS$_DEVOFFLINE Another ? on GNV Re: Another ? on GNV/ Re: Another unix/linux cert ... but not on VMS!  Re: Any ISP on VMS?  Re: Any ISP on VMS? 
 crash dump Re: crash dump9 Re: CSWS PHP v0101: MMS gives abort status when compiling ! DEC C/TCPWARE/globalvalue problem % Re: DEC C/TCPWARE/globalvalue problem % Re: DEC C/TCPWARE/globalvalue problem , Re: Do's and don'ts about cleaning keyboards7 Re: Download service for VMS and Tru64 layered products % Re: DSSI disks and allocation classes % Re: DSSI disks and allocation classes % Re: DSSI disks and allocation classes % Re: DSSI disks and allocation classes % Re: DSSI disks and allocation classes % Re: DSSI disks and allocation classes % Re: DSSI disks and allocation classes 
 EMC on VMS Re: EMC on VMS Re: EMC on VMS Re: EMC on VMS Re: EMC on VMS% FA: (finally) 600MHz EV56 164LX Tower  Re: Firewall for VMS / TRU64; Re: Fortran compiler was (Re: HP to port VMS to Oopsteron?) ; Re: Fortran compiler was (Re: HP to port VMS to Oopsteron?) + Re: FORTRAN-Debugger more than one session? + Re: FORTRAN-Debugger more than one session?  Re: GNV and ODS-5  Re: GNV and ODS-5  Re: GNV and ODS-5  Re: GNV and ODS-5  Re: GNV and ODS-5  Re: GNV and ODS-5 P GS1280 upgrade for Bank~Austria Creditanstalt's OpenVMS disaster-tolerant cluste GS160/GS320 complete USD450,000 + Help Required Regarding DECNET Copy Command / Re: Help Required Regarding DECNET Copy Command 4 HELP/MESSAGE problem under SIMH with OpenVMS VAX 6.28 Re: HELP/MESSAGE problem under SIMH with OpenVMS VAX 6.28 Re: HELP/MESSAGE problem under SIMH with OpenVMS VAX 6.28 Re: HELP/MESSAGE problem under SIMH with OpenVMS VAX 6.2P Hewlett-Packard said to be interested in buying software and services companies.P Re: Hewlett-Packard said to be interested in buying software and services compan Hoobyist licence wait  Re: Hoobyist licence wait $ HP Powers More TOP500 Supercomputers( Re: HP Powers More TOP500 Supercomputers( Re: HP Powers More TOP500 Supercomputers Re: HP to drop hpux? Re: HP to drop hpux?6 Re: Hp warns about Suns future ... or lack thereof .../ Re: IBM won't buy VMS ... has NIH syndrome too! / Re: IBM won't buy VMS ... has NIH syndrome too! / Re: IBM won't buy VMS ... has NIH syndrome too!   is there a problem with Info-VAX! is there a problem with Info-VAX? 4 Josh Fisher of HP Labs receives Eckert-Mauchly Award Laptop recommendation? Re: Laptop recommendation?" Re: lib$spawn and persona services" Re: lib$spawn and persona services Memo:  Re:  Memo:  Rant  Re: Mozilla on old hardware  Re: Mozilla on old hardware  Re: Mozilla on old hardware  Re: Mozilla on old hardware  Re: Mozilla on old hardware  Re: New AlphaServer DS15 ??? Re: New AlphaServer DS15 ??? Re: New AlphaServer DS15 ??? Re: New AlphaServer DS15 ??? Re: New AlphaServer DS15 ??? Re: New AlphaServer DS15 ???6 New Business View Magazine for HP Enterprise Customers: Re: New Business View Magazine for HP Enterprise Customers: Re: New Business View Magazine for HP Enterprise Customers! Re: OpenVMS (TM) VAX Version X7G7 - Ottawa OpenVMS Technical Seminar (June 25-26)  physical drive replacement Re: physical drive replacement Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death  Re: Portents of Itanium death " Re: Problem with SAN attached SDLT" Re: Problem with SAN attached SDLT" Re: Problem with SAN attached SDLT" Re: Problem with SAN attached SDLT Re: Problems with SYS$GETRMI Re: Problems with SYS$GETRMI Re: Problems with SYS$GETRMI Re: Problems with SYS$GETRMII Secondhand MicroVAX / AlphaStation systems/parts resellers in Australia ? . SET HOST/LAT and cut and paste with DECwindows SIMH V3.0-0 released Re: SIMH V3.0-0 released Re: SIMH V3.0-0 released Re: SIMH V3.0-0 released Re: SIMH V3.0-0 released Re: SIMH V3.0-0 released Re: SIMH V3.0-0 released Slow FTP Gets. Re: Slow FTP Gets. Re: Slow FTP Gets. Re: Slow FTP Gets. Re: Slow FTP Gets. SSH2 ON VMS? Re: SSH2 ON VMS? Re: SSH2 ON VMS? Re: SSH2 ON VMS? Re: SSH2 ON VMS? Re: SSH2 ON VMS? Re: SSH2 ON VMS? Re: SSH2 ON VMS? Re: Sun to be the next DEC!  system disk on ODS-5 Re: system disk on ODS-5 Re: system disk on ODS-5 Re: system disk on ODS-5 Re: system disk on ODS-5 Re: TCPIP problem with 7.3-1 TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1 Re: TCPIP problem with 7.3-1& Re: Tracing  EXE -> OBJ -> SRC modules& Re: Tracing  EXE -> OBJ -> SRC modules VAX - VMS license transfer Re: VAX - VMS license transfer Re: VAX - VMS license transfer Re: VAX - VMS license transfer Re: VAX - VMS license transfer& Re: Volume Locking on ANAL/DISK/REPAIR& Re: Volume Locking on ANAL/DISK/REPAIR& Re: Volume Locking on ANAL/DISK/REPAIR Re: Who can you trust, then? WRITEMAP.SYS Re: WRITEMAP.SYS XFC docs/SYSGEN inconsistencies : Re: [OT] SCO tells IBM no more AIX ... IBM better buy VMS!) Re: [SPAM] Re: VAX - VMS license transfer  Re: __PAL_REMQUEQ from C Re: __PAL_REMQUEQ from C Re: __PAL_REMQUEQ from C Re: __PAL_REMQUEQ from C  F ----------------------------------------------------------------------    Date: 23 Jun 2003 08:57:05 -0700$ From: gspamtackett@yahoo.com (Galen)9 Subject: $brkthruw from a symbiont returns SS$_DEVOFFLINE = Message-ID: <bdc65a53.0306230757.21da7390@posting.google.com>   B I have a bit of C code to write a breakthru message to a specified# user. It looks something like this:        #include <brkdef.h>      ... 9     $DESCRIPTOR(msg_desc, "Here is a breakthru message"); %     $DESCRIPTOR(user_desc, "SYSTEM"); &     unsigned short io_status_block[4];     ...      status = sys$brkthruw(0,$                           &msg_desc,%                           &user_desc, )                           BRK$C_USERNAME, +                           &io_status_block,                            0,0,&                           BRK$C_QUEUE,                           5,                           0,0);   : (any errors in this snippet are probably due to inaccurate& transcription as I copied it by hand.)  D This works as expected if executed from within a typical interactive! program, even a detached process.   E If executed from a symbiont, though, the I/O status block gets an I/O  status of SS$_DEVOFFLINE.   C Can someone tell me what's happening and how I might get around it?    ------------------------------  # Date: Mon, 23 Jun 2003 20:36:36 GMT 4 From: brad@.gateway.2wire.net (Bradford J. Hamilton) Subject: Another ? on GNV + Message-ID: <oxJJa.2144$Ab2.4169@sccrnsc01>   J My manager asked me to remove GNV on one of the systems here at work todayN (apparently, a previous employee installed it without permission).  Has anyoneN had any experience with removing GNV?  I *know* that I sould be able to removeK it with PRODUCT REMOVE - I just want to know if anyone has had any problems  removing it.  E I'll make a backup of the system disk before I try it, of course.	:-)   A _________________________________________________________________ 0 Bradford J. Hamilton			"All opinions are my own"= bMradAhamiPltSon-at-atMtAbi.cPoSm	"Lose the MAPS, and replace  					'-at-' with @"    ------------------------------  # Date: Tue, 24 Jun 2003 14:50:49 GMT < From: "John E. Malmberg" <Malmberg@dskwld.zko.dec.compaq.hp> Subject: Re: Another ? on GNV 0 Message-ID: <dzZJa.3260$Qj.166@news.cpqcorp.net>   Bradford J. Hamilton wrote: L > My manager asked me to remove GNV on one of the systems here at work todayP > (apparently, a previous employee installed it without permission).  Has anyoneP > had any experience with removing GNV?  I *know* that I sould be able to removeM > it with PRODUCT REMOVE - I just want to know if anyone has had any problems  > removing it.  H I have had no visible problems with using PRODUCT REMOVE on GNV.  I did B not inspect the results as I then installed a pre-release version.  G > I'll make a backup of the system disk before I try it, of course.	:-)   F Of course there is a matter of finding any command files that are now  using GNV utilities.   -John ! malmberg@dskwld.zko.dec.compaq.hp  Personal Opinion Only    ------------------------------  % Date: Mon, 23 Jun 2003 13:01:29 +0100 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> 8 Subject: Re: Another unix/linux cert ... but not on VMS!0 Message-ID: <bd6q6u$fno$1@new-usenet.uk.sun.com>   Bob Ceculski wrote: q > zinser@decus.de (Martin P.J. Zinser) wrote in message news:<220d7b8e.0306180830.1169c105@posting.google.com>...  > ` >>david20@alpha2.mdx.ac.uk (David Webb) wrote in message news:<bcna4i$apc$1@aquila.mdx.ac.uk>... >>Hello, >> >>k >>>In article <d7791aa1.0306170515.6890946d@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:  >>> 9 >>>>this would never happen on VMS, but on os's like unix 9 >>>>and linux with such poor security models, it does ...  >>>>Z >>>>http://www.computerworld.com/securitytopics/security/holes/story/0,10801,82197,00.html >>> - >>>Are you sure VMS isn't affected by this :-  >>>  >>>"Q >>>Adobe Systems Inc.'s Acrobat Reader 5.06 is affected by the problem, as is the 9 >>>open-source reader Xpdf 1.01, according to the report.  >>>  >>>" >>>  >>>XPDF is available on VMS. >>> R >>>Note the problem is to do with the PDF reader calling out to run other programsM >>>when it encounters a link and the potential for someone to write malicious R >>>code which would allow them to affect other files OWNED by the USER running the >>>pdf reader.   >>>  >>>" >>M >>Maybe it is worth mentioning that there is a new version of Xpdf (2.02pl1)  L >>available since a couple of days already which addresses this problem for  >>Unix users.  >>K >>Also upon examination of the patch I did decide this is not a problem on  L >>OpenVMS since the exploit actually does rely on backtick expansion, which L >>is not something DCL usually does. Which is also the reason why I have notJ >>yet updated the OpenVMS port. Xpdf (and the libraries it uses) certainly> >>has some shortcomings (the ones I am aware of are listed on M >>http://www.decus.de:8080/www/vms/sw/notes/xpdf.htmlx), but these are worked  >>on.  >>K >>Also if Bob rather wants to run Windows, this is obviously his choice ;-)  >> >>Greetings, Martin  >  > A > VMS security saves the day again ... and as for windows, a thin ? > vt emulation client with ie and outlook express is alot safer A > than putting unknown unix code on a vms box ... as we have seen > > vms time after time can bail you out of bugs, but as ucx andA > apache proves not every time ... to me a windows client is what  > windows does best ...       Now you have completely lost it.  ; All you are doing by running a thin client with Outlook and ; IE is moving the security problem elsewhere and making your - organisation more vunerable at the same time.   = Outlook/IE are just about the worst offendors of any products @ from a security standpoint. Running IE/Outlook using RDP implies= that you have a NT/Win2000 box also a major offendor. It also = implies that you will probably be using Exchange another huge  hole.   6 Not doing email, web serving and browsing etc is not a6 vote for OpenVMS security and pushing the problem onto4 the most insecure system you could find isn't clever either.   7 And OpenVMS security didn't save the day, DCL not doing 5 backtick expansion isn't a security feature, its just 
 a feature.   Regards  Andrew Harrison    ------------------------------  # Date: Mon, 23 Jun 2003 13:02:02 GMT * From: "Mark E. Levy" <levy@sysman-inc.com> Subject: Re: Any ISP on VMS?/ Message-ID: <eTCJa.112024$YZ2.278989@rwcrnsc53>   : "Larry Kilgallen" <Kilgallen@SpamCop.net> wrote in message- news:q0xzYOLbQfob@eisner.encompasserve.org... = > In article <3EF3CEE2.4E2943F6@fsi.net>, "David J. Dachtera"  <djesys.nospam@fsi.net> writes:  > / > >> If he's still in that business, Mark Levy.  > > F > > Mark sold FSInet to DLS some time back after an unfortunate event. > C > Said event being no fault of Mark's and having nothing to do with 
 computers.   Thanks for the support, guys.    Mark   ------------------------------    Date: 23 Jun 2003 19:32:59 -0700# From: dooleys@snowy.net.au (dooley)  Subject: Re: Any ISP on VMS?= Message-ID: <1ca82fc6.0306231832.3370d504@posting.google.com>   y "Peter Weaver" <WeaverConsultingServices@sympatico.ca> wrote in message news:<bcvhia$o2pcs$1@ID-141708.news.dfncis.de>... < > Because of some new work I will be doing I need to setup a= > small web site for myself. Since I'm a VMS type guy I would ; > rather find a ISP who can host my site on a VMS system. I @ > recall seeing someone post here before that their company does= > this but every key word I can think of has too many hits on 	 > Google.  > < > Is there anyone out there who can host a (currently small)@ > commercial site on a VMS machine. The site name will be in the= > .CA domain, and I would prefer dealing with a Canadian site > > just to make billings, payments and budgets easier, but that > is not 100% necessary. > ) When you open your Australian office, try  http://www.sysworks.com.au/  Phil   ------------------------------  + Date: Tue, 24 Jun 2003 22:54:33 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) Subject: crash dump $ Message-ID: <bdakr9$vjk$1@online.de>  . MACHINECHK, Machine check while in kernel mode  ! What is a possible cause of this?   G One of my ALPHAs crashed on me.  I was 500 km away, logged in remotely. E It came back up automatically, after just the normal reboot time; the F TCPIP cluster alias failed over to a VAX in the cluster and so I couldE reconnect right away. (Nice to see that something I have never tested G live---a hard crash like this, with automatic restart of all software I ? want running, TCPIP failover etc---actually works as planned!)    H This is a 255/233 running 7.2-1, soon to be upgraded.  (Before I do the C upgrade, I want to bring a third voting member into the cluster to  I eliminate downtime (and learn some new things), thus the other thread on   DSSI allocation classes.)   E The machine has been running almost continuously for over 6 years; I  I think this is the first "internal" crash I've had (as opposed to reboots   caused by power failures etc).  " The machine is back to normal now.  E Another question: with the cluster alias pointing to the ALPHA, FTP,  E both normal and anonymous, works fine.  With it pointing to the VAX,  H normal FTP works fine, but anonymous doesn't.  The SYSUAF is not on the H system disk; a logical points to it on both systems.  (Presumably, this I has nothing to do with the cluster alias, but is some difference between  G the machines, which I just noticed when the cluster alias shifted from   ALPHA to VAX.)   I get   C    %RMS-E-PRV, insufficient privilege or file protection violation    - in DISK$USER:[ANONYMOUS]TCPIP$FTP_SERVER.LOG.   H The account itself, file protections etc seem OK.  They ARE the same as " when the login goes via the ALPHA.  I The LOGIN.COM DOES get executed, and the login gets recorded (last login   in AUTHORIZE, for example).    At the other end, I get:  + 331 Guest login OK, send ident as password. 	 Password:  530 Login incorrect.+ %TCPIP-E-FTP_LOGREJ, login request rejected  425 Session is disconnected.  
 Any ideas?   ------------------------------  % Date: Wed, 25 Jun 2003 01:08:10 +0100 * From: "John Travell" <john@travell.uk.net> Subject: Re: crash dump 5 Message-ID: <bdap8p$pqjdt$1@ID-120847.news.dfncis.de>   L "Phillip Helbig---remove CLOTHES to reply" <helbig@astro.multiCLOTHESvax.de>/ wrote in message news:bdakr9$vjk$1@online.de... 0 > MACHINECHK, Machine check while in kernel mode > # > What is a possible cause of this?  > I > One of my ALPHAs crashed on me.  I was 500 km away, logged in remotely. G > It came back up automatically, after just the normal reboot time; the H > TCPIP cluster alias failed over to a VAX in the cluster and so I couldG > reconnect right away. (Nice to see that something I have never tested I > live---a hard crash like this, with automatic restart of all software I @ > want running, TCPIP failover etc---actually works as planned!) > I > This is a 255/233 running 7.2-1, soon to be upgraded.  (Before I do the D > upgrade, I want to bring a third voting member into the cluster toJ > eliminate downtime (and learn some new things), thus the other thread on > DSSI allocation classes.)  > F > The machine has been running almost continuously for over 6 years; IJ > think this is the first "internal" crash I've had (as opposed to reboots  > caused by power failures etc). > $ > The machine is back to normal now. > - A machine check is a HARDWARE detected error. J The evidence will be in the errorlog. I cannot remember offhand if the 255F is one of the machines supported in ERF, or whether it needs DECevent.' In any case the commands are similar... = $ analyse/error/include=(cpu,mac,mem) sys$errorlog:errlog.sys 4 $ diag/include=(cpu,mac,mem) sys$errorlog:errlog.sys  J Read through the plain text comments on the errorlog entries. The cause ofI the crash may, or may not, be obvious. If not, post the relevant errorlog : entries here. Someone, if not me, will give you an answer.   -- John Travell  VMS crashdump expertise for hire john@travell.uk.net  http://www.travell.uk.net/           --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.492 / Virus Database: 291 - Release Date: 24/06/2003    ------------------------------  # Date: Mon, 23 Jun 2003 17:20:10 GMT 5 From: Matthew Doremus <Matthew.Doremus@hp.com.NoSpam> B Subject: Re: CSWS PHP v0101: MMS gives abort status when compiling, Message-ID: <3EF7317E.6060902@hp.com.NoSpam>  I The problem you're having is due to the fact that PHP 1.1 was built with  I a version of the DECC compiler which did not support the UTIME function.  H  The simplest way to fix this problem for your environment is to remove E the build of utime.c from the PHP_ROOT:[OpenVMS]OpenVMS.MMS descript  G file and rebuild or, more quickly, delete the utime object module from  < the PHP_ROOT:[OpenVMS.OBJ]LIBOPENVMS.OLB library and relink.   Chris Guthrey wrote:  P >I'm trying to compile the PHP V0101 sources on Alpha VMS 7.3-1 with both TCP/IPF >and Multinet installed, but with Multinet active. Here is my session: >  >$ set def php_root:[build]  >$ MMS/descrip=build.mms >  > - >! PHP_ROOT:[BUILD]PHP_EDIT.DAT is up to date  >  > 8 >! Built PHP_ROOT:[BUILD.OBJ]PHPSHRP.EXE for OpenVMS PHP > 3 >%LINK-W-MULDEF, symbol DECC$UTIME multiply defined L >        in module DECC$SHR_EV56 file SYS$COMMON:[SYSLIB]DECC$SHR_EV56.EXE;17 >%LINK-W-MULDEF, symbol DECC$INET_ATON multiply defined L >        in module DECC$SHR_EV56 file SYS$COMMON:[SYSLIB]DECC$SHR_EV56.EXE;1L >%MMS-F-ABORT, For target PHP_ROOT:[BUILD.OBJ]PHPSHR.EXE, CLI returned abort >status: %X10648268. > M >Although I am aware that CSWS is not officially supported with Multinet, the L >binary distributions seem to work fine.  I want to see if I can compile theN >myssql extension for PHP but I'd like to understand the above PHP build error >first!  > $ >Any assistance gratefully received! >  >Cheers, Chris Guthrey.  >    ------------------------------  % Date: Tue, 24 Jun 2003 20:59:06 +0100 9 From: "covendotartdottalk21dotcom" <postmaster@127.0.0.1> * Subject: DEC C/TCPWARE/globalvalue problem3 Message-ID: <btydnbtyupQRMGWjXTWJiA@brightview.com>   
 Hi, folks.  = I'm trying to provide an interim solution prior to some Vaxen : (5.5-2H4) being replaced with Alphas running somewhat more< recent versions of OVMS, which necesssitates me developing a: C program (to avoid the hideous error-handling problems of* DCL) to FTP some files to/from a U*ix box.  8 However, I'm encountering a problem whilst trying to use9 mandatory globalvalue "definitions" that TCPWARE uses for ' its error codes (TCPWARE_NONODE &etc.).   < Okay, things are obviously a little bit out of date, but the< TCPWARE 5.1 programmer's reference manual says that in order? to access these codes, they need to be defined as globalvalues.   = I don't have a problem with this (I've never used globalvalue 9 before, only globaldef and globalref, so you'll pardon my < ignorance if I don't understand why TCPWARE doesn't just use= header files and #defines), but I'm having a problem with the  use of the values.  & If I try to reference the values using   int a = TCPWARE_NONODE ;   or  ; printf("\nValue of TCPWARE_NONODE is %d", TCPWARE_NONODE) ;   ( then this doesn't prove to be a problem.      However, if I do something like:   switch (tcpware_error) {     case TCPWARE_NONODE :     /* blah */     break ;       default :     break ;  }     ; Then the DEC C compiler complains, saying that (in the case 9 statement) the TCPWARE_NONODE needs to be a constant, but < actually isn't (dialled in from PC at home at the moment, so( I don't recall the exact error message).  ; I've done a search on Google, and tried using references to    #pragma extern_model save   #pragma extern_model globalvalue globalvalue TCPWARE_NONODE ; #pragma extern_model restore  > from a posting by Wayne Sewell (with me adding the globalvalue+ reference as it appears here), to no avail.   @ On rereading the posting, I may have misspelt the "extern_model"B as "external_model" (but can't check just now), but the use of theB pragmas makes no difference, nor does it make any difference as to9 whether or not /STANDARD=VAXC is used on the compilation.   > Can anyone tell me what I'm doing wrong (and more importantly,@ why?), and what benefits there are to using globalvalue on DEC C2 or VAX C on a Vax, or plain old DEC C on an Alpha?  ? Are there any benefits at all, or is this just a peculiarity of ? (extensions to) ANSI which would only be of use when porting to  other O/Ses?  9 I've also seen references to using "extern" as opposed to < "globalvalue" when using DEC C, and that it is an acceptable; alternative (according to one of the posts that the OpenVMS : Wizard shows on www.hp.com), but isn't (due to the obvious< default inference of "int") according to a comp.os.vms post, and am somewhat perplexed.   Regards,  ? Mark (who doesn't read his Vax and Alpha architecture reference  manuals on a regular basis)    ------------------------------  # Date: Tue, 24 Jun 2003 22:05:09 GMT < From: "John E. Malmberg" <Malmberg@dskwld.zko.dec.compaq.hp>. Subject: Re: DEC C/TCPWARE/globalvalue problem2 Message-ID: <pW3Ka.3290$O31.1893@news.cpqcorp.net>  ! covendotartdottalk21dotcom wrote:  > Hi, folks. > ? > I'm trying to provide an interim solution prior to some Vaxen < > (5.5-2H4) being replaced with Alphas running somewhat more> > recent versions of OVMS, which necesssitates me developing a< > C program (to avoid the hideous error-handling problems of, > DCL) to FTP some files to/from a U*ix box.  G It may be simpler to solve that problem with DCL.  I use an FTP script  H that copies the file to a temporary name and then renames it.  That way B the receiver never sees a partial transfer.  The $status code and I checking the output of the script can both be used to determine success.  F   The SEARCH command can search for appropriate codes in a simple log " with out having to parse the logs.  : > However, I'm encountering a problem whilst trying to use; > mandatory globalvalue "definitions" that TCPWARE uses for ) > its error codes (TCPWARE_NONODE &etc.).  > > > Okay, things are obviously a little bit out of date, but the> > TCPWARE 5.1 programmer's reference manual says that in orderA > to access these codes, they need to be defined as globalvalues.  > ? > I don't have a problem with this (I've never used globalvalue ; > before, only globaldef and globalref, so you'll pardon my > > ignorance if I don't understand why TCPWARE doesn't just use? > header files and #defines), but I'm having a problem with the  > use of the values.  & A globalvalue is a link time constant.  ( > If I try to reference the values using >  > int a = TCPWARE_NONODE ; >  > or > = > printf("\nValue of TCPWARE_NONODE is %d", TCPWARE_NONODE) ;  > * > then this doesn't prove to be a problem. >  > " > However, if I do something like: >  > switch (tcpware_error) > {  >    case TCPWARE_NONODE : >     /* blah */
 >     break ;  >  >    default :
 >     break ;  > }  >  > = > Then the DEC C compiler complains, saying that (in the case ; > statement) the TCPWARE_NONODE needs to be a constant, but > > actually isn't (dialled in from PC at home at the moment, so* > I don't recall the exact error message).  C The C compiler probably requires a case label to be a compile time  , constant, so a global value can not be used.  = > I've done a search on Google, and tried using references to  >  > #pragma extern_model save " > #pragma extern_model globalvalue > globalvalue TCPWARE_NONODE ; > #pragma extern_model restore  J The extern model has nothing to do with how a global value is interpreted.  I The extern model determines how "extern" declarations are handled by the  	 compiler.   I The default for DEC/COMPAQ/HP C for OpenVMS is to treat extern variables  H as global variables for other programming languages.  This is what most  programmers would expect.   G The default for VAXC is to treat extern variables as psects or FORTRAN   common blocks.  E globalref and globaldef were needed so that VAXC could access global   variables from other languages.   E globalref is the same as extern with out assigning a value to to the   variable.  "extern int foo;"  I globaldef is the same as extern when you assign a value to the variable.     "extern int foo = 1".   @ > from a posting by Wayne Sewell (with me adding the globalvalue- > reference as it appears here), to no avail.  > B > On rereading the posting, I may have misspelt the "extern_model"D > as "external_model" (but can't check just now), but the use of theD > pragmas makes no difference, nor does it make any difference as to; > whether or not /STANDARD=VAXC is used on the compilation.  > @ > Can anyone tell me what I'm doing wrong (and more importantly,B > why?), and what benefits there are to using globalvalue on DEC C4 > or VAX C on a Vax, or plain old DEC C on an Alpha?  ; globalvalue may be easier to code than the ANSI equivalent.   - Lets see if I can get this right from memory.   H The C compiler really only knows external symbols as addresses, so with 4 out a globalvalue specification, you have to declare       globalvalue int foo;   as:        extern int *foo;  8 And then make sure that everything can accept a pointer.   Or you can use:        extern int foo;   I And then make sure that you reference &foo when ever you need it's value.   A > Are there any benefits at all, or is this just a peculiarity of A > (extensions to) ANSI which would only be of use when porting to  > other O/Ses?  A The use of /STANDARD=VAXC will supress some diagnostics from the  / compiler.  I would recommend avoiding using it.   I I would use /WARN=ENABLE=(LEVEL4, QUESTCODE) instead and fixing the bugs  H that it reports, or using the #pragma message codes to disable the ones * that you think will not hurt your program.   -John ! malmberg@dskwld.zko.dec.compaq.hp  Personal Opinion Only    ------------------------------  % Date: Tue, 24 Jun 2003 23:48:12 +0100 9 From: "covendotartdottalk21dotcom" <postmaster@127.0.0.1> . Subject: Re: DEC C/TCPWARE/globalvalue problem3 Message-ID: <FvKdnZASeZmlSGWjXTWJhg@brightview.com>   G "John E. Malmberg" <Malmberg@dskwld.zko.dec.compaq.hp> wrote in message , news:pW3Ka.3290$O31.1893@news.cpqcorp.net.../ > Lets see if I can get this right from memory.  > I > The C compiler really only knows external symbols as addresses, so with 6 > out a globalvalue specification, you have to declare >  >     globalvalue int foo; >  > as:  >  >     extern int *foo; > : > And then make sure that everything can accept a pointer. >  > Or you can use:  >  >     extern int foo;  > K > And then make sure that you reference &foo when ever you need it's value.   J However, the TCPWARE programmer's reference manual to which I have access,J doesn't say whether or not these "constant"s are char, unsigned char, int,K unsigned int, long, unsigned long, or exactly what "type" of variable, so I > have no idea whether or not "int" is the corect variable type!  I I can't imagine that it is "safe" to just assume "int" (and that this, if J I'm currently correct, DEC C will continue to assume so for <indiscernible$ time in the future>), or am I wrong?  ? I still don't see what PSC have to benefit by effectively using G globalvalue FOO as extern (*)FOO (and hence, by default, int (*) FOO in 2 both cases) - why the h*ll not just use #define???     MarkJ (hoping to try out suggestions at work tomorrow if the $Ford_motor_vehicleK doesn't keep on spouting "EAC [Electronic ACcelerator] Fail" every time the H ignition is switched on, much less fail at $exceeding_speed_limit_MPH in lane 2 of the motorway...)   ------------------------------   Date: 23 JUN 2003 15:22:15 GMT4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher)5 Subject: Re: Do's and don'ts about cleaning keyboards 6 Message-ID: <23JUN03.15221566@thuria.waisman.wisc.edu>  A In a previous article, p_sture@elias.decus.ch (Paul Sture) wrote:    .. WD 40 ignites...   B ->And just to follow on, your fire suppression system is Halon, soH ->once it has dumped that, you are legally not allowed to replace it, so ->have to buy a new system.   D If you sell the halon before it dumps you might be able to pay for a@ new system with the proceeds (and could have cash left if it's a large amount).   --G -- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madison : --                   karcher.nomorespxm@waisman.wisc.edu     ------------------------------  % Date: Mon, 23 Jun 2003 11:40:47 -0400 , From: "warren sander" <warren.sander@hp.com>@ Subject: Re: Download service for VMS and Tru64 layered products, Message-ID: <3ef7241d$1@usenet01.boi.hp.com>  L They are the folks that run the service for us. They have folks that used toJ be part of digital (software testing and release group). They still do theJ same work for us. they had the infrastucture to do the distributions (theyK have the db of who gets what software product libraries etc). So they admin I the site for us. I guess it was easier to use their site in hp brand than ; get a hp license plate (would you what h28382.www4.hp.com?)    -warren   6 "Barry Treahy, Jr." <Treahy@MMaz.com> wrote in message! news:3EF39521.3070001@MMaz.com...  > warren sander wrote: > J > >As a result of the feedback from comp.os.vms they have decided to allow FTP  > >access to the kits. > >   > >there are 2 ways to do this.. > > 6 > >1) via FTP command level etc  (browsing is allowed)? > >2) via ftp://user:pass@www1.aclabs.com (browsing is allowed)  > >  > > * > What is the ACLABS.COM domain all about? > 
 > Registrant:  > Ajilon LLC (IQXNOYSLCD)  >    40 Simon St.  >    Nashua, NH 03060  >    US  >  > Barry  >  > --   > A > Barry Treahy, Jr  *  Midwest Microwave  *  Vice President & CIO  > C > E-mail: Treahy@mmaz.com * Phone: 480/314-1320 * FAX: 480/661-7028  >  >    ------------------------------  # Date: Mon, 23 Jun 2003 17:59:10 GMT " From: Alfred Falk <falk@arc.ab.ca>. Subject: Re: DSSI disks and allocation classes9 Message-ID: <Xns93A379ED9809Bfalkarcabca@205.233.108.180>   D helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to , reply) wrote in news:bd3roa$edf$2@online.de:  I > I've successfully installed VMS on a VAX 4000 100A I inherited.  There  I > is an internal RF35 disk called $1$DIA3.  It appears to have about 800  0 > MB.  I used this as the system disk (for now). > <etc>   : You've already got lots of answers, but here's my version.G The allocation class of a DSSI drive matters only if you want a system  I to serve the disk to other cluster members.  A node can only serve disks   with the same allocation class.W  = The commands you are looking for are run from the >>> prompt.uG It is also possible, as someone has pointed out to do it from DCL on a   live system.   >>> SET HOST/DUP/DSSI 3 PARAMS PARAMS> SHOW ALLCLASSn/ PARAMS> SET ALLCLASS 255    	    	(or whatever)s/ PARAMS> SHOW ALLCLASS    	    	    	(to verify)s  E If you don't like the default node name that shows up in SHOW CLUSTERl  
 SHOW NODENAME  SET NODENAME whateverb  > There are other parameters, most of which are best left alone. SHOW *  8 PARAMS> WRITE    	    	    	    	Don't forget this line!  
 >>> SHOW DSSI    --  @ ----------------------------------------------------------------A   A L B E R T A         Alfred Falk               falk@arc.ab.ca s@ R E S E A R C H         Information Systems Dept   (780)450-5185+   C O U N C I L         250 Karl Clark Road 1                         Edmonton, Alberta, CanadaV http://www.arc.ab.ca/   T6N 1E4   http://www.arc.ab.ca/staff/falk/   ------------------------------  % Date: Mon, 23 Jun 2003 10:05:39 +0200e% From: "Fred Zwarts" <F.Zwarts@KVI.nl>P. Subject: Re: DSSI disks and allocation classes. Message-ID: <bd6c3l$qn9$1@info.service.rug.nl>  , "Phillip Helbig---remove CLOTHES to reply" =4 <helbig@astro.multiCLOTHESvax.de> wrote in message = news:bd3udt$gqh$1@online.de...H > In article <8hg8MrR4RlLw@elias.decus.ch>, p_sture@elias.decus.ch (Paul > Sture) writes:=20S >=20C > > > Is the allocation class used for DSSI disks the same as the =  "normal"=20q7 > > > allocation class (NOT the port allocation class)?m > > >=20G > > > The reason I ask is that otherwise there is no allocation class =V defined;F > > > the device shows up as $1$DIA3 automatically, even after a freshH > > > install.  Apparently, this is the default for DSSI disks whereas = for=20G > > > SCSI disks there is no default (or rather it is 0 and the NODE$ =i names=20 > > > are used). > >=20G > > It's been a long time since I used DSSI disks, but IIRC you have torC > > connect to the controller and do it from there. Something like:s > >=20 > > $ mc sysman io conn =m8 fya0:/noad/driver=3Dsys$loadable_images:SYS$FYDRIVER.EXE > > $ set proc/priv=3DdiagnoseB > > $ set host /dup /task=3Dcli /server=3Dmscp$dup controller_name > >=20B > > Where controller_name can be seen from a SHOW CLUSTER display. >=20F > Interesting.  Even though I just did a fresh install of VMS, SHOW=20I > CLUSTER does output something, including the controller name.  OK, so =p I=20" > can try your command suggestion. >=20J > Still, what is the relationship between this allocation class and the=20F > regular allocation class, which one would normally change via the=20C > cluster-configuration procedure or via MODPARAMS.DAT and AUTOGEN?e  C It is a normal allocation class. But, in some way, a DSSI disk is aSB "node" (not a member) in a DSSI cluster. It will show up in the=20C SHOW CLUSTER output and it has its own allocation class defined.=20e? For your VAX, a DSSI disk is not considered as a local disk.=20iI On your VAX, you define whether you want to serve disks of other nodes=20iG (e.g., DSSI disks) to other members. (Check the MSCP_SERVE_ALL system =a parameter.)n  H Also check your VAX's documentation. We have two VAX 4000-300 systems, = whicheH allow you already from the console mode to login onto the DSSI disk to = changeC parameters, like the allocation class. You don't need to start VMS.n. I don't now whether your VAX allows that, too.   ------------------------------  + Date: Mon, 23 Jun 2003 18:55:15 +0000 (UTC)u7 From: moroney@world.std.spaamtrap.com (Michael Moroney) . Subject: Re: DSSI disks and allocation classes( Message-ID: <bd7iej$51u$1@pcls4.std.com>  3 keithparris_NOSPAM@yahoo.com (Keith Parris) writes:c  E >I detect an incorrect assumption that the allocation classes must beIB >different to avoid a conflict.  This is not the case.  The deviceE >name, as you say, is DIA3, which would never conflict with any localsE >device names on a VMS node (SCSI devices would be DKcn:, DSA deviceso >DUcn:, etc.  I That is true.  As long as you're using a recent DUDRIVER (7.1+patches or -G 7.2 and beyond) the only possible conflict is another DSSI disk callingVF itself DIA3:.  If its allocation class is different even this is not a	 conflict.P   --   -Mikew   ------------------------------  + Date: Mon, 23 Jun 2003 18:58:34 +0000 (UTC)e7 From: moroney@world.std.spaamtrap.com (Michael Moroney)S. Subject: Re: DSSI disks and allocation classes( Message-ID: <bd7ikq$51u$2@pcls4.std.com>  $ Alfred Falk <falk@arc.ab.ca> writes:  H >The allocation class of a DSSI drive matters only if you want a system J >to serve the disk to other cluster members.  A node can only serve disks   >with the same allocation class.  E No longer true.  I fixed this in dudriver for V7.2. (and V7.1+latest e patch kits)  -- C -Mikeb   ------------------------------    Date: 23 Jun 2003 10:20:39 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)E. Subject: Re: DSSI disks and allocation classes= Message-ID: <cf15391e.0306230920.15c6dfdf@posting.google.com>s  | helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) wrote in message news:<bd3roa$edf$2@online.de>...I > I've successfully installed VMS on a VAX 4000 100A I inherited.  There iI > is an internal RF35 disk called $1$DIA3.  It appears to have about 800 T0 > MB.  I used this as the system disk (for now). > C > Apparently, this is a DSSI disk.  I have no experience with these-J > beasts.  Show device/full shows that it has allocation class 1.  I want J > to put it in a cluster where there is already a machine with allocation I > class 1, thus I want to change the allocation class on the new machine.:  D I detect an incorrect assumption that the allocation classes must beA different to avoid a conflict.  This is not the case.  The device7D name, as you say, is DIA3, which would never conflict with any localD device names on a VMS node (SCSI devices would be DKcn:, DSA devices DUcn:, etc.e  B > the device shows up as $1$DIA3 automatically, even after a freshG > install.  Apparently, this is the default for DSSI disks whereas for iH > SCSI disks there is no default (or rather it is 0 and the NODE$ names  > are used).  D No, RF disks have an SCS nodename, and the default device name wouldF be nodename$DIAn:.  It is only when a non-zero allocation class is set3 on the RF disk that the device appears as $a$DIAn:.    ------------------------------  # Date: Mon, 23 Jun 2003 18:47:54 GMT)" From: Alfred Falk <falk@arc.ab.ca>. Subject: Re: DSSI disks and allocation classes9 Message-ID: <Xns93A3823098518falkarcabca@205.233.108.180>f  G Alfred Falk <falk@arc.ab.ca> wrote in news:Xns93A379ED9809Bfalkarcabca@e 205.233.108.180:  F > helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to . > reply) wrote in news:bd3roa$edf$2@online.de:  G An afterthought: On a 100A, (as opposed to 100) there should be 2 DSSI 2 buses.  Look at output fromm        >>> SHOW DSSIE to see which bus your drive is connected to.  Normally it would be 0,J5 but if it is connected to bus 1, the command would ber*     	>>> SET HOST/DUP/DSSI/BUS:1 id PARAMS   --  @ ----------------------------------------------------------------A   A L B E R T A         Alfred Falk               falk@arc.ab.ca I@ R E S E A R C H         Information Systems Dept   (780)450-5185+   C O U N C I L         250 Karl Clark Roadg1                         Edmonton, Alberta, Canadae http://www.arc.ab.ca/   T6N 1E4_  http://www.arc.ab.ca/staff/falk/   ------------------------------  + Date: Mon, 23 Jun 2003 19:06:11 +0000 (UTC)M7 From: moroney@world.std.spaamtrap.com (Michael Moroney)c. Subject: Re: DSSI disks and allocation classes( Message-ID: <bd7j33$51u$3@pcls4.std.com>  R helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) writes:  J >In that case, it would be more logical to have the DSSI allocation class F >be different than the SCSI allocation class on the same machine---at A >least if (for various reasons) each machine in my cluster has a aF >different allocation class.  (After all, if the controller should be # >thought of as being its own node.)b  C The configuration requirements are that if you are using SCSI port hC allocation classes, that allocation class must be unique within thesH cluster.  For this reason the dssi allocation class must be different to follow the configuration rules.o  I Beyond this, I feel setting each 'bus' to a unique allocation class helps I keep down confusion.  Each SCSI bus using port allocation classes already H must follow this rule, but if, say, you have lots of DSSI disks, you mayF want to have all those on one dssi bus have the same unique allocationI class, all those on the second have the same second allocation class etc.r -- _ -Mikea   ------------------------------  # Date: Tue, 24 Jun 2003 19:47:14 GMTe" From: Lee <lytmah@telusplanet.net> Subject: EMC on VMSu/ Message-ID: <3EF8AAB2.51554A33@telusplanet.net>e  8 I'll be testing four EMC (8.2GB) disks on my VMS cluster9 shortly for compatibility and functionality and I requiree5 info from someone who may already have gone down thisk7 path.  My initial step will be to connect two FC cables58 (for redundancy) from one node (L) of the cluster to the) EMC SAN.  I have the following questions:r;     - Can I hook up the cables without taking down the nodec#         down to the console prompt? :     - All the VMScluster nodes have Sysgen ALLOCLASS OF 2,A         as do the HSJ disk controllers, resulting in our physical'?         drives and tapes with the prefix $2$.  I'm told the EMCa7         drives will have $1$ as their allocation class.o;         Will VMS be accepting of this, especially for HBVS?      -- Lee>   lytmah@telusplanet.net   ------------------------------    Date: 24 Jun 2003 14:58:08 -0500+ From: young_r@encompasserve.org (Rob Young)  Subject: Re: EMC on VMSr3 Message-ID: <sDAT0uUuNwPN@eisner.encompasserve.org>o  T In article <3EF8AAB2.51554A33@telusplanet.net>, Lee <lytmah@telusplanet.net> writes:: > I'll be testing four EMC (8.2GB) disks on my VMS cluster; > shortly for compatibility and functionality and I requirea7 > info from someone who may already have gone down this 	 > path.  u  < 	You mean 8.6+ Gbyte drives - 17677440 blocks to be precise.  2 > My initial step will be to connect two FC cables: > (for redundancy) from one node (L) of the cluster to the+ > EMC SAN.  I have the following questions:t= >     - Can I hook up the cables without taking down the nodee% >         down to the console prompt?t  ( 	Assumption:  HBAs are already in place.  B 	Yes, you can plug cables into HBAs and switches without downtime.5 	You can then SYSMAN io auto to make devices show up.w  < >     - All the VMScluster nodes have Sysgen ALLOCLASS OF 2,C >         as do the HSJ disk controllers, resulting in our physicaltA >         drives and tapes with the prefix $2$.  I'm told the EMCe9 >         drives will have $1$ as their allocation class.   8 	Fibre in general.  Look in various docs.  Fibre drives  	show up as $1$   = >         Will VMS be accepting of this, especially for HBVS?   < 	Yes.  However, kiss mini-merge goodbye.  Keep that in mind.   				Robr   ------------------------------  # Date: Tue, 24 Jun 2003 20:44:16 GMTr" From: Lee <lytmah@telusplanet.net> Subject: Re: EMC on VMSe/ Message-ID: <3EF8B814.DFCF451A@telusplanet.net>b  M HBA's are already in place.  EMC is already being used for non-VMS platforms.O& Does the HP SAN allow for mini-merges?7 Can you use EMC disks for system disks?  I.e., $1$DUAx? < Can you use shadowed EMC disks as system disks?  I.e., DSAx?     Rob Young wrote:  V > In article <3EF8AAB2.51554A33@telusplanet.net>, Lee <lytmah@telusplanet.net> writes:< > > I'll be testing four EMC (8.2GB) disks on my VMS cluster= > > shortly for compatibility and functionality and I require 9 > > info from someone who may already have gone down this 	 > > path.  >aE >         You mean 8.6+ Gbyte drives - 17677440 blocks to be precise.2 >u4 > > My initial step will be to connect two FC cables< > > (for redundancy) from one node (L) of the cluster to the- > > EMC SAN.  I have the following questions:n? > >     - Can I hook up the cables without taking down the noden' > >         down to the console prompt?E >E1 >         Assumption:  HBAs are already in place.. >pK >         Yes, you can plug cables into HBAs and switches without downtime. > >         You can then SYSMAN io auto to make devices show up. > > > >     - All the VMScluster nodes have Sysgen ALLOCLASS OF 2,E > >         as do the HSJ disk controllers, resulting in our physical C > >         drives and tapes with the prefix $2$.  I'm told the EMC ; > >         drives will have $1$ as their allocation class.  >?@ >         Fibre in general.  Look in various docs.  Fibre drives >         show up as $1$ >l? > >         Will VMS be accepting of this, especially for HBVS?I >4E >         Yes.  However, kiss mini-merge goodbye.  Keep that in mind.n > % >                                 Rob.   -- Leea   lytmah@telusplanet.net   ------------------------------  # Date: Wed, 25 Jun 2003 00:06:15 GMTn1 From: Michael Austin <maustin@firstdbasource.com>  Subject: Re: EMC on VMS 2 Message-ID: <3EF8E590.4E489645@firstdbasource.com>  
 Lee wrote: > O > HBA's are already in place.  EMC is already being used for non-VMS platforms. ( > Does the HP SAN allow for mini-merges?9 > Can you use EMC disks for system disks?  I.e., $1$DUAx?t> > Can you use shadowed EMC disks as system disks?  I.e., DSAx? >  > Rob Young wrote: > X > > In article <3EF8AAB2.51554A33@telusplanet.net>, Lee <lytmah@telusplanet.net> writes:> > > > I'll be testing four EMC (8.2GB) disks on my VMS cluster? > > > shortly for compatibility and functionality and I requireC; > > > info from someone who may already have gone down thisj > > > path.> > >AG > >         You mean 8.6+ Gbyte drives - 17677440 blocks to be precise.o > >s6 > > > My initial step will be to connect two FC cables> > > > (for redundancy) from one node (L) of the cluster to the/ > > > EMC SAN.  I have the following questions:aA > > >     - Can I hook up the cables without taking down the nodeh) > > >         down to the console prompt?n > >t3 > >         Assumption:  HBAs are already in place.  > >rM > >         Yes, you can plug cables into HBAs and switches without downtime.n@ > >         You can then SYSMAN io auto to make devices show up. > >e@ > > >     - All the VMScluster nodes have Sysgen ALLOCLASS OF 2,G > > >         as do the HSJ disk controllers, resulting in our physicaloE > > >         drives and tapes with the prefix $2$.  I'm told the EMCn= > > >         drives will have $1$ as their allocation class.h > >sB > >         Fibre in general.  Look in various docs.  Fibre drives > >         show up as $1$ > >sA > > >         Will VMS be accepting of this, especially for HBVS?i > > G > >         Yes.  However, kiss mini-merge goodbye.  Keep that in mind.n > >t' > >                                 Roba >  > -- > Lee  >  > lytmah@telusplanet.net  E In some environments, relying on the mirror/raid functionality of theaF controllers is more than sufficient redundancy rather than the expenseD of using shadowing.  However, there are times when procedural issues( come into play and you need to use HBVS.   -- i Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163    ------------------------------  % Date: Tue, 24 Jun 2003 20:50:54 -0500i1 From: "David J. Dachtera" <djesys.nospam@fsi.net>d Subject: Re: EMC on VMS(' Message-ID: <3EF8FFFD.A2E8414C@fsi.net>   
 Lee wrote: > O > HBA's are already in place.  EMC is already being used for non-VMS platforms. ( > Does the HP SAN allow for mini-merges?  C Careful there. A SAN is nothing more than a "storage area network":-G fibre cables and switches. HBAs, HSGs and such are network attachments,hH not part of the network itself. The SAN neither facilitates nor prevents2 anything, it simply provides connectivity. Period.  9 > Can you use EMC disks for system disks?  I.e., $1$DUAx?v  C FC disks appear as DGA, not DUA. ...and yes, FC volumes can be boot = disks. Not sure about EMC, specifically, but it does work foru HSG-attached disks.n  > > Can you use shadowed EMC disks as system disks?  I.e., DSAx?  G EMC does not support certain SCSI operations necessary for full supporthH of HBVS. As I understand it, if a disk block were to go bad and the dataE be unrecoverable, VMS would be unable to write the forced error flag.sG Therefore, HBVS would have no option other than to drop the member fromwD the shadow-set under a condition that would normally be recoverable.   --   David J. Dachtera  dba DJE Systemso http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/e   ------------------------------  % Date: Tue, 24 Jun 2003 01:51:58 -0600a- From: "Brian Zurbach" <briansz@sitemouse.com>a. Subject: FA: (finally) 600MHz EV56 164LX Tower: Message-ID: <pan.2003.06.24.07.51.41.294019@sitemouse.com>  H On fleaBay, no reserve, $9.99 minimum.  Fast machine for the $$, flashedC to the latest SRM firmware.  Will ship anywhere you pay to send it.e  = http://cgi.ebay.com/ws/eBayISAPI.dll?ViewItem&item=3032083375m   ** Check my feedback rating **   Regards,   Brian S. Zurbach briansz@sitemouse.comv   ------------------------------  % Date: Mon, 23 Jun 2003 17:15:45 +0100uO From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>l% Subject: Re: Firewall for VMS / TRU64o0 Message-ID: <bd793i$kuj$1@new-usenet.uk.sun.com>   Hoff Hoffman wrote:b > .. > J >   I'd not use Solaris as a firewall, either -- running a general-purposeK >   operating system as a firewall is not something that I would choose to oL >   do, regardless of the options and availabilies and filters and platform. >   < However this is what a lot of appliances are under the hood.  ; General purpose OS hardened and supplied with a Firewall asa
 an appliance.2   Regards- Andrew Harrison-K >   Can a general-purpose system be secured sufficiently?  Very likely yes,hJ >   at least for the more secure platforms.  Is the effort and the support< >   time involved cost-effective?  I don't believe it to be. > I >   Running a firewall requires a support effort, and -- with small-scale H >   firewalls available with prices starting around US$100 -- not a taskH >   that I could reasonably justify choosing to perform myself, as it isJ >   an ONGOING support task and one that is not without substantial risks. > J >   Corporate IP firewalls with stateful inspection and filtering and DDoSI >   defenses and subscription services and mail-scanning capabilities aretK >   obviously somewhat more expensive, but are still readily available fromt >   security products vendors. , > I >   I here refer to the requirement that the firewall must be reviewed as I >   users and software are added and removed and as packages are upgradedoC >   -- the up-front part of configuring ipchains or other firewall PF >   mechanisms is the easy part, and some of the newer GUIs make theseE >   tasks even easier.  Where this gets expensive is with KEEPING theIK >   system secure.  Regardless of whether the platform is Solaris, OpenVMS,$ >   Windows, or otherwise. L > D >   I've seen suggestions of removing chunks off the general-purposeD >   box, and this is certainly feasible -- but there are still costsE >   involved with maintaining and with upgrading the firewall that is 0 >   on this locally-produced firewall appliance. > H >   A dedicated firewall appliance eliminates much of the support effortF >   discussed above, and provides me with a dedicated solution where IB >   can also contract for vendor support and upgrades as required. > H >   This is all a question of priorities, and particularly one of costs. > J >   This decision is a no-brainer for most implementations and most sites. > P >  ---------------------------- #include <rtfaq.h> -----------------------------L >       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.comP >  --------------------------- pure personal opinion ---------------------------G >         Hoff (Stephen) Hoffman   OpenVMS Engineering   hoff[at]hp.como >    ------------------------------  % Date: Mon, 23 Jun 2003 18:09:19 +1000n1 From: Paddy O'Brien <paddy.o'brien@tg.nsw.gov.au>TD Subject: Re: Fortran compiler was (Re: HP to port VMS to Oopsteron?), Message-ID: <3EF6B5AF.3080500@tg.nsw.gov.au>   Phillip Helbig wrote:aI >>Fortran is essentially in maintenance mode on VMS -- at F77 on VAX and d >>at F95 on Alpha. >  > H > The Fortran90 standard came out in 1991 or something, and ALPHA in in J > 1992 or whatever.  There was a decision made not to develop a Fortran90 I > compiler for the VAX.  I think this was a correct decision.  There are  J > two types of Fortran people, those who do cutting-edge number crunching J > and those running old, well tested code.  The former will of course use J > the newest hardware and compiler techniques, so obviously Fortran90 and J > ALPHA were a good match.  No serious number cruncher would use VAX when H > ALPHA is available, so these guys don't care about Fortran90 on VAX.  I > Folks with old code running a VAX at VMS 5.5-2 obviously don't want to cB > risk changing anything, so these guys wouldn't be interested in  > Fortran90 for VAX either.  > C Yes, I can agree with your judgement on no F95 for VAX, but if the  C compiler is common and only the GEM backend different, why does it mD matter?  We are probably in the area of what you call "cutting-edge F number crunching".  I have a 4-CPU ES40 for our production work which H runs at about 400% most days of the month.  Our engineers do very large A dynamic analyses of our 4-state interconnected system where they tE represent all the generator control systems and induction motors and   non-linear loads in detail.e  G Another use is for similarly large Monte Carlo runs to optimise use of   power producing resources.  9 We quickly backed off VAXes when Alpha came on the scene.n    E > It looks like Fortran2000 will actually appear only in 2005 or so, VJ > probably AFTER Itanium systems which are faster than the fastest ALPHAs  > become available.  p  H Oh, Oh.  This is a problem area.  From many threads here (and I am only G on the mailing list), Alpha has been held back in capacity and Itanium s& has not yet been proven in speed, etc.  H As I said, which you quote below, our organisation is unlikely to allow @ us any more VMS boxes.  We still want to be able to improve our H performance using updated compilers even to the F95 standard.  There is > still much scope in the compilers to match the easier to read G vector?matrix operations with old-fashioned DO loops.  COMPLEX MATMUL,  " for example, is a horrendous slug.  H >                   So here the corresponding argument against having a < > new Fortran standard on older hardware is even stronger.    H Why even stronger?  The compiler code is still being developed at Intel I from our beloved folk like Steve Lionel and fellow team members for CVF. sE   It is the GEM back-end that translates the compiler into optimised n
 machine code.o  I Regardless of when the new standard comes out (you hypothesise 2005), it aI may or may not beat Itanium.  Regardless, TRs are currently being issued sG which vendors are incorporating into their F95+ compilers.  An example  I that would be pertinent to one of our applications is dynamic allocation eH of arrays in structures/types.  A colleague wrote into this application F this array of pointers thing to simulate this and the code is hard to E maintain.  (We do not like any use of pointers in our Fortran codes).e  @ I am not interested in the full wealth of F2K, just some of the H "trivial" aspects.  But I expect enhancements in vector/array handling. G   I still maintain loops which are harder to read because the compiler eH generates code many times slower for vector/array operations.  I assess C my bottleneck routines using PCA (also not too brilliant on Alpha).o  F I would also love to get compiler cross-referencing back;  SCA is not  always a panacea.w  I >                                                            Add to that fK > the fact that the lack of marketing etc has meant that academia---in the  H > past a source of number crunchers using Fortran on DEC hardware---has : > largely left DEC and the argument becomes even stronger. > D True (are you John Smith in disguise with this marketing pitch :-). F However, for people who have bought CVF licences, there is an upgrade F without money going back to the vendor -- as with all our licences on ( VMS, but we do pay software maintenance.   > I >>Since our organisation has eschewed VMS (Gartner followers, and lack ofrH >>software), if/when the IA64 thingy (copyright J-F) comes into being weH >>shall not be able to downgrade ourselves.  This means that we shall beC >>using our Alphas for as long as possible and as far as Fortran is J >>concerned we shall be stuck in maintenance mode.  TRs are coming out forG >>some of the F2K stuff, yet we shall never be able to use it.  I wouldsJ >>think that full F2K compilers will be out before the IA64 thingy, yet we5 >>shall still be on the F95 standard Alpha compiler. d >  > J > I would have thought so too, until recently, but it looks like F2K will - > be a BIG change, even bigger than F77--F90!s    I Yes, and as with much of F90, I have little use for many of the changes. rG   I have no desire to totally rewrite proven code, but when I am doing eI maintenance or a new feature, I am quite happy to use new notation since oH it is more readable, but only IF IT DOES NOT DENIGRATE PERFORMANCE IN A  BOTTLENECK ROUTINE.   I Secondary comment: Why did CXML go to Intel when they already have their o= own optimised BLA and LAPACK routines for their architecture?C   Regards, Paddy      G ***********************************************************************o  C "This electronic message and any attachments may contain privilegedl> and confidential information intended only for the use of the B addressees named above.  If you are not the intended recipient of C this email, please delete the message and any attachment and adviseoB the sender.  You are hereby notified that any use, dissemination, 7 distribution, reproduction of this email is prohibited.   A If you have received the email in error, please notify TransGrid mA immediately.  Any views expressed in this email are those of the n= individual sender except where the sender expressly and with nC authority states them to be the views of TransGrid.  TransGrid usesn> virus scanning software but excludes any liability for viruses contained in any attachment.  < Please note the email address for TransGrid personnel is now$ firstname.lastname@transgrid.com.au"  G ***********************************************************************    ------------------------------  % Date: Mon, 23 Jun 2003 14:15:05 -0400;+ From: Steve Lionel <Steve.Lionel@intel.com>mD Subject: Re: Fortran compiler was (Re: HP to port VMS to Oopsteron?)8 Message-ID: <96gefvcn15knmk0rr1radca5cssb7i0qo7@4ax.com>  1 On Fri, 20 Jun 2003 17:54:34 +1000, Paddy O'Brienr$ <paddy.o'brien@tg.nsw.gov.au> wrote:  H >Fortran is essentially in maintenance mode on VMS -- at F77 on VAX and  >at F95 on Alpha.   I I would disagree with this characterization for Alpha. (I agree fully for K VAX).  I don't speak for HP, but I'm not aware of any decision to not offer(N new Fortran features on Alpha.  I will comment that F95 is the current FortranK standard and will continue to be so for the next couple of years, at least. K HP Fortran for OpenVMS Alpha has some features from the Fortran 2000 draft, L and most language-oriented changes that go into the new Intel compilers alsoE go into the Alpha compiler.  The Alpha compiler has not been put into: "maintenance mode".v  H > From what I have read on c.l.f, the DEC Fortran compilers have always I >had a very good reputation and CVF most be at least close to the top of 3 >vendor's compilers on PCs..  / It's the best selling Fortran on that platform.o  D >Not sure why CXML had to go over as Intel has always had their own G >optimised libraries for BLAs and LAPACK.  Perhaps they were after the  H >other juicy stuff that CXML contains.  It might have been an automatic C >follow-on from giving the compiler folks away, since CXML (though uG >previously licenced) has for the past several years been bundled into   >the Fortran PCSI.  N To clarify - CXML itself did not go over to Intel.  Some of the CXML engineers did.    I >Since our organisation has eschewed VMS (Gartner followers, and lack of eH >software), if/when the IA64 thingy (copyright J-F) comes into being we H >shall not be able to downgrade ourselves.  This means that we shall be C >using our Alphas for as long as possible and as far as Fortran is  J >concerned we shall be stuck in maintenance mode.  TRs are coming out for G >some of the F2K stuff, yet we shall never be able to use it.  I would  J >think that full F2K compilers will be out before the IA64 thingy, yet we 3 >shall still be on the F95 standard Alpha compiler.s  J You already have the allocatable components TR on Alpha, and other TRs mayG still make it there.  Since F2K isn't even close to being approved as aiM standard yet - the technical work isn't even finished - I don't expect to see,L a new Fortran standard before the end of 2004 at the earliest.  F2K is a bigJ change from F95 - bigger than from F77 to F90 - and there are fewer peopleM working on Fortran compilers than there were a decade ago.  Most of the majortK vendors haven't expressed a lot of enthusiasm for implementing the full F2K 	 language.i   Steve Lionel Software Products Development) Intel Corporation0   ------------------------------    Date: 23 Jun 2003 01:36:20 -0700- From: norbert.hanowski@dillinger.de (Norbert)e4 Subject: Re: FORTRAN-Debugger more than one session?= Message-ID: <6f378eb9.0306230036.785ab8ee@posting.google.com>h   Hi,l sorry, what i mean:yE On my old machine with VMS 6.3 I start my main programm with debuggert@ and within this programm I start a second process with the "call lib$spawn" command@ On the new VMS 7.3 Alpha nothing happens when i start the secondD process. First i have to cancel the first debugger process and after ctrly I can go on.  D I think it could be a wrong debugger or system option?? Is there any options?  	 thank youg    g "labadie" <tonari_no_tottoro@127.0.0.1> wrote in message news:<IPAIa.3030$6e3.1412@news.cpqcorp.net>...t< > "Norbert" <norbert.hanowski@dillinger.de> wrote in message9 > news:6f378eb9.0306200110.56944021@posting.google.com...a > > Hi,tE > > With my old VMS 6.3-H1 (on Alpha 1000) software I first started adG > > debugger session with my main programm and then i started a furtherrH > > debugger-session after the command: CALL LIB$SPAWN ('@'//ProgNa//' ' > > ....H > > Now I am using the fortran debugger with VMS 7.3-1 ( on Alpha DS10).D > > The first session runs but after the spawn the session hangs up. > > What can I do? > >h > >d > >f > > Norbertm > 	 > post uni > $ show system 1 > to see if the process is in an intersting state  > go to sda>B > and see what the process is doing, has opened, is waiting for...   ------------------------------    Date: 23 Jun 2003 19:23:40 -0700# From: dooleys@snowy.net.au (dooley) 4 Subject: Re: FORTRAN-Debugger more than one session?= Message-ID: <1ca82fc6.0306231823.2f11af50@posting.google.com>a  r norbert.hanowski@dillinger.de (Norbert) wrote in message news:<6f378eb9.0306200110.56944021@posting.google.com>... > Hi,nC > With my old VMS 6.3-H1 (on Alpha 1000) software I first started a-E > debugger session with my main programm and then i started a furtherlF > debugger-session after the command: CALL LIB$SPAWN ('@'//ProgNa//' ' > ....F > Now I am using the fortran debugger with VMS 7.3-1 ( on Alpha DS10).B > The first session runs but after the spawn the session hangs up. > What can I do?0 Apologies if this message appears more than once2 I don't know if it is relevant in this case but...+ Have a look at the logical name DBG$PROCESSt and use " DBG> help debugging_considerations to find out what it means' Phil   ------------------------------  % Date: Mon, 23 Jun 2003 08:20:09 -0400 5 From: David Beatty <David.Beatty@qwertysasasdfgh.com>e Subject: Re: GNV and ODS-52 Message-ID: <ve=2PlPAEV3cqWxnAh+IItiY5dc+@4ax.com>  C     We have GNV V1.3-5 installed on an ODS-2 system disk on V7.3-1.h< However, we don't have BZIP2 on there.  What version are you installing?o   David R. Beattys  F On Sun, 22 Jun 2003 22:17:07 GMT, brad@.gateway.2wire.net (Bradford J. Hamilton) wrote:   >Folks,  >tI >The subject line says it all.  I attempted to install GNV on my hobbyistk@ >Alpha (V7.3-1).  The install failed with the following message: >MQ >%PCSI-E-OPENIN, error opening DISK$RABBIT_0731:[SYS0.SYSCOMMON.][GNV.SRC.GNV.BZIf$ >P2]BZIP2^.1.PREFORMATTED;* as input0 >-RMS-E-FND, ACP file or directory lookup failed. >-SYSTEM-W-BADFILEVER, bad file version number >-L >I'm assuming that because the filename referenced contains two "dots", thatK >ODS-5 is a prerequisite, even though it is not listed at the hp website as5 >such: > : >http://h71000.www7.hp.com/opensource/gnvreadme_first.html > N >Anyone know if I need ODS-5?  Anyone have any problems using ODS-5?  I recallL >that I had problems restoring an incremental backup on an ODS-5 disk a year' >or so ago, when I was "testing" ODS-5.a   ------------------------------  # Date: Mon, 23 Jun 2003 14:37:08 GMT 4 From: brad@.gateway.2wire.net (Bradford J. Hamilton) Subject: Re: GNV and ODS-5/ Message-ID: <ogEJa.112471$YZ2.279828@rwcrnsc53>a  j In article <ve=2PlPAEV3cqWxnAh+IItiY5dc+@4ax.com>, David Beatty <David.Beatty@qwertysasasdfgh.com> writes: > D >    We have GNV V1.3-5 installed on an ODS-2 system disk on V7.3-1.= >However, we don't have BZIP2 on there.  What version are you( >installing?  > This is the "latest and greatest" on the hp web site, V1.4-10:8 http://h71000.www7.hp.com/opensource/opensource.html#gnv   >e >David R. Beatty >a    A _________________________________________________________________h0 Bradford J. Hamilton			"All opinions are my own"= bMradAhamiPltSon-at-atMtAbi.cPoSm	"Lose the MAPS, and replace  					'-at-' with @"r   ------------------------------  % Date: Mon, 23 Jun 2003 20:51:43 +0200R. From: "Thomas H. Pauli" <thomaspauli@arcor.de> Subject: Re: GNV and ODS-5, Message-ID: <oprq8hchirloujp4@news.arcor.de>  ' Did you ty a $ SET PROCESS/PARSE=EXT ??$   Thomas  7 On Sun, 22 Jun 2003 22:17:07 GMT, Bradford J. Hamilton s  <brad@.gateway.2wire.net> wrote:   > Folks, >0J > The subject line says it all.  I attempted to install GNV on my hobbyistA > Alpha (V7.3-1).  The install failed with the following message:= >   > %PCSI-E-OPENIN, error opening 4 > DISK$RABBIT_0731:[SYS0.SYSCOMMON.][GNV.SRC.GNV.BZI% > P2]BZIP2^.1.PREFORMATTED;* as inputt1 > -RMS-E-FND, ACP file or directory lookup failedo/ > -SYSTEM-W-BADFILEVER, bad file version numberh > I > I'm assuming that because the filename referenced contains two "dots", v > thatJ > ODS-5 is a prerequisite, even though it is not listed at the hp website  > as > such:  > ; > http://h71000.www7.hp.com/opensource/gnvreadme_first.htmlh >0I > Anyone know if I need ODS-5?  Anyone have any problems using ODS-5?  I   > recallI > that I had problems restoring an incremental backup on an ODS-5 disk a   > year( > or so ago, when I was "testing" ODS-5. >e       -- i Thomas H. Paulib Hammersteinstr. 19 14199 Berlin Germanya   ------------------------------  % Date: Mon, 23 Jun 2003 13:32:34 -0400 5 From: David Beatty <David.Beatty@qwertysasasdfgh.com>m Subject: Re: GNV and ODS-52 Message-ID: <Pzn3PoX8jrdclcYcMC4zeMaf5Ogv@4ax.com>  F On Mon, 23 Jun 2003 14:37:08 GMT, brad@.gateway.2wire.net (Bradford J. Hamilton) wrote:  k >In article <ve=2PlPAEV3cqWxnAh+IItiY5dc+@4ax.com>, David Beatty <David.Beatty@qwertysasasdfgh.com> writes:  >>E >>    We have GNV V1.3-5 installed on an ODS-2 system disk on V7.3-1.-> >>However, we don't have BZIP2 on there.  What version are you
 >>installing?  >t? >This is the "latest and greatest" on the hp web site, V1.4-10:s9 >http://h71000.www7.hp.com/opensource/opensource.html#gnvt >c >> >>David R. Beatty, >> >  >,B >_________________________________________________________________1 >Bradford J. Hamilton			"All opinions are my own"i> >bMradAhamiPltSon-at-atMtAbi.cPoSm	"Lose the MAPS, and replace >					'-at-' with @"  C     The one we are using is on the Open Source V1.0 CD (June 2002).d   David R. Beattyf   ------------------------------  # Date: Mon, 23 Jun 2003 19:59:54 GMTt4 From: brad@.gateway.2wire.net (Bradford J. Hamilton) Subject: Re: GNV and ODS-5; Message-ID: <__IJa.2037$e26.1217@rwcrnsc52.ops.asp.att.net>c  ] In article <oprq8hchirloujp4@news.arcor.de>, "Thomas H. Pauli" <thomaspauli@arcor.de> writes:t( >Did you ty a $ SET PROCESS/PARSE=EXT ??  I I've been doing a little reading; it *looks* as though I can run that wayeN without causing damage - perhaps I'll try it soon, *after* I get a good backup of the system disk.	:-)   O I'm keeping John's caustion in mind, as well; if the risks are too great, I maya! just wait for the "fix", instead.-   >--  >Thomas H. Pauli >Hammersteinstr. 192
 >14199 Berlinm >Germany  A _________________________________________________________________o0 Bradford J. Hamilton			"All opinions are my own"= bMradAhamiPltSon-at-atMtAbi.cPoSm	"Lose the MAPS, and replace  					'-at-' with @"o   ------------------------------  % Date: Tue, 24 Jun 2003 00:05:51 -0400'  From: John Santos <JOHN@egh.com> Subject: Re: GNV and ODS-55 Message-ID: <1030624000408.2835C-100000@Ives.egh.com>t  0 On Mon, 23 Jun 2003, Bradford J. Hamilton wrote:  l > In article <ve=2PlPAEV3cqWxnAh+IItiY5dc+@4ax.com>, David Beatty <David.Beatty@qwertysasasdfgh.com> writes: > > F > >    We have GNV V1.3-5 installed on an ODS-2 system disk on V7.3-1.? > >However, we don't have BZIP2 on there.  What version are you  > >installing? > @ > This is the "latest and greatest" on the hp web site, V1.4-10:: > http://h71000.www7.hp.com/opensource/opensource.html#gnv  C I installed GMV 1.4-10 (from the web site) on two systems on 30Mar.eD Didn't have any problems.  Both system disks are ODS-5 with hardlink support.   --   John Santosv Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------    Date: 24 Jun 2003 16:24:55 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)sY Subject: GS1280 upgrade for Bank~Austria Creditanstalt's OpenVMS disaster-tolerant cluste < Message-ID: <cf15391e.0306241524.6b0b036@posting.google.com>  B Trading in the global treasury market 24 hours a day, Bank~AustriaE Creditanstalt depends on an OpenVMS disaster-tolerant cluster.  A newDF success story discusses how the next-generation EV7-based AlphaserversF cut daily financial end-of-day roll-up processing time in half and whyD the enhanced features of the OpenVMS V7.3-1 operating system make itE the ideal platform to support business-critical operations.  Customer9A executives explain how AlphaServer technology enables the bank tod% expand its business into new markets.   9 http://h71000.www7.hp.com/openvms/brochures/bank-austria/u   ------------------------------  % Date: Mon, 23 Jun 2003 08:29:36 -0400a, From: "Island" <dbturner@nospamislandco.com>( Subject: GS160/GS320 complete USD450,000/ Message-ID: <vfdsoetgnp8i12@news.supernews.com>u  & Incoming system configured as follows:  7 1. AlphaServer GS160 Tru64/TruCluster with Accessories:-  ; 2x DA-160CC-AB AlphaServer GS160 Model 16 incl. a total of: ) * 16x Alpha CPU 21264 6/731MHz, 4MB Cachea * 16x 2GB CB Memory Option0 *  4x Master PCI-Drawer (16 PCI Buses, 28 Slots)) * 12x PCI-Drawer (48 PCI Buses, 84 Slots) ! *  2x CCMHB Memory Channel II Hub,, *  4x CCMAB PCI to Memory Channel II Adapter# * 12x DEGPA-SA GigaBit Ethernet NIC-5 * 96x KGPSA-CA LP8000 FC HBA 1Gbit (6 per PCI Drawer):I *  2x 3X-DS8BA-AA GS160 Console System with US Keyboards, Mouse & Monitorc1 *  4x HA9A20 Cabinet (additional GS160 I/O Racks)e1 *  nx required MC, data, control and power cablest$ *  2x QL-6BRAQ-AA TruCluster License0 *  2x QL-0EGAQ-AA AdvFS System Utilities LicenseC *  2x QL-5U29M-3E Advanced Server for Unix 50 Client License (form.i
 PathWorks)" *  2x Tru64 V5.1b Media & Docu Kit  ! List price is almost 2Million US$o   We will let it go for $450,000 Less than 2 years oldt   Serious enquiries onlyA Product is currently in EU but can be shipped worldwide for $3000d       --   David B Turner Island Computers US Corporationi 2700 Gregory St., Suite 180a Savannah GA 31404o Tel: 912 447 6622q Fax: 912 201 0402s Email: dbturner@hpaq.net http://www.hpaq.net    ------------------------------    Date: 24 Jun 2003 08:09:30 -0700' From: hari@transcomm.uk.com (HariHaran)l4 Subject: Help Required Regarding DECNET Copy Command= Message-ID: <2de080f3.0306240709.64335003@posting.google.com>l  L We have 2 VAx 4100 and 2 Alpha DS10 machines connected to our corporate Lan.J To copy a file from ALpha to Vax I have to use the following copy command : copy *.* nodename"userid password"::disk1:[destination]*.*  T Is there any possibility to override the username and password in the above command.@ I suspect that there should be a configuration change in Decnet.
 Is that true?l Can you please advice....n Thanks in Advance.   ------------------------------  % Date: Tue, 24 Jun 2003 16:26:28 +0100e* From: "John Travell" <john@travell.uk.net>8 Subject: Re: Help Required Regarding DECNET Copy Command5 Message-ID: <bd9qj6$pthrm$1@ID-120847.news.dfncis.de>o  4 "HariHaran" <hari@transcomm.uk.com> wrote in message7 news:2de080f3.0306240709.64335003@posting.google.com...oI > We have 2 VAx 4100 and 2 Alpha DS10 machines connected to our corporater Lan.K > To copy a file from ALpha to Vax I have to use the following copy commandn< > copy *.* nodename"userid password"::disk1:[destination]*.* > G > Is there any possibility to override the username and password in thes above command.B > I suspect that there should be a configuration change in Decnet. > Is that true?n > Can you please advice....- > Thanks in Advance.   Look at DECnet proxies.   L DECnet 'knows' who you are and if a valid proxy for your user identity (bothF node and username) is found on the destination node you do not need to supply the login information.e  L If 'disk1:[destination]' is the same as sys$login for the specified usernameI on the target node you do not need to specify that detail. If you want tooD preserve the same filename you can omit that as well. So, with proxyH working, copy to the same filename in (remote) sys$login is as simple as $ copy *.* nodename::F     -- John Travell  VMS crashdump expertise for hire john@travell.uk.netA http://www.travell.uk.net/       --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.489 / Virus Database: 288 - Release Date: 10/06/2003c   ------------------------------    Date: 23 Jun 2003 21:43:11 -07000 From: Jason.Armistead@otis.com (Jason Armistead)= Subject: HELP/MESSAGE problem under SIMH with OpenVMS VAX 6.2 = Message-ID: <48ac369b.0306232043.4fabcc28@posting.google.com>    Hi  E When running SIMH V2.10-4 on Win2K (OpenVMS 6.2), I get the following  problem:  	 If I typea   $ HELP/MESSAGE NOTFOUNDe  4 I get a string of unknown error message errors, i.e.   Message number 05FD8013S Message number 05FD8023p Message number 05FD801Bs Message number 05FD8033g Message number 05FD801Bm Message number 05FD8043- Message number 05FD804B  Message number 05FD801B2 Message number 05FD8053l Message number 05FD8013u Message number 05FD800Be  F Pressing RETURN and it goes on for a few more pages before dropping me at the DCL prompt again.  4 But I can't look up those error codes anywhere (HELP3 /MESSAGE/STATUS=%x [error code]  .... doesn't work.   @ I haven't applied any sort of OpenVMS ECOs to this system.  It's out-of-the-box OpenVMS 6.2/VAX.   F Running OpenVMS Alpha 6.2-1H3 on a regular AlphaServer 400 or 1000A, I don't have this problem.   Any ideas / ECO suggestions ?o  F Perhaps someone out there with SIMH running can see if this happens to
 them also.   Cheers   Jasonh   ------------------------------  % Date: Tue, 24 Jun 2003 01:27:04 -0400   From: John Santos <JOHN@egh.com>A Subject: Re: HELP/MESSAGE problem under SIMH with OpenVMS VAX 6.2l5 Message-ID: <1030624011921.2835B-100000@Ives.egh.com>e  & On 23 Jun 2003, Jason Armistead wrote:   > Hi > G > When running SIMH V2.10-4 on Win2K (OpenVMS 6.2), I get the following 
 > problem: >  > If I typer >  > $ HELP/MESSAGE NOTFOUNDn > 6 > I get a string of unknown error message errors, i.e. >  > Message number 05FD8013e > Message number 05FD8023  > Message number 05FD801Ba > Message number 05FD8033. > Message number 05FD801Bc > Message number 05FD8043L > Message number 05FD804Bs > Message number 05FD801Bs > Message number 05FD8053  > Message number 05FD8013p > Message number 05FD800Bm > H > Pressing RETURN and it goes on for a few more pages before dropping me > at the DCL prompt again. > 6 > But I can't look up those error codes anywhere (HELP5 > /MESSAGE/STATUS=%x [error code]  .... doesn't work.e > B > I haven't applied any sort of OpenVMS ECOs to this system.  It's! > out-of-the-box OpenVMS 6.2/VAX.   A Know V6.2 bug.  Apply the ECO's.  (I think one of them fixed it.)tD Or grab the MSGHLP$ENGLISH file from V6.1.  Or install a recent VMS. V6.2 is 8 years old.  H > Running OpenVMS Alpha 6.2-1H3 on a regular AlphaServer 400 or 1000A, I > don't have this problem. >  > Any ideas / ECO suggestions ?r > H > Perhaps someone out there with SIMH running can see if this happens to > them also.   Nothing to do with SIMH.   >  > Cheers >  > Jasond >  >    -- $ John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------    Date: 24 Jun 2003 20:01:19 -07000 From: Jason.Armistead@otis.com (Jason Armistead)A Subject: Re: HELP/MESSAGE problem under SIMH with OpenVMS VAX 6.2 = Message-ID: <48ac369b.0306241901.4d90852a@posting.google.com>C  ] John Santos <JOHN@egh.com> wrote in message news:<1030624011921.2835B-100000@Ives.egh.com>...o( > On 23 Jun 2003, Jason Armistead wrote: >  > > Hi > > I > > When running SIMH V2.10-4 on Win2K (OpenVMS 6.2), I get the followingd > > problem: > > 
 > > If I typel > >  > > $ HELP/MESSAGE NOTFOUNDe > > 8 > > I get a string of unknown error message errors, i.e. > >  > > Message number 05FD8013  > > Message number 05FD8023  > > Message number 05FD801Bs > > Message number 05FD8033n > > Message number 05FD801Be > > Message number 05FD8043B > > Message number 05FD804Bi > > Message number 05FD801Bn > > Message number 05FD8053e > > Message number 05FD8013  > > Message number 05FD800B  > > J > > Pressing RETURN and it goes on for a few more pages before dropping me > > at the DCL prompt again. > > 8 > > But I can't look up those error codes anywhere (HELP7 > > /MESSAGE/STATUS=%x [error code]  .... doesn't work.  > > D > > I haven't applied any sort of OpenVMS ECOs to this system.  It's# > > out-of-the-box OpenVMS 6.2/VAX.  > C > Know V6.2 bug.  Apply the ECO's.  (I think one of them fixed it.) F > Or grab the MSGHLP$ENGLISH file from V6.1.  Or install a recent VMS. > V6.2 is 8 years old.  D How silly of me not to RTFNG (newsgroup) archives before firing thisB question off.  I did that almost as soon as I'd hit send, and knew0 that someone would point this out to me already.  D I couldn't find any reference to an ECO, no even on HP's own OpenVMSD support sites, but did get the V6.1 file OK from a backup I had from- my MicroVAX 2000 (gotta love my first VAX !).   D Agree that V6.2 is a bit long in the tooth these days, but it's warmB and familiar, and besides, with SIMH, I really want my PC spendingD more time running VAX applications and doing it as fast as possible,C rather than running the latest and greatest VMS 7.3 with no doubt ae; bit more OS overhead and thus less performance that VMS 6.2e   J > > Running OpenVMS Alpha 6.2-1H3 on a regular AlphaServer 400 or 1000A, I > > don't have this problem. > > ! > > Any ideas / ECO suggestions ?r > > J > > Perhaps someone out there with SIMH running can see if this happens to > > them also. >  > Nothing to do with SIMH. >  > > 
 > > Cheers > > 	 > > Jasond > >  > >v   ------------------------------   Date: 24 Jun 2003 23:10 CDT$' From: carl@gerg.tamu.edu (Carl Perkins)UA Subject: Re: HELP/MESSAGE problem under SIMH with OpenVMS VAX 6.2s- Message-ID: <24JUN200323102000@gerg.tamu.edu>d  4 Jason.Armistead@otis.com (Jason Armistead) writes...E }Agree that V6.2 is a bit long in the tooth these days, but it's warmsC }and familiar, and besides, with SIMH, I really want my PC spendingwE }more time running VAX applications and doing it as fast as possible,iD }rather than running the latest and greatest VMS 7.3 with no doubt a< }bit more OS overhead and thus less performance that VMS 6.2   You are probably mistaken.  A Newer versions of VMS tend to be faster than older versions. TheyeD do also tend to use slightly more memory (or more than just slightlyD when you consider file caching), but PC memory is very cheap so thatB is not likely to be a problem when using an emulated VAX. It's not= like Windows where each new version needs twice a much as the > previous version and usually runs slower on the same hardware.   --- Carl   ------------------------------  % Date: Mon, 23 Jun 2003 16:32:24 -0400  From: norm.raphael@metso.comY Subject: Hewlett-Packard said to be interested in buying software and services companies. ? Message-ID: <OF07350873.0105A04A-ON85256D4E.00709247@metso.com>o   Is HP going shopping?-H Hewlett-Packard is said to be interested in buying software and service= si
 companies. June 23, 2003: 1:49 PM EDT- By Paul R. La Monica, CNN/Money Senior Writero    F NEW YORK (CNN/Money) - A little more than a year since Hewlett-PackardH completed its purchase of Compaq, the company may be gearing up for mor= ei deals.    D Such rumors have been circulating since CEO Carly Fiorina said at anF analysts meeting earlier this month that services and software are twoC areas where the company might look to acquire companies. HP was nots- available for further comment for this story.e    B Wall Street has warmed to HP (HPQ) in recent months as the companyD continues to narrow losses in its enterprise division -- which sellsF storage and servers -- and makes improvements in its personal computer8 segment, which competes aggressively with Dell Computer.    H Still, further moves to diversify away from these two notoriously tough=  H businesses would probably be welcomed. "Anything HP can do to move away=  H from the eroding margin business of hardware would make sense," said Ad= amH Adelman, senior technology analyst with Philippe Asset Management, a Ne= w D York-based asset management firm. He does not have a position in HP.    /                             A "bear" of a deal?i    H Fiorina already has tried to make a big move into services, when it off= eredH to buy the consulting business of PricewaterhouseCoopers in 2000. The t= woD companies could not agree on a price. IBM wound up scooping up PwC's consulting division last year.    H HP increased its presence in the services business through the Compaq d= eal H but it still has a long way to go to catch IBM, said John Rutledge, man= agerB of the Evergreen Technology fund, which owns shares of HP and IBM.    H The services business is attractive because of its relative stability -= -hH big companies typically sign multi-year outsourcing contracts, such as = HP'sH recent $3 billion, 10-year deal to run Procter & Gamble's IT services -= -sH and it could also give HP the opportunity to sell more of its computers= ,n servers and printers.e    H "HP would benefit from a services acquisition. It's a great entr=E9e fo= re$ additional business," Rutledge said.    H But Rutledge said he doubts that HP would want to do a truly large deal=  sooH soon after the Compaq merger. So that probably would rule out companies=  H like the struggling EDS, which has an $11 billion market value, as well=  asoC other major players in tech services, including Affiliated ComputeraH Services ($6.4 billion market cap), Computer Sciences ($7.8 billion) an= d-' Accenture ($16.8 billion market value).i    C Still, Rutledge said that BearingPoint (BE), formerly known as KPMG-H Consulting, could be an interesting takeover candidate since its market=  E value is less than $2 billion. A spokesman for BearingPoint would not  comment.    0                           Searching for software    < On the software side, don't expect HP to get involved in theH PeopleSoft-Oracle fracas, even though PeopleSoft disclosed in a regulat= ory3D filing Friday that it would consider a so-called "white knight" bid.    H Robert Cihra, an analyst with Fulcrum Global Partners, said HP would be=  H unlikely to buy a major application software company like PeopleSoft, w= hichH sells products that help companies manage their customers, supplier and=  H human resource functions. He doesn't own the stock and Fulcrum does not=   perform investment banking.p    D But Cihra said a storage software company or infrastructure softwareC company would be a good fit with HP's enterprise business division.r    H Storage software developer Legato Systems (LGTO) could be a fairly easy=  H deal for HP to do, according to Michael Mahoney, managing director of E= GMH Capital, which runs a hedge fund focusing on technology, telecom and me= dian$ companies. Legato would not comment.    C Mahoney pointed out that HP already is Legato's largest reseller ofiD software so the companies are fairly familiar with each other. Plus,E Legato's market value is only $887 million. Legato was the subject ofEH takeover rumors earlier this year. EMC, a competitor of HP in the stora= ge# business, was the mentioned suitor.u    H BEA Systems, which develops infrastructure software that helps companie= smH manage e-commerce capabilities, has been mentioned as a possible target=  forH HP as well. Rutledge said BEA (BEAS), which competes against IBM's soft= wareH division, could be a target since it has struggled as of late. The comp= anysH reported in May that its first quarter licensing revenue fell from a ye= ar ago.    H And despite the big tech rally, shares of BEA are down 4.5 percent year=  tokH date. Still, BEA would be a bigger deal for HP to digest, with a market=  D value of nearly $4.5 billion. BEA did not return a call for comment.    E HP's stock has surged 42 percent since March 11, which could make its H shares more attractive to takeover targets. Plus, the company had $13.7=  H billion in cash on its balance sheet as of April 30. So even though HP = maybH not be itching to a deal just yet, it clearly has the financial strengt= h to pull something off.r     Find this article at: 7 http://money.cnn.com/2003/06/23/technology/hp/index.htmt     =h   ------------------------------  # Date: Wed, 25 Jun 2003 02:09:07 GMTt# From: "John Smith" <a@nonymous.com>MY Subject: Re: Hewlett-Packard said to be interested in buying software and services compansG Message-ID: <7v7Ka.12364$O31.1831@news02.bloor.is.net.cable.rogers.com>a  C I think buying an advertising agency that has a clue would be moneyl
 better spent.d    ) <norm.raphael@metso.com> wrote in messageo9 news:OF07350873.0105A04A-ON85256D4E.00709247@metso.com...k     Is HP going shopping?m? Hewlett-Packard is said to be interested in buying software andn services
 companies. June 23, 2003: 1:49 PM EDT- By Paul R. La Monica, CNN/Money Senior Writer     F NEW YORK (CNN/Money) - A little more than a year since Hewlett-PackardC completed its purchase of Compaq, the company may be gearing up forl more deals.    D Such rumors have been circulating since CEO Carly Fiorina said at anF analysts meeting earlier this month that services and software are twoC areas where the company might look to acquire companies. HP was not - available for further comment for this story.t    B Wall Street has warmed to HP (HPQ) in recent months as the companyD continues to narrow losses in its enterprise division -- which sellsF storage and servers -- and makes improvements in its personal computer8 segment, which competes aggressively with Dell Computer.    A Still, further moves to diversify away from these two notoriously  toughvB businesses would probably be welcomed. "Anything HP can do to move awayD from the eroding margin business of hardware would make sense," said AdamD Adelman, senior technology analyst with Philippe Asset Management, a NewtD York-based asset management firm. He does not have a position in HP.    /                             A "bear" of a deal?     C Fiorina already has tried to make a big move into services, when iti offered:E to buy the consulting business of PricewaterhouseCoopers in 2000. Ther two D companies could not agree on a price. IBM wound up scooping up PwC's consulting division last year.    E HP increased its presence in the services business through the Compaqe dealC but it still has a long way to go to catch IBM, said John Rutledge,s managernB of the Evergreen Technology fund, which owns shares of HP and IBM.    ; The services business is attractive because of its relative  stability --F big companies typically sign multi-year outsourcing contracts, such as HP's< recent $3 billion, 10-year deal to run Procter & Gamble's IT services -->= and it could also give HP the opportunity to sell more of itsy
 computers, servers and printers.9    F "HP would benefit from a services acquisition. It's a great entre for$ additional business," Rutledge said.    B But Rutledge said he doubts that HP would want to do a truly large deal so = soon after the Compaq merger. So that probably would rule oute	 companies B like the struggling EDS, which has an $11 billion market value, as well asmC other major players in tech services, including Affiliated ComputerlD Services ($6.4 billion market cap), Computer Sciences ($7.8 billion) andd' Accenture ($16.8 billion market value).n    C Still, Rutledge said that BearingPoint (BE), formerly known as KPMGi@ Consulting, could be an interesting takeover candidate since its marketE value is less than $2 billion. A spokesman for BearingPoint would notd comment.    0                           Searching for software    < On the software side, don't expect HP to get involved in the? PeopleSoft-Oracle fracas, even though PeopleSoft disclosed in al
 regulatoryD filing Friday that it would consider a so-called "white knight" bid.    D Robert Cihra, an analyst with Fulcrum Global Partners, said HP would beE unlikely to buy a major application software company like PeopleSoft,T whichdC sells products that help companies manage their customers, suppliern andiC human resource functions. He doesn't own the stock and Fulcrum does  nots perform investment banking.s    D But Cihra said a storage software company or infrastructure softwareC company would be a good fit with HP's enterprise business division.d    B Storage software developer Legato Systems (LGTO) could be a fairly easyE deal for HP to do, according to Michael Mahoney, managing director of  EGM-D Capital, which runs a hedge fund focusing on technology, telecom and media $ companies. Legato would not comment.    C Mahoney pointed out that HP already is Legato's largest reseller ofeD software so the companies are fairly familiar with each other. Plus,E Legato's market value is only $887 million. Legato was the subject ofrA takeover rumors earlier this year. EMC, a competitor of HP in the7 storagen# business, was the mentioned suitor.a    > BEA Systems, which develops infrastructure software that helps	 companiess@ manage e-commerce capabilities, has been mentioned as a possible
 target forB HP as well. Rutledge said BEA (BEAS), which competes against IBM's softwareB division, could be a target since it has struggled as of late. The companyrD reported in May that its first quarter licensing revenue fell from a year ago.    B And despite the big tech rally, shares of BEA are down 4.5 percent year top@ date. Still, BEA would be a bigger deal for HP to digest, with a marketD value of nearly $4.5 billion. BEA did not return a call for comment.    E HP's stock has surged 42 percent since March 11, which could make itsfA shares more attractive to takeover targets. Plus, the company hadn $13.7lF billion in cash on its balance sheet as of April 30. So even though HP mayw? not be itching to a deal just yet, it clearly has the financiala strength too pull something off.      Find this article at:v7 http://money.cnn.com/2003/06/23/technology/hp/index.htm    ------------------------------  % Date: Tue, 24 Jun 2003 18:39:49 +0100v/ From: Tim ffrench-Lynch <nojunk@baesystems.com>. Subject: Hoobyist licence wait. Message-ID: <3EF88CE5.DB6725BD@baesystems.com>  F How long does it usually take for Mantagar to recognise a new EncompasH US membership number for getting a hobbyist licence? It's been more thanC a week now and I'm starting to wonder at what point to chase it up.s   Time   ------------------------------  % Date: Tue, 24 Jun 2003 21:13:59 +0200  From: "B.Eckstein" <eck@ivu.de>k" Subject: Re: Hoobyist licence wait- Message-ID: <bda7tq$3q6$3@swifty.westend.com>l  3 Tim ffrench-Lynch schrub im Jahre 24.06.2003 19:39:FH > How long does it usually take for Mantagar to recognise a new EncompasJ > US membership number for getting a hobbyist licence? It's been more thanE > a week now and I'm starting to wonder at what point to chase it up.   H My first Lic with german DECUS took two weeks, AFAIR. But this was some G years back. Sunday it took less than a minute to get a new license for rH my new AS1200. I think the first shot always could take a little longer.     -- mH B.Eckstein, be@epost.de         Cheap, Fast, Good - pick any two of them9 Die FAQ zu de.comp.hardware.netzwerke: http://how.to/dchneE Tips und Tricks zum Mozilla und Netscape 6/7: http://www.hmetzger.de/aH I want your VAX http://monster-island.org/tinashumor/humor/iwantyou.html   ------------------------------    Date: 24 Jun 2003 15:28:18 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)j- Subject: HP Powers More TOP500 Supercomputers?= Message-ID: <cf15391e.0306241428.2185fab5@posting.google.com>h  A "For the third consecutive time, HP (NYSE:HPQ) ranks as the No. 1tB supercomputing provider on the TOP500 Supercomputer list,(1) whichD categorizes the 500 most powerful installed technical and commercial computer systems worldwide.   A With 159 entries, representing more than 30 percent of the postedpB sites, HP has more installations on the TOP500 list than any otherE technology supplier. Led by its powerful HP Superdome and AlphaServerlE systems, HP not only increased its share on this exclusive list since$E its November 2002 showing of 137 entries, but once again secured four. of the top 10 spots."  ...6D "HP Superdome systems accounted for 127 of the company's entries. HPC AlphaServer systems represented 21 entries -- three of those in theeE top 10. Eight sites were powered by Intel Itanium 2-based systems -- D including an HP rx2600 server cluster installed at Pacific Northwest/ National Laboratory that placed in the top 10."   9 http://www.hp.com/hpinfo/newsroom/press/2003/030623b.htmlb   ------------------------------  # Date: Wed, 25 Jun 2003 02:05:57 GMT # From: "John Smith" <a@nonymous.com> 1 Subject: Re: HP Powers More TOP500 SupercomputerstF Message-ID: <9s7Ka.12345$O31.670@news02.bloor.is.net.cable.rogers.com>  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0306241428.2185fab5@posting.google.com...(C > "For the third consecutive time, HP (NYSE:HPQ) ranks as the No. 1gD > supercomputing provider on the TOP500 Supercomputer list,(1) whichF > categorizes the 500 most powerful installed technical and commercial > computer systems worldwide.. > C > With 159 entries, representing more than 30 percent of the postedeD > sites, HP has more installations on the TOP500 list than any other; > technology supplier. Led by its powerful HP Superdome and( AlphaServer'A > systems, HP not only increased its share on this exclusive listo since B > its November 2002 showing of 137 entries, but once again secured four > of the top 10 spots."g > ...tA > ".... HP AlphaServer systems represented 21 entries -- three ofi thosec >  in the top 10.l; > http://www.hp.com/hpinfo/newsroom/press/2003/030623b.htmlu  > Not bad for a dead processor. Imagine (where have I heard that= before....) what further development work on Alpha could have 8 achieved.....and what artificially 'holding it back' has4 prevented.....a clean sweep of the top 10 positions.  C Nope....can't have Alpha looking too good...makes IA-64 and PA-Risc2 look like yesterday's news.e   ------------------------------    Date: 24 Jun 2003 22:02:35 -0700' From: icerq4a@spray.se (David Svensson)01 Subject: Re: HP Powers More TOP500 Supercomputers.= Message-ID: <734da31c.0306242102.250ebc97@posting.google.com>l  q "John Smith" <a@nonymous.com> wrote in message news:<9s7Ka.12345$O31.670@news02.bloor.is.net.cable.rogers.com>... @ > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message9 > news:cf15391e.0306241428.2185fab5@posting.google.com...ME > > "For the third consecutive time, HP (NYSE:HPQ) ranks as the No. 1dF > > supercomputing provider on the TOP500 Supercomputer list,(1) whichH > > categorizes the 500 most powerful installed technical and commercial > > computer systems worldwide.5 > >cE > > With 159 entries, representing more than 30 percent of the posted F > > sites, HP has more installations on the TOP500 list than any other= > > technology supplier. Led by its powerful HP Superdome and  >  AlphaServerC > > systems, HP not only increased its share on this exclusive list- >  sinceD > > its November 2002 showing of 137 entries, but once again secured >  fourr > > of the top 10 spots.": > > ...OC > > ".... HP AlphaServer systems represented 21 entries -- three of< >  those > >  in the top 10.x= > > http://www.hp.com/hpinfo/newsroom/press/2003/030623b.html  > @ > Not bad for a dead processor. Imagine (where have I heard that? > before....) what further development work on Alpha could have(: > achieved.....and what artificially 'holding it back' has6 > prevented.....a clean sweep of the top 10 positions. > E > Nope....can't have Alpha looking too good...makes IA-64 and PA-Risc_ > look like yesterday's news._  , It seems to be impossible to be positive? ;)   SGI makes good Itanium boxes...aJ http://www.sgi.com/newsroom/press_releases/2003/june/altix_benchmarks.htmlI http://www.sgi.com/newsroom/press_releases/2003/june/altix_benchmarks.pdfT  A HP have a Madison SuperDome that also look pretty good on TOP500.e   /David   ------------------------------    Date: 23 Jun 2003 07:43:03 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)d Subject: Re: HP to drop hpux?23 Message-ID: <vkxo0Fe$VwNR@eisner.encompasserve.org>s  h In article <d7791aa1.0306191512.61365658@posting.google.com>, bob@instantwhip.com (Bob Ceculski) writes:? > it says here that its a rumor ... maybe Carly is smarter thanI > we think! :) > ? > http://www.infoworld.com/article/03/06/18/HNsapfiorina_1.htmlR  >    Yep, and Microsoft is about to drop Windows, but Bill Gates!    keeps denying that rumor, too.    ------------------------------  % Date: Mon, 23 Jun 2003 21:29:09 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net>s Subject: Re: HP to drop hpux? ' Message-ID: <3EF7B775.7AD9C9D6@fsi.net>h   David Froble wrote:d > [snip]< > Bill, are you saying that these people are (gasp!), liars? > P > I guess that if COMMITMENTS don't mean a thing, then breaking promises isn't a0 > problem either.  (Pretty much the same thing.) > B > Maybe it's a memory thing.  They just forget about the promises.  @ I recall an old ad campaign from misty moons ago... "Promise herB anything, but giver her Arpege." Compaq (and now HP?) seem to haveF adapted that: "Promise them anything, just give 'em VMS, then give 'em the shaft".4  A > You don't think it has anything to do with comparison to IA-64?.  	 Well, ...a   -- . David J. Dachtera  dba DJE Systems= http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Mon, 23 Jun 2003 13:10:16 +0100eO From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> ? Subject: Re: Hp warns about Suns future ... or lack thereof ...:0 Message-ID: <bd6qnd$fno$2@new-usenet.uk.sun.com>   Bob Ceculski wrote:B7 > it's right here ... Sun has also made a bad choice in_7 > betting againset the alpha engineers and not choosingl7 > Itanium ... this will be the nail in their coffin ...m > + > http://www.theinquirer.net/?article=10064"    3 So you didn't read the right down to the end of theD article did you !!!i    I "It wants to know dozens of answers to other questions too, but all this 5H demonstrates to us is that the bitter fight for share in a strangulated E IT business is causing the HP competitive marketing folk to slightly n	 lose it."a   regardsq Andrew Harrison.   ------------------------------    Date: 23 Jun 2003 10:08:14 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) 8 Subject: Re: IBM won't buy VMS ... has NIH syndrome too!= Message-ID: <cf15391e.0306230908.531f5dfe@posting.google.com>g  d "Bill Todd" <billtodd@metrocast.net> wrote in message news:<4QqdnZs0KMh2Fm6jXTWJkg@metrocast.net>...N > Does anyone happen to know what's happening to the the Compaq storage peopleM > these days?  Richie left - and my impression is was encouraged to leave, as_N > part of a general cut-back in senior people to save short-term cash - before
 > the merger.M  D Compaq's storage people definitely did NOT want Richie Lary to leaveD -- they even convinced him to do consulting for Compaq after he went out on his own.   C Richie had worked for DEC-classic for a very long time (he was evenD= involved in the original VAX/VMS architecture team).  I'm not2@ surprised he wanted a change of pace, and a bit of rest from theE constant pressure.  His technical expertise and knowledge of industrywE trends and directions is certainly widely recognized in the industry.s  D Steve Sicola tried to fill Richie's shoes after he left, but I thinkB he got worn out in the attempt -- Richie is a tough act to follow.  D But the Storage organization isn't dependent on any single person --. there are many brilliant people working there.   ------------------------------  # Date: Mon, 23 Jun 2003 21:26:29 GMTt# From: "John Smith" <a@nonymous.com>p8 Subject: Re: IBM won't buy VMS ... has NIH syndrome too!F Message-ID: <9gKJa.2376$H9q1.394@news04.bloor.is.net.cable.rogers.com>  > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message7 news:cf15391e.0306230908.531f5dfe@posting.google.com...a7 > "Bill Todd" <billtodd@metrocast.net> wrote in message3. news:<4QqdnZs0KMh2Fm6jXTWJkg@metrocast.net>...A > > Does anyone happen to know what's happening to the the Compaqr storage peopleE > > these days?  Richie left - and my impression is was encouraged toc	 leave, ashB > > part of a general cut-back in senior people to save short-term
 cash - beforeo > > the merger.a >aF > Compaq's storage people definitely did NOT want Richie Lary to leaveF > -- they even convinced him to do consulting for Compaq after he went > out on his own.p >nE > Richie had worked for DEC-classic for a very long time (he was evenn? > involved in the original VAX/VMS architecture team).  I'm notaB > surprised he wanted a change of pace, and a bit of rest from the> > constant pressure.  His technical expertise and knowledge of industry= > trends and directions is certainly widely recognized in the@	 industry.e >cF > Steve Sicola tried to fill Richie's shoes after he left, but I thinkD > he got worn out in the attempt -- Richie is a tough act to follow. >dF > But the Storage organization isn't dependent on any single person --0 > there are many brilliant people working there.  D It may be time to get some of those brilliant people to take a break= from Engineering and try VMS Marketing & Advertising instead.n   ------------------------------  % Date: Mon, 23 Jun 2003 21:23:37 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> 8 Subject: Re: IBM won't buy VMS ... has NIH syndrome too!' Message-ID: <3EF7B629.5C536203@fsi.net>O   John Smith wrote:r > @ > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message9 > news:cf15391e.0306230908.531f5dfe@posting.google.com... 9 > > "Bill Todd" <billtodd@metrocast.net> wrote in message(0 > news:<4QqdnZs0KMh2Fm6jXTWJkg@metrocast.net>...C > > > Does anyone happen to know what's happening to the the Compaq4 > storage peopleG > > > these days?  Richie left - and my impression is was encouraged ton > leave, asoD > > > part of a general cut-back in senior people to save short-term > cash - beforeG > > > the merger.8 > >aH > > Compaq's storage people definitely did NOT want Richie Lary to leaveH > > -- they even convinced him to do consulting for Compaq after he went > > out on his own.- > > G > > Richie had worked for DEC-classic for a very long time (he was evengA > > involved in the original VAX/VMS architecture team).  I'm nottD > > surprised he wanted a change of pace, and a bit of rest from the@ > > constant pressure.  His technical expertise and knowledge of
 > industry? > > trends and directions is certainly widely recognized in thek > industry.n > > H > > Steve Sicola tried to fill Richie's shoes after he left, but I thinkF > > he got worn out in the attempt -- Richie is a tough act to follow. > >.H > > But the Storage organization isn't dependent on any single person --2 > > there are many brilliant people working there. > F > It may be time to get some of those brilliant people to take a break? > from Engineering and try VMS Marketing & Advertising instead.r  H ...or maybe SysAdmin'ing, so they'll understand that storage controllers: that cannot be commanded from a VMS batch job are useless.   -- o David J. Dachtera1 dba DJE Systems1 http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/l   ------------------------------  + Date: Tue, 24 Jun 2003 15:58:54 +0100 (MET)*9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com>n) Subject: is there a problem with Info-VAXp; Message-ID: <01KXH8HBZRJIAOKN0V@sysdev.deutsche-boerse.com>     is there a problem with Info-VAX   ------------------------------  + Date: Tue, 24 Jun 2003 22:39:54 +0000 (UTC)yP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)* Subject: is there a problem with Info-VAX?$ Message-ID: <bdajvp$v1q$1@online.de>  I Is there a problem with Info-VAX?  I'm subscribed at work and didn't get   any messages today.u   ------------------------------    Date: 24 Jun 2003 14:14:54 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)y= Subject: Josh Fisher of HP Labs receives Eckert-Mauchly Awards= Message-ID: <cf15391e.0306241314.3c7cdace@posting.google.com>r  B The Association for Computing Machinery (www.acm.org) and the IEEEF Computer Society (www.computer.org) presented the Eckert-Mauchly Award? to HP Labs senior fellow Joseph A. (Josh) Fisher of HP Labs "in/E recognition of 25 years of seminal contributions to instruction-level-; parallelism, pioneering work on VLIW architectures, and ther; formulation of the Trace Scheduling compilation technique."b  - http://www.acm.org/announcements/em_2003.html    ------------------------------    Date: 24 Jun 2003 11:57:54 -0700' From: mark_tarka@yahoo.com (Mark Tarka)  Subject: Laptop recommendation?f= Message-ID: <6b70c71c.0306241057.4f6ad5bb@posting.google.com>w   So...maybe it's off-topic....e  . Panasonic has a ruggedized unit, in particular1 a Toughbook 28 -- Pentium III with PC 133 memory. / I'm currently living the life of leisure on the-. streets of NYC since my sister and some looney0 feminist nuked my crib in Montana (a reminder of- my good buddy Murphy -- some people really doT4 have shit for brains; Hey!  Did you know the grocery. stores will buy back aluminum cans at a nickel0 a pop :-)  Homelessness doesn't necessarily mean- lack of WWW connectivity -- libraries provide / hardware and limited access time.  Here in NYC,s- you can get a seat at a port if your laptop's 1 got the right card, and, WiFi zones are coming upe. like mosquitos after the spring rains all over/ Manhattan.  Starbuck's has ports available, butn0 who really likes coffee and the evil people that use it?h  1 The point -- A Toughbook would run something like-/ $4K with options, more with insurance.  Laptop/ 3 notebooks come in at half that, with faster memory,/7 and I could put one into a Pelican water/dust resistant-1 hard case.  What I'd lose would be the supposedlyn4 specialized shock-mounting Panasonic claims for it's0 toys (seems to me if a HDD is rated at resisting4 x Gees out of the box, why shock-mount -- comments?)  1 Anyone with experience along this line...a budgetn. minded system for _rough_ service; or have the1 big-boys clinched the market specialized featuress0 like sealed KBs and displays that don't crack at3 the drop of a hat and can be seen in full sunlight?L        Marku  4 FWIW, "shit for brains" is a very utilitarian phrase6 and extrordinarily effective in the right hands.  Over2 on sci.chem a lurker/small-time poster suprised us6 with a plagiarized rant (search keywords to find other8 variations)...appended.  It's been a month.  I've gotten6 over it and would almost thank the poster for pointing7 out that the text and the underlying thought existed --M4 but why mess with something that worked in the first place.     ***w  ' From: chang_edison@hotmail.com (EdiSon)o Newsgroups: sci.chem) Subject: Re: Negative expansion materialsr  Date: 24 Apr 2003 03:48:41 -0700' Organization: http://groups.google.com/-	 Lines: 31g= Message-ID: <60c02aa9.0304240248.54a29b06@posting.google.com> - References: <3EA703A4.ECF445C0@hate.spam.net>p! NNTP-Posting-Host: 203.58.230.170n, Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 8bit P X-Trace: posting.google.com 1051181322 1879 127.0.0.1 (24 Apr 2003 10:48:42 GMT)( X-Complaints-To: groups-abuse@google.com+ NNTP-Posting-Date: 24 Apr 2003 10:48:42 GMT7    ( Uncle Al <UncleAl0@hate.spam.net> wrote:   Nothing.   :  5 I cannot believe how incredibly stupid Uncle Al is.   F I mean rock-hard stupid.  Blazing hot mid-day sun on Mercury stupid.  C Surface of Venus under 80 atmospheres of red hot carbon dioxide and0 sulfuric? acid vapor dehydrated for 300 million years rock-hard stupid.   B Stupid so stupid that it goes way beyond the stupid we know into a wholew? different sensorium of stupid. Uncle Al is trans-stupid stupid.EC Meta-stupid.  Stupid so collapsed upon itself that it is within its E own Schwarzschild radius.  Black hole stupid.  Stupid gotten so dense D and massive that no intellect can escape.  Singularity stupid. UncleB Al emits more stupid/second than our entire galaxy otherwise emitsF stupid/year.  Quasar stupid.  Nothing else in the universe can be thisD stupid.  Uncle Al is an oozingly putrescent primordial fragment from2 the original Big Bang of Stupid, a pure essence ofE stupid so uncontaminated by anything else as to be beyond the laws of 9 physics that define maximally extrapolated hypergeometricUF n-dimensional backgroundless stupid as we can imagine it.  Uncle Al isD Planck stupid, a quantum foam of stupid, a vacuum decay of stupid, aD grand unified theory of stupid. Uncle Al is the epiphany of stupid.  Uncle Al is stooopid.aF Uncle Al is a swine. Uncle Al is a vulgar little maggot. Uncle Al is aF worthless bag of filth.  As they say in Texas,  Uncle Al couldn't pourE piss out of a boot with instructions printed on the heel. Uncle Al isdD a canker, an ulcerous sore that won't go away.  I would rather watchB Affirmative Action lawyers suck-start Harleys with their bungholes/ than suffer the troll abominations of Uncle Al.c   ------------------------------  % Date: Tue, 24 Jun 2003 20:29:52 -0500t1 From: "David J. Dachtera" <djesys.nospam@fsi.net>'# Subject: Re: Laptop recommendation?t' Message-ID: <3EF8FB10.B996314B@fsi.net>e   Mark Tarka wrote:l >  > So...maybe it's off-topic....i > 0 > Panasonic has a ruggedized unit, in particular3 > a Toughbook 28 -- Pentium III with PC 133 memory.e1 > I'm currently living the life of leisure on theM0 > streets of NYC since my sister and some looney' > feminist nuked my crib in Montana ...   C Sorry - this is so far outside my paradigm that my brain bugcheckedr5 (some would say I *AM* a bugcheck, but I digress...).m  @ Would not the money be better spent on shelter? ...a job search?   Re: Bugcheck...   H I'm reminded of a story from years ago. A manframe jock was overheard onG a cell phone at a social gathering. He drew a lot of strange looks wheneG he was heard to tell the person (an operator) on the other end to "takeu& a dump, and I'll look at it tomorrow".   -- ) David J. Dachterat dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------    Date: 24 Jun 2003 09:51:41 -07005 From: forrest.cahoon@merrillcorp.com (Forrest Cahoon)a+ Subject: Re: lib$spawn and persona servicese= Message-ID: <8369d643.0306240851.64ed017b@posting.google.com>y  s JONESD@er6.eng.ohio-state.edu (David Jones) wrote in message news:<bctam9$il3$1@charm.magnus.acs.ohio-state.edu>...oG > If you want to use minimum privileges, create a rights identifier foreO > each process slot for each process index (e.g. IMPERSONATE_JT_ACCESS_nnn) andVQ > then set the ACL to allow access for the identifier that corresponds to the joboQ > master process index.  Then when you create the persona, use ISS$_ADD_RIGHTS to-2 > include that identifier in the security profile.  D I'm getting somewhere with this approach, but I can't figure out howF to add rights with ISS$_ADD_RIGHTS successfully.  The documentation is incredibly vague on this.6  E The data structure which you're supposed to point to when you use theiC ISS$_ADD_RIGHTS code in your itmlst appears to be illustrated underu ISS$_RIGHTS.  C There are pairs of longs identified as (Id value / Id flags) for asKE many identifiers as you want to add.  The "Id value" should certainlyeD be the binary value returned by sys$asctoid(), but what the heck areC "Id flags"? I have no idea where the heck to find that information.   A I've been taking potshots at it, but obviously I've been guessingtD wrong, because when I run "show process /priv /id=<the impersonating/ process id>" I get a rights list that ends withh  , %SYSTEM-F-IVIDENT, invalid identifier format  = Does anyone know where I can find this info?  Has anyone used F sys$persona_create with the ISS$_ADD_RIGHTS itmlst entry successfully?   (The doc I'm looking at is atpE http://h71000.www7.hp.com/doc/731FINAL/4527/4527pro_084.html#mar_106)l   ------------------------------   Date: 25 Jun 2003 03:45:39 GMT1 From: JONESD@er6.eng.ohio-state.edu (David Jones)g+ Subject: Re: lib$spawn and persona servicesm: Message-ID: <bdb5t3$f81$1@charm.magnus.acs.ohio-state.edu>  = In message <8369d643.0306240851.64ed017b@posting.google.com>,,9   forrest.cahoon@merrillcorp.com (Forrest Cahoon) writes:aE >I'm getting somewhere with this approach, but I can't figure out howlG >to add rights with ISS$_ADD_RIGHTS successfully.  The documentation isi >incredibly vague on this.  F I believe your item list must also specify an ISS$_ENABLED item with aB mask of ISS$M_ENABLED_PERSONA to actually make the process use the per-persona rights..  D >There are pairs of longs identified as (Id value / Id flags) for asF >many identifiers as you want to add.  The "Id value" should certainlyE >be the binary value returned by sys$asctoid(), but what the heck areeD >"Id flags"? I have no idea where the heck to find that information.  K Look at the docs for SYS$GRANTID().  Using 0 is suffcient for simple accessn checks.t   >Has anyone usedG >sys$persona_create with the ISS$_ADD_RIGHTS itmlst entry successfully?    Yes.      < David L. Jones               |      Phone:    (614) 292-6929- Ohio State University        |      Internet:eL 140 W. 19th St. Rm. 231a     |               jonesd@er6s1.eng.ohio-state.edu: Columbus, OH 43210           |               vman+@osu.edu  1 Disclaimer: I'm looking for marbles all day long.a   ------------------------------  % Date: Tue, 24 Jun 2003 13:36:57 +0100t From: paul.beaudoin@hsbc.com  Subject: Memo:  Re:  Memo:  RantE Message-ID: <OF37958C28.B2D127D0-ON80256D4F.0045311B@systems.uk.hsbc>m   Paul  F Thanks for the link to Hunter's page (and more to the point, thanks to Hunter).   Paul          ' ** HSBC's website is at www.hsbc.com **d  D ********************************************************************A This message and any attachments are confidential to the ordinary A user of the e-mail address to which it was addressed and may also = be privileged. If you are not the addressee you may not copy,e7 forward, disclose or use any part of the message or itseB attachments and if you have received this message in error, pleaseA notify the sender immediately by return e-mail and delete it fromn
 your system."   < Internet communications cannot be guaranteed to be secure or@ error-free as information could be intercepted, corrupted, lost,= arrive late or contain viruses. The sender therefore does notn> accept liability for any errors or omissions in the context of> this message which arise as a result of Internet transmission.  C Any opinions contained in this message are those of the author and t> are not given or endorsed by the HSBC Group company or office < through which this message is sent unless otherwise clearly @ indicated in this message and the authority of the author to so 2 bind the HSBC entity referred to is duly verified.  D ********************************************************************   ------------------------------  # Date: Mon, 23 Jun 2003 11:32:50 GMTh' From: Colin Blake <colin@theblakes.com>r$ Subject: Re: Mozilla on old hardware, Message-ID: <3EF6E560.7090109@theblakes.com>   Richard Brodie wrote:7  W >It was meant to be a rhetorical question; since Mozilla runs perfectly happily withoutoS >a Java Runtime installed, I don't believe there is any Java code in it at all. I'md >ready to be corrected though. >tD I can confirm that there is absolutely not a single line of Java in 
 Mozilla/CSWB.e   ------------------------------  # Date: Mon, 23 Jun 2003 17:12:24 GMTn' From: Colin Blake <colin@theblakes.com>a$ Subject: Re: Mozilla on old hardware, Message-ID: <3EF734F7.1040309@theblakes.com>  I There is a lot of JavaScript. The XPFE (cross-platform front end) is all h JS, XUL and CSS. h   ------------------------------  % Date: Mon, 23 Jun 2003 09:58:03 +0100i* From: "Richard Brodie" <R.Brodie@rl.ac.uk>$ Subject: Re: Mozilla on old hardware+ Message-ID: <bd6fes$g1e@newton.cc.rl.ac.uk>o  ` "John Reagan" <john.reagan@hp.com> wrote in message news:3XjIa.2965$sq2.1300@news.cpqcorp.net...  " > >>But mainly, Java is just slow. >i1 > > Maybe but what's that got to do with Mozilla?S > >n/ > Because Mozilla uses a lot of Java thats why.   V It was meant to be a rhetorical question; since Mozilla runs perfectly happily withoutR a Java Runtime installed, I don't believe there is any Java code in it at all. I'm ready to be corrected though.e   ------------------------------  # Date: Mon, 23 Jun 2003 14:26:08 GMTu& From: John Reagan <john.reagan@hp.com>$ Subject: Re: Mozilla on old hardware2 Message-ID: <46EJa.3135$X%6.2618@news.cpqcorp.net>   Colin Blake wrote:  F > I can confirm that there is absolutely not a single line of Java in  > Mozilla/CSWB.s >   I I stand corrected.  I thought there was lots of it.  Learn something new O
 every day!     -- t John Reagan ' Compaq Pascal/{A|I}MACRO Project Leader  Hewlett-Packard Companyi   ------------------------------    Date: 23 Jun 2003 07:27:35 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)O$ Subject: Re: Mozilla on old hardware3 Message-ID: <4DVYjN8seby3@eisner.encompasserve.org>   w In article <01KXB7CSGJHEAM67GX@sysdev.deutsche-boerse.com>, Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> writes:d > G > What about the speed of Mozilla on the 3000/600?  Fine, usable, or a o > pain?h  C    Somewhat slow.  This can range from a slight annoyance to nearly "    unseable depending on the page.  D    A bigger probem is that it soaks up memory in my VXT 2000+ beyondD    my InfoServer 1000's ability to page, and doesn't release it evenE    on exit.  This doesn't happend too often, but I have to reboot ther    VXT when it does.   ------------------------------  % Date: Mon, 23 Jun 2003 19:12:22 -0400 2 From: rdeininger@mindspring.com (Robert Deininger)% Subject: Re: New AlphaServer DS15 ???HL Message-ID: <rdeininger-2306031912220001@user-uinj0cp.dialup.mindspring.com>  P In article <8%HJa.78713$RM6.1090856@news.chello.at>, peter@langstoeger.at wrote:  E >In article <BgGJa.3164$0a7.1899@news.cpqcorp.net>, "Fred Kleinsorge"o) <my-last-name@stardotzko.dec.com> writes: J >>The DS15 and DS25 and ES45 are all EV68 based systems that use the TitanL >>core logic chipset.  Titan is the follow-on to the Tsunami core logic thatN >>was in the DS10/DS20/ES40 - and is the main reason why this isn't just a CPU
 >>upgrade. >>M >>So, faster CPU & faster IO.  The simple differences between the systems are D >>the number of CPUs (1, 2, 4).  Great follow-ons to the DS10/20/40. >cI >And why is then a delay of almost a year between the ES45 and the DS15 ?    About 2 years, actually.  E Because the DS15 team didn't start work until after the HP merger wast1 complete.  Compaq wasn't going to do this system.    ------------------------------  # Date: Mon, 23 Jun 2003 16:53:53 GMTr9 From: "Fred Kleinsorge" <my-last-name@stardotzko.dec.com>i% Subject: Re: New AlphaServer DS15 ???t2 Message-ID: <BgGJa.3164$0a7.1899@news.cpqcorp.net>  H The DS15 and DS25 and ES45 are all EV68 based systems that use the TitanJ core logic chipset.  Titan is the follow-on to the Tsunami core logic thatL was in the DS10/DS20/ES40 - and is the main reason why this isn't just a CPU upgrade.  K So, faster CPU & faster IO.  The simple differences between the systems arelB the number of CPUs (1, 2, 4).  Great follow-ons to the DS10/20/40.    1 "Michael Unger" <unger@decus.de> wrote in message / news:bd1s4u$nc754$1@ID-152801.news.dfncis.de... ( > On 21-Jun-2003 14:09, bob smith wrote: >tJ > > There is also a DS25 listed, above the DS15, with a date of June 2003.J > > are there any tidbits about this? Am I correct in assuming this is new > > cpu in a DS style package? >gJ > The DS25 itself is no _new_ system; the earliest QuickSpecs I have savedH > (PDF file) is dated 5-Aug-2002. In the QuickSpecs dated 24-Mar-2003 asH > well as in the earlier one the "Alpha 21264C 1000 MHz processor with 8, > MB L2 DDR cache" is explicitely mentioned. >d > >tL > > If someone tells me that the motherboard is the same and the cpu is just& > > a plug in, that will make my day!! >l	 > Michaelt >e > -- >lB > Please do *not* send "Security Patch Notifications" or "SecurityC > Updates"; this system isn't running a Micro$oft operating system.t? > And don't annoy me <mailto:postmaster@[127.0.0.1]> please ;-)g >    ------------------------------  # Date: Mon, 23 Jun 2003 18:51:48 GMT 6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)% Subject: Re: New AlphaServer DS15 ???l4 Message-ID: <8%HJa.78713$RM6.1090856@news.chello.at>  n In article <BgGJa.3164$0a7.1899@news.cpqcorp.net>, "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> writes:I >The DS15 and DS25 and ES45 are all EV68 based systems that use the Titan.K >core logic chipset.  Titan is the follow-on to the Tsunami core logic thatoM >was in the DS10/DS20/ES40 - and is the main reason why this isn't just a CPUr	 >upgrade.s >lL >So, faster CPU & faster IO.  The simple differences between the systems areC >the number of CPUs (1, 2, 4).  Great follow-ons to the DS10/20/40.e  H And why is then a delay of almost a year between the ES45 and the DS15 ?   -- p Peter "EPLAN" LANGSTOEGERa% Network and OpenVMS system specialists E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Mon, 23 Jun 2003 18:29:40 -0400o, From: "David Turner" <dbturner@islandco.com>% Subject: Re: New AlphaServer DS15 ??? / Message-ID: <vff0409tvb7d6b@news.supernews.com>n  K Because Compaq/HP had a DS10-Lake, and were trying to use up their reservest before realeasing the DS15   Make sense right ?   DT   ------------------------------  % Date: Mon, 23 Jun 2003 19:13:18 -0400M2 From: rdeininger@mindspring.com (Robert Deininger)% Subject: Re: New AlphaServer DS15 ???nL Message-ID: <rdeininger-2306031913180001@user-uinj0cp.dialup.mindspring.com>  > In article <vff0409tvb7d6b@news.supernews.com>, "David Turner" <dbturner@islandco.com> wrote:  L >Because Compaq/HP had a DS10-Lake, and were trying to use up their reserves >before realeasing the DS15- >- >Make sense right ?-  E It might "make sense", but it isn't true.  Too bad for the conspiracyj theorists...   ------------------------------  # Date: Tue, 24 Jun 2003 00:24:01 GMTe1 From: Michael Austin <maustin@firstdbasource.com>a% Subject: Re: New AlphaServer DS15 ???m2 Message-ID: <3EF79854.3960364F@firstdbasource.com>   Fred Kleinsorge wrote: > J > The DS15 and DS25 and ES45 are all EV68 based systems that use the TitanL > core logic chipset.  Titan is the follow-on to the Tsunami core logic thatN > was in the DS10/DS20/ES40 - and is the main reason why this isn't just a CPU
 > upgrade. > M > So, faster CPU & faster IO.  The simple differences between the systems areeD > the number of CPUs (1, 2, 4).  Great follow-ons to the DS10/20/40. > 3 > "Michael Unger" <unger@decus.de> wrote in messager1 > news:bd1s4u$nc754$1@ID-152801.news.dfncis.de...w* > > On 21-Jun-2003 14:09, bob smith wrote: > >eL > > > There is also a DS25 listed, above the DS15, with a date of June 2003.L > > > are there any tidbits about this? Am I correct in assuming this is new  > > > cpu in a DS style package? > >nL > > The DS25 itself is no _new_ system; the earliest QuickSpecs I have savedJ > > (PDF file) is dated 5-Aug-2002. In the QuickSpecs dated 24-Mar-2003 asJ > > well as in the earlier one the "Alpha 21264C 1000 MHz processor with 8. > > MB L2 DDR cache" is explicitely mentioned. > >a > > >tN > > > If someone tells me that the motherboard is the same and the cpu is just( > > > a plug in, that will make my day!! > >u > > Michaele > >r > > -- > >rD > > Please do *not* send "Security Patch Notifications" or "SecurityE > > Updates"; this system isn't running a Micro$oft operating system. A > > And don't annoy me <mailto:postmaster@[127.0.0.1]> please ;-)i > >a  H I think this should be DS15 (1 CPU and up to 4GB memory),DS25 (Max 2CPU B x 8GB) and ES47 4 x 16GB. We currently have a total of 20 of theseB systems on order  (8,8,4). plus an additional 16 or so refurbished DS20's.h  F BTW, working with SSSU (Storage Systems Scripting Utility) and the EVAG creating snapshots is really cool stuff.  Automating processes that cansG get and add WWID presentation to the EVA can be a real challenge unlessdG you have DECNET and/or RSH along with an advanced understanding of what $ you can do with the pipe command....  D OCT  (off current topic)  -- NOT TESTED, but should be very close to	 accurate.e  D $! get the wwid from a running system running 7.3 and above assuming systems have 2 HBAs)G $pipe rsh 'p1' /user='p2' /pass='p3' "show dev fg/full" | sear sys$pipe1 "port name"/out=sys$ouput | -o 		(	read sys$pipe rec1; -b 			read sys$pipe rec2; -# 			define/job/nolog $REC1 &$rec1; -0# 			define/job/nolog $REC2 &$rec2  -< 		 );o $write sys$output f$element(2,"M- ",f$edit(f$trnlnm("$REC1"),"TRIM,COMPRESS")))1 $write sys$output f$element(2,"c- ",f$edit(f$trnlnm("$REC2"),"TRIM,COMPRESS")))_ $deas/job $REC1  $deas/job $REC2   H Does any one have a good example of programmatically getting this out of SDA    $anal/system show dev fga show dev fgb  B Currently there is not a lexical function to retrieve this info...    	   -- e Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163i   ------------------------------    Date: 24 Jun 2003 15:35:56 -07001 From: keithparris_NOSPAM@yahoo.com (Keith Parris)y? Subject: New Business View Magazine for HP Enterprise Customers = Message-ID: <cf15391e.0306241435.68e51a8c@posting.google.com>o  C "This week HP launches Business View, a new, quarterly magazine andc$ Webzine for HP enterprise customers.  F Business View focuses on helping business and IT executives understand> how they can solve real business problems and benefit from new% opportunities in technology advances.o  F Business View will help to shape both the perception and understandingA of HP as a global enterprise company, demonstrating the return onr5 investment technology (RoIT) available to businesses.   ; While the HP-produced publication will showcase the talent,iC capabilities, innovations, and global reach of HP and HP's business @ partners, Business View isn't strictly about HP. It will feature@ stories about customers who face business challenges that can be$ answered with technology solutions."  " See http://www.hpbusinessview.com/  B And to save you the trouble, yes, OpenVMS is indeed mentioned. :-)   ------------------------------  # Date: Wed, 25 Jun 2003 00:44:33 GMTi4 From: brad@.gateway.2wire.net (Bradford J. Hamilton)C Subject: Re: New Business View Magazine for HP Enterprise Customers , Message-ID: <Rf6Ka.10073$R73.2481@sccrnsc04>  q In article <cf15391e.0306241435.68e51a8c@posting.google.com>, keithparris_NOSPAM@yahoo.com (Keith Parris) writes:  <snip># >See http://www.hpbusinessview.com/i > C >And to save you the trouble, yes, OpenVMS is indeed mentioned. :-)n  + Thanks, Keith - it's good to see a mention.i  A _________________________________________________________________ 0 Bradford J. Hamilton			"All opinions are my own"= bMradAhamiPltSon-at-atMtAbi.cPoSm	"Lose the MAPS, and replaces 					'-at-' with @"i   ------------------------------  % Date: Tue, 24 Jun 2003 20:58:53 -0500a1 From: "David J. Dachtera" <djesys.nospam@fsi.net> C Subject: Re: New Business View Magazine for HP Enterprise Customerst& Message-ID: <3EF901DD.BEAF306@fsi.net>   Keith Parris wrote:/ > [snip]D > And to save you the trouble, yes, OpenVMS is indeed mentioned. :-)  + ...but is it __A_D_V_E_R_T_I_S_E_D__ ???!!!h  H HP doesn't need to do a webzine to emphasise it's enterprise stuff, just1 "come out of the closet" and proudly advertise!!!   F Why is it that only morons like me with two-digit IQs understand that?   -- g David J. Dachteran dba DJE Systemsk http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/x   ------------------------------  # Date: Mon, 23 Jun 2003 13:38:05 GMTa3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond)2* Subject: Re: OpenVMS (TM) VAX Version X7G71 Message-ID: <1pDJa.3122$lR6.896@news.cpqcorp.net>5  w In article <bd3gu5$640$1@online.de>, helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) writes:eG >Subject says it all: what is this strange version number?  It appears 0 >when booting the CD.s >eJ >I borrowed a VAX CD to do a fresh install of VMS 7.3 on an old VAX in my # >hobbyist cluster.  It IS a 7.3 CD::  I "X7G7" is a version string that represents an internal OpenVMS baselevel.k  B If you boot either Standalone Backup (e.g. BOOT DKA400) or OpenVMSC (e.g. BOOT /R5=10000000 DKA400)  from the official V7.3 CD you willsD see this version string displayed.  (I just verified this.)  On the C installed (upgraded) system disk this is changed to display "V7.3".   . This is an anomaly that you can safely ignore.  E (X7G7 was the final baselevel from which OpenVMS VAX V7.3 was built.)    -- ]J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Mon, 23 Jun 2003 11:23:20 -0400g) From: "Neil Rieck" <n.rieck@sympatico.ca>t6 Subject: Ottawa OpenVMS Technical Seminar (June 25-26)9 Message-ID: <KXEJa.1299$Fe3.246800@news20.bellglobal.com>   I The Canadian Association of Compaq Users, in collaboration with HP Canadai andmL HP Corporation, will be hosting an OpenVMS 2-day technical seminar in Ottawa= on June 25 and 26, 2003. (on the University of Ottawa Campus)P  H All members, users and customers of HP technology from across Canada areF invited to attend this information-packed seminar for IT ProfessionalsK involved in HP technology platforms. Registration is now open. Please visit ; our seminar pages regularly to find out the latest updates.$   http://www.canacu.org/) http://www.communitywerke.com/seminar.htm.  
 Neil Rieck Kitchener/Waterloo/Cambridge,n Ontario, Canada.! http://www3.sympatico.ca/n.rieck/t   ------------------------------  # Date: Tue, 24 Jun 2003 15:27:31 GMTe2 From: "frank brown" <frank.brown@ci.seattle.wa.us># Subject: physical drive replacementr0 Message-ID: <D5_Ja.8$%E1.4993@news-west.eli.net>  I I'm getting UNCORRECTABLE ECC errors on DUA12 which is actually a pair oftI striped RZ29L (SCSI disks) shadowed with an identical 2-volume stripeset. G (We have 4 disks, configured as 2 shadowed stripesets.)  The drives areHG attached to an HSD10 DSSI-SCSI controller in a Storageworks BA350 tower-$ connected to a VMS 5.5-2 VAXcluster.  K My first problem is determining which of the 2 physical drives in stripeseteC DUA12 is throwing the errors.  ANALYZE/ERROR identifies the unit as L _RAID1$DUA12.  Since DUA12 is a stripset how can I tell which physical drive it is?  H Once I identify the drive with the errors, I'd like to replace it with aI spare.  Ideally I'd love to simply dismount the volume, pop the old drivegJ module out of the enclosure, replace the drive inside the module, slide itH back into the enclosure, initialize the drive, remount it as part of theL shadowset and have the system rebuild the volume.  However I realize this is improbable.a  I 1. The HSD10 manual says I can warm-swap the drive but I need to 'quiesceyE the SCSI bus'.  Is there a way to perform this operation from the VMS H command line or do I need to shutdown VMS to get to >>> console to enter HSD10 commands?tJ 2. Will I need to recreate the stripeset at the HSD10 or will the existingI stripeset definition work with the replacement drive (since it's the samet model in the same slot)?D 3. Any other thoughts or suggestions on dealing with this situation.   -Frank Brown Seattle Fire Dept. http://www.inwa.net/~frog/   ------------------------------  % Date: Tue, 24 Jun 2003 23:12:42 -0400   From: John Santos <JOHN@egh.com>' Subject: Re: physical drive replacement 5 Message-ID: <1030624224635.2835D-100000@Ives.egh.com>:  ' On Tue, 24 Jun 2003, frank brown wrote:!  K > I'm getting UNCORRECTABLE ECC errors on DUA12 which is actually a pair ofiK > striped RZ29L (SCSI disks) shadowed with an identical 2-volume stripeset.6I > (We have 4 disks, configured as 2 shadowed stripesets.)  The drives areeI > attached to an HSD10 DSSI-SCSI controller in a Storageworks BA350 towero& > connected to a VMS 5.5-2 VAXcluster. > M > My first problem is determining which of the 2 physical drives in stripesetiE > DUA12 is throwing the errors.  ANALYZE/ERROR identifies the unit as N > _RAID1$DUA12.  Since DUA12 is a stripset how can I tell which physical drive > it is? > J > Once I identify the drive with the errors, I'd like to replace it with aK > spare.  Ideally I'd love to simply dismount the volume, pop the old driveeL > module out of the enclosure, replace the drive inside the module, slide itJ > back into the enclosure, initialize the drive, remount it as part of theN > shadowset and have the system rebuild the volume.  However I realize this is
 > improbable.n > K > 1. The HSD10 manual says I can warm-swap the drive but I need to 'quiescehG > the SCSI bus'.  Is there a way to perform this operation from the VMSlJ > command line or do I need to shutdown VMS to get to >>> console to enter > HSD10 commands?u  E Frank - since no one else has responded yet (though I only see 36 newrE messages today, so my ISP might be having news server problems), I'lle throw in my 2 cents.  E I haven't used an HSD10, but on HSZ & HSJ controllers there are a setgD of buttons on the front, one for each bus.  You hold down the buttonD for a few seconds and it starts flashing.  This means the controllerF has noticed the button press and is stalling all I/O to that bus.  YouG then have about 30 seconds to make changes.  (I'm not sure, but I thinkeD you can press the button again to resume activity when you are down,B but the controller will time out after about 30 seconds and resumeE even if you do nothing.  While quiesced, the controllor will continuetC to accept I/O requests for drives on the bus, but won't do anythingmC about them.  To the hosts, it just looks like the disks have gottensC really slow, but nothing breaks.  You probably want to do this when-H the system is relatively idle, just to keep your users from complaining.  L > 2. Will I need to recreate the stripeset at the HSD10 or will the existingK > stripeset definition work with the replacement drive (since it's the samet > model in the same slot)?  1 Sorry, don't know how HSD's handle stripe sets...n  F > 3. Any other thoughts or suggestions on dealing with this situation.  ( You say "shadowed stripesets"... HBVS?    @ My guess, since half the blocks will need to be replaced is thatD the write-logging stuff in later VMS (probably not available in V5.5A anyway) wouldn't help in this and you'll have to do a full shadow B copy, but I think you should be able to dismount the bad stripesetA from the shadowset, if it hasn't already been kicked out (do thisoE before pulling the bad drive), replace the broken drive, reconstitute B the stripe set (Don't know how to do this), init the reconstituted> DUA12: with the same volume label as your original stripe set,D and mount it into the shadow set, which should trigger a shadow copyD (not Merge!) to it.  (Half the blocks, more or less, should still beA identical to the source, but the other half will be blank or tests> patterns or old data, so you definitely want it to do a copy.)@ In an hour or so, the copy should complete and you should be all set.  9 No down time for the application, provided you don't mind ? running without the shadow backup you normally have.  If reallyB< paranoid, you can shut down the application, backup the good= remaining stripeset (the good half of the shadow set), do the-C drive replace and shadow set rebuild, and then turn the application-D back on.  This method will result in considerable downtime, probablyG about an hour plus whatever time it takes to backup the good disks and  E to swap out the bad disk and rebuild the stripeset, but you will haven a good backup at all times.m  = Other people recently have discussed the benefits of doing ansB physical backup of a shadow set to a new disk that you want to add@ to the shadow set, in order to reduce the amount of copying thatA the shadow copy needs to do.  (I think shadow copying operates ine? a very cautious way.  Something like read from the source disk,hC read-check again to verify it reads okay, read from the destinationpD disk (looking for a potential bad block), read-check the destinationA disk, compare the source and destination, and if different, writeoD to the destination disk, and then writecheck the stuff just written.C If a bad block or check failure is found anywhere in all this, theno? the bad block replacement process is initiated.  I think if youaB backup/physical the source disk to the destination disk first, you@ save the copy from having to do the writes and write-checks, but@ of course the system still has to write all the data while doing> the backup, and has to read the data an extra time, so I don'tB see how this saves you much, especially if you /verify the backup.A A Google search should find the thread that discussed this, a fewi months ago.f     >  > -Frank Brown > Seattle Fire Dept. > http://www.inwa.net/~frog/  - Fire Dept?  Maybe you want to be paranoid :-)u   -- p John Santoso Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------    Date: 23 Jun 2003 15:54:29 -0500+ From: young_r@encompasserve.org (Rob Young) & Subject: Re: Portents of Itanium death3 Message-ID: <Jx3h0zb0Y3Jc@eisner.encompasserve.org>C  _ In article <iIOcnaz24LPXpmqjXTWJjQ@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:u >    >   Now obviouslyt@ >> tpmC tuning comes into play here.  But to take 2 configs that< >> are nearly a year apart, differing versions of the OS andB >> database and extrapolate frequency scaling from there is a poor >> comparison at best. > H > That's right:  it gives an unfair advantage to the later system, whichN > should be faster by virtue of whatever OS and database improvements occurredL > plus an additional year's tuning experience on that hardware.  And *still*L > the 17% faster clock rate (plus twice the memory) on the later system only! > boosted the TPC-C score by 11%.e > J > I guess you don't even pay much attention to your own observations, Rob. >    	Right, faster memory.   	The 44,942 versus the 38,386:   		9% faster cpun 		1/4 less memory' 		17% higher tpmCp  4 	Much contributed by faster memory as you point out.  > 	Coming full circle to the 2.0 GHz Xeon MP versus 2.8 GHz Xeon 	MP tpmC number...   	Slower PC1600 memory:  E http://www.tpc.org/results/individual_results/HP/ES_HP_DL580G2_4P.pdfd  
 	77,905 tpmC.w  @ 	So Bill, tell us about the frequency scaling there.  Shouldn't ? 	we expect at least a 20% increase in tpmC and quite a boost in ? 	price performance (no doubt able to trim quite a bit of memory B 	to achieve a high tpmC)?  If so, shouldn't we expect that 2.8 GHz6 	Xeon 4-way to shoot past a 4-way Opteron quite a bit?  F 	By the way, thanks for pointing out that discrepancy.  I hadn't until4 	now realized that the RackServer had PC2100 memory.  B 	Quite obvious the faster memory system contributes to higher tpmC5 	as Andrew points out latency is the important issue.P  A 	Thanks for helping, I think you've confirmed there are some nice-B 	2.8 GHz Xeon MP numbers waiting in the wings.  Interestingly, youA 	can order a 4-way from HP today with PC2100 and yet the Dell ands< 	HP 4-ways both have PC1600 memory (the PC2100 memory option  	was available in January 2003).  @ 	Now maybe I'm mixed up about this, I'm sure you will correct me& 	or more than likely just not respond.   				Robu   ------------------------------  % Date: Mon, 23 Jun 2003 13:44:14 +01008O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> & Subject: Re: Portents of Itanium death0 Message-ID: <bd6sn3$ghc$1@new-usenet.uk.sun.com>   Rob Young wrote:a > In article <4qacnQRQZ4qO4HKjXTWJiA@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:l > : >>"Rob Young" <young_r@encompasserve.org> wrote in message/ >>news:jomTkTPmcupf@eisner.encompasserve.org...  >>= >>>In article <bcno9701tlr@enews2.newsguy.com>, "Dann Corbit"e >> >><dcorbit@connx.com> writes:o >>< >>>>"Rob Young" <young_r@encompasserve.org> wrote in message >>> J >>>>Depends on what you are doing.  For database benchmarks and for things >>>m >>that >>6 >>>>need huge memory or disk, 32 bits craps out early. >>>> >>>d
 >>>Oh really?t >>M >>Actually, by definition in the case of 'things that need huge memory'.  AndrL >>that includes current TPC-C configurations (at least those using > 4 GB ofM >>physical memory):  IA32 *can* address it, but nowhere nearly as efficientlys0 >>(let alone easily) as a flat 64-bit model can. >  > A > 	But well enough to crank high tpmC in a 4-processor footprint.U >   ? The maximum memory you can address with a 32 bit OS in a singleh> flat model is 4GB, generally for OS reasons its slightly lower than this 3.75 GB.  = The AMD TPC-C system used 32 GB of RAM and it would have beent; more efficient from an OS and DBMS standpoint to be able tol0 address all of this rather than chunks of 3-4GB.  < The fastest Xeon MP results also use 32 GB of RAM which will- not be addressable in one flat address space.9  = Since both configurations would benefit from 64bit addressingu9 if they could use it your point about well enough for a 49+ CPU TPC-C result isn't very relevant is it.      > H >>There's no obvious reason that the architecture word size has anythingD >>whatsoever to do with heavy disk activity, though:  it's Opteron'sG >>significant bandwidth advantage that makes it superior to IA32 there.a >> >  > ? > 	There being where exactly?  In 4-processor CPU tpmC metrics?o6 > 	If so, how long does that lead last?  2 more weeks? >   > Most DBMS's spend between 50 and 75% of their time waiting forC main memory. Unless the 2.8 GHz Xeon MP also has much lower latency @ memory then you won't get the frequency speedup and thats before# you consider the other bottlenecks.t   Regardsd Andrew Harrisonn   > I >>>>Here is a favorable TPC report, even using a database compiled for 32  >>>u >>bits:  >>B >>>>http://www.tpc.org/results/FDR/TPCC/RackSaverQuatreX-64FDR.pdf >>> > >>>Good thing to bring up I suppose.  Bill has chided me, that? >>>is a 82000 tpmC result if memory serves.  You should be able-7 >>>to find its corresponding Xeon equivalent competitora/ >>>- a 4-CPU 2.0 GHz Xeon Dell - at 78000 tpmC.R >>>b? >>>When the Xeon goes from 2.0 GHz to 2.8 GHz, the Opteron wille >>>be trailing the Xeon. >>N >>That's probably true in the specific case of TPC-C (Opteron's lead over XeonL >>in other benchmarks probably won't be threatened, because other benchmarksH >>don't seem nearly as cache-size-sensitive), though it's not clear fromE >>analysis of the patterns of TPC-C scores just how much better TPC-CcH >>performance that 40% boost in clock rate will translate to (since evenJ >>though they'll supposedly be available within weeks no one seems to have@ >>posted TPC-C results for them yet:  do you still consider that >>irresponsible?)  >  > C > 	No.  Marketing.  Your dig is referring to my dig at AMD.  We seeh? > 	Opteron roll-out in April with tpmC 4-way 1.8 GHz , hardwareuB > 	availability in July 2003.  My dig is: if 2.0 GHz was coming in" > 	July, simply bench it in April. > @ > 	tpmC is no doubt part of a rollout strategy.  I suspect partsB > 	announced on June 30th and vendors rolling out tpmC in a day or > 	. > < >>- nor how long they'll retain whatever lead they manage to= >>eke out (since the 2 GHz Opterons are also imminently due, d >  > > > 	But as you would be quick to acknowledge, going from 1.8 toB > 	2.0 GHz certainly isn't going to do much regarding an increase.4 > 	It won't be close to the 2.8 GHz MP Xeon in tpmC. >  >  >>followed by 2.2a >  >  > 	In September, maybe., >  > * >>and 2.4 GHz versions before year's end). >  > L >>What *is* clear is that those Xeon MPs will still cost close to 5 times asL >>much as their Opteron counterparts, and that the systems they're availableL >>in will still cost close to twice as much (i.e., far costlier $/tpmC).  SoG >>the only reason you'd buy one would be if your performance needs weretI >>precisely in the (SWAG) 85K - 100K tpmC range:  lower, and Opteron willgN >>satisfy your needs far more cost-effectively; higher, and Xeon won't satisfy >>them at all. >> >  > B > 	You will buy a Xeon because that is what your tier 1 vendor you> > 	buy from today will be selling.  Sure, some will experiment# > 	with Opteron.  A small minority.A >  >  >>...s >>( >>  But my point was/is Xeon gets a huge >>7 >>>speedup in the 2 processor space - going to 3.06 GHz  >>H >>From its current 2.8 GHz:  do you consider a 9% speed increase 'huge'? >> >  > ? > 	No.  I guess I got thinking about Opteron in comparison.  Mya@ > 	mistake.  And 3.06 GHz is shipping today with 512KByte cache, > 	400 MHz FSB.e > E > http://www.tpc.org/tpcc/results/tpcc_result_detail.asp?id=103053003t > ; > 	Note too that was posted May 29th, hardware availabilityd > 	May 27th.   >  > 
 >> and the >>G >>>Xeon MP gets a big speed bump going from 2.0 to 2.8 GHz.  June 30th.H >>>y >>>sK >>>>Someday in the near future, due to active development by AMD, it may be  >>>. >>a  >> >>>>better than the Alpha chip.  >>>:E >>>Sure.  Athens - the next spin on Opteron - is "the great hope" now  >>K >>No:  current Opterons will be more than adequate as long as they meet thes@ >>dates for 2, 2.2, and 2.4 GHz versions that AMD has set forth. >> >  > > > 	No.  Because Opteron will be at 2.2 GHz in September.  Will< > 	it really be equivalent in performance to a 2.8 GHz Xeon? >  > = >>>as Xeon will be faster than Opteron for at least 6 months.7 >>% >>Not if AMD keeps to its schedule:    >  > ; > 	Sure.  Maybe not across all segments but in quite a few.t >  > ) >>2.2 GHz Opterons in 2 - 3 months shouldsN >>about equal the $3692 2.8 GHz Xeon MPs in TPC-C and 2.4 GHz Opterons in late9 >>fall should surpass them - still at far lower prices.  n >  > @ > 	But no tier 1 vendors.  That $3700 list price isn't the priceB > 	Dell pays for them.  What would Dell pay, $1600?   You see partC > 	of the problem is the Opteron is actually too cheap.  At $900 or > > 	so a pop, AMD may have a sweetheart deal of $600 to a largeC > 	OEM.  It is in Dell's and other's best interest to sell a highertC > 	priced part.  Especially if they are paing quite a bit less thand? > 	list price for the part.  It isn't a stretch to presume they C > 	are making a very nice chunk on each Xeon they sell.  They wouldeA > 	make a lot less on Opteron.  Smart business to sell/push thoset > 	4-way Xeons.a >  >  >>The late-fall 90 nmnF >>3.4 GHz Xeons with 1 MB of on-chip cache should be considerably moreN >>cost-effective for most uses than the Xeon MP dinosaurs and just about equalD >>the 2.4 GHz Opterons in both performance and price (though only inM >>applications where multi-processor bandwidth is not critical, since OpteronnN >>promises to lead there for the foreseeable future) - though since they'll beJ >>a full process generation ahead of Opteron that's hardly impressive, and; >>Opteron is scheduled to move to 90 nm before next summer.c >> >  > B > 	But that 3.4 GHz Xeon is for the 1-2 CPU space.  Prior to that,J > 	3.06 GHz with 1 MByte cache and 533 MHz bus which ought to bump tpmC.  J > 	That in the 3rd quarter and priced less than Opteron.  That link above: > E > http://www.tpc.org/tpcc/results/tpcc_result_detail.asp?id=103053003c > < > 	43000 tpmC with 2 processors and a decent $3.71 per tpmC.@ > 	Where is Opteron in that space?  That space gets more crowdedF > 	when Xeon goes to 3.4 GHz 1 MB before year-end and 533 MHz bus, 667E > 	MHz bus later (1).   That 3.4 GHz 1 MByte 533 MHz FSB part at $6903G > 	should be a shutout for Xeon in the 1-2 processor space.  After all,(A > 	a 2.0 GHz, 1 MByte , 400 MHz is doing 78000 tpmC.  I don't see-E > 	Opteron doing well at all in the 1-2 processor server space versusR? > 	Xeon in the next 6 months.  Opteron's niche gets real narrowa
 > 	real soon.- > D > 	But no matter, we will know more in a few weeks.  But I think theF > 	only metric Opteron will be leading at in 3 months will be $/metric > 	and SSL.a > 	 > 				Rob  > * > http://www.theinquirer.net/?article=9765   ------------------------------    Date: 23 Jun 2003 10:07:22 -0500+ From: young_r@encompasserve.org (Rob Young)o& Subject: Re: Portents of Itanium death3 Message-ID: <iSIBjuqet6fu@eisner.encompasserve.org>J  _ In article <cbSdnVukpIk2jG2jXTWJgQ@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:e   > $ >>  In 4-processor CPU tpmC metrics? > F > Er, no, Rob:  do you really think that TPC-C is bandwidth-intensive? >   6 	Yes.  Is it the most bandwidth intensive metric?  No.   > ...  > B >> >> When the Xeon goes from 2.0 GHz to 2.8 GHz, the Opteron will >> >> be trailing the Xeon.o >> >L >> > That's probably true in the specific case of TPC-C (Opteron's lead over > XeonD >> > in other benchmarks probably won't be threatened, because other > benchmarksK >> > don't seem nearly as cache-size-sensitive), though it's not clear from@H >> > analysis of the patterns of TPC-C scores just how much better TPC-CK >> > performance that 40% boost in clock rate will translate to (since eveneM >> > though they'll supposedly be available within weeks no one seems to havevC >> > posted TPC-C results for them yet:  do you still consider that  >> > irresponsible?) >>; >> No.  Marketing.  Your dig is referring to my dig at AMD.r > J > Indeed - specifically, at your suggestion that the lack of AMD benchmarkN > results for its faster Opterons said to be coming soon indicated that eitherM > they were vapor or that AMD was remiss in not having submitted them - whichaJ > seemed rather inconsistent with your other observation that AMD was justK > 'teasing' people by publishing 4-processor Opteron results in April for aiJ > system not available until July (not to mention the fact that Intel doesH > exactly the same things, depending upon which way the wind seems to be > blowing).a >   @ 	The benchmark in April for hardware available in October.  They> 	benchmark 1.8 GHz parts.  Since there is an allowable 6 month> 	lead time in tpc submissions, one would think they might have 	benched 2.0 GHz parts.a  < > But of course consistency has never been your strong suit.   	Sure.   >  > ...w > > >> >- nor how long they'll retain whatever lead they manage to? >> > eke out (since the 2 GHz Opterons are also imminently due,> >>> >> But as you would be quick to acknowledge, going from 1.8 toB >> 2.0 GHz certainly isn't going to do much regarding an increase.4 >> It won't be close to the 2.8 GHz MP Xeon in tpmC. > N > I think you might be wise not to count those particular tpmC chickens beforeI > they've hatched.  How much a 40% clock-rate boost will help the Xeon MPaM > scores depends more than anything else on the degree to which clock-rate is2' > the bottleneck in the current scores.i >    	Right.t  K > For example, take the two dual-processor Xeon ProLiant ML530G2T2P systems N > tested.  One was a 2.4 GHz (512 KB cache) system with 8 GB of RAM, the otherH > a 2.8 GHz (512 KB cache) system with 16 GB of RAM.  The 2.4 GHz systemK > scored 34473, the 2.8 GHz system scored 38386:  only 11% higher, a figuremI > which could be explained entirely by the increase in RAM based on otherOJ > evidence in the chart, which if true would mean that the 17% increase inK > clock speed contributed - nothing.  And while I think that interpretationdM > would be extreme, it remains clear that the 17% clock-rate increase yielded-N > *at best* an 11% better score even if you go to the other extreme and assumeM > that doubling the RAM contributed nothing; if the truth is somewhere in therM > middle, one might infer that each change contributed equally, in which casesL > extrapolating to the MP situation would imply that a 40% increase in clockK > rate would yield only about a 13% better TPC-C score if the RAM were heldr > constant at 32 GB. >    	Bad analysis , see below.  N > Not that I'd expect you to have bothered to examine such data before jumpingM > to the conclusion that your preferred platform will just naturally wipe out L > all competition:  that's always the attitude you had about Alpha, and thenM > more recently Itanic (at Alpha's expense), and now you seem to be extending M > it to all things Intel with the same uncritical (and uninformed) enthusiasm J > you've always shown.  If you didn't have such an obvious agenda I'd justJ > consider you something of a fool; as it is, I find you fairly obnoxious. >   C 	Well, that is sweet.  But back to your analysis.  You're comparingo@ 	the same hardware is a nice touch.  But they might as well haveA 	been running two different flavors of Linux running two separate @ 	versions of Oracle.  It would have been as relevent.  Why?  The> 	two you chose to compare are running two different OSes, two  	differing versions of SQL.  /  E 	Secondly, the 38,386 tpmC number was *obviously* pushed out the door F 	as an initial .NET number.  It was submitted January 8, 2003.  A much+ 	better number came along in 3 months time.o  C 	Take a 2 processor Xeon config, decrease its memory by 4 GBytes torD 	12 GBytes and Windows Enterprise running the same database version,D 	bump the clock from 2.8 GHz to 3.06 GHz and that 38,386 number goes? 	to 44,943.  So by decreasing the memory by 1/4, increasing thea> 	clock by 9%, the tpmC number increases by 17%.  Now obviously> 	tpmC tuning comes into play here.  But to take 2 configs that: 	are nearly a year apart, differing versions of the OS and@ 	database and extrapolate frequency scaling from there is a poor 	comparison at best.   > L > Since as noted above Opteron at 1.8 GHz should be at the very least a fullK > match for the current 3.06 GHz dual Xeon box, and while the Xeon clock isdJ > only going up 10% by the end of the year while Opteron's is scheduled toC > increase 33%, expecting anything like a 'shutout' for Xeon soundsm8 > suspiciously like more of your characteristic hot air. >    	We'll see.e   				Robe   ------------------------------  % Date: Mon, 23 Jun 2003 08:42:51 -0700n3 From: Robert Klute <robert_klute_removethis@hp.com>1& Subject: Re: Portents of Itanium death8 Message-ID: <ir7efvg2t7k8ip9ki7kbamksq1dqafcpqh@4ax.com>  E On Mon, 23 Jun 2003 13:44:14 +0100, Andrew Harrison SUNUK Consultancy . <Andrew_No.Harrison_No@nospamn.sun.com> wrote:    @ >The maximum memory you can address with a 32 bit OS in a single? >flat model is 4GB, generally for OS reasons its slightly lowerr >than this 3.75 GB.a  A This is usually hardware reasons, not OS regions.  The ability toeH utililize only 3.75GB of 4GB real memory is a pretty good indicator of a# system that uses memory-mapped I/O.e   ------------------------------  % Date: Mon, 23 Jun 2003 16:57:09 +0100fO From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>(& Subject: Re: Portents of Itanium death0 Message-ID: <bd780m$ke8$1@new-usenet.uk.sun.com>   Rob Young wrote:a > In article <cbSdnVukpIk2jG2jXTWJgQ@metrocast.net>, "Bill Todd" <billtodd@metrocast.net> writes:  >  > $ >>> In 4-processor CPU tpmC metrics? >>F >>Er, no, Rob:  do you really think that TPC-C is bandwidth-intensive? >> >  > 8 > 	Yes.  Is it the most bandwidth intensive metric?  No.  < TPC-C isn't bandwidth intensive its latency intensive. it is3 after all an OLTP benchmark not a DW/DSS benchmark.n   Regardsp Andrew Harrison2 >  >  >>...l >> >>A >>>>>When the Xeon goes from 2.0 GHz to 2.8 GHz, the Opteron willL >>>>>be trailing the Xeon. >>>>K >>>>That's probably true in the specific case of TPC-C (Opteron's lead overe >>>t >>Xeon >>C >>>>in other benchmarks probably won't be threatened, because otherr >>>  >>benchmarks >>J >>>>don't seem nearly as cache-size-sensitive), though it's not clear fromG >>>>analysis of the patterns of TPC-C scores just how much better TPC-CnJ >>>>performance that 40% boost in clock rate will translate to (since evenL >>>>though they'll supposedly be available within weeks no one seems to haveB >>>>posted TPC-C results for them yet:  do you still consider that >>>>irresponsible?)o >>>6; >>>No.  Marketing.  Your dig is referring to my dig at AMD.g >>J >>Indeed - specifically, at your suggestion that the lack of AMD benchmarkN >>results for its faster Opterons said to be coming soon indicated that eitherM >>they were vapor or that AMD was remiss in not having submitted them - which J >>seemed rather inconsistent with your other observation that AMD was justK >>'teasing' people by publishing 4-processor Opteron results in April for atJ >>system not available until July (not to mention the fact that Intel doesH >>exactly the same things, depending upon which way the wind seems to be >>blowing).r >> >  > B > 	The benchmark in April for hardware available in October.  They@ > 	benchmark 1.8 GHz parts.  Since there is an allowable 6 month@ > 	lead time in tpc submissions, one would think they might have > 	benched 2.0 GHz parts.S >  > < >>But of course consistency has never been your strong suit. >  >  > 	Sure. >  >  >>...C >> >>> >>>>- nor how long they'll retain whatever lead they manage to> >>>>eke out (since the 2 GHz Opterons are also imminently due, >>>v> >>>But as you would be quick to acknowledge, going from 1.8 toB >>>2.0 GHz certainly isn't going to do much regarding an increase.4 >>>It won't be close to the 2.8 GHz MP Xeon in tpmC. >>N >>I think you might be wise not to count those particular tpmC chickens beforeI >>they've hatched.  How much a 40% clock-rate boost will help the Xeon MP:M >>scores depends more than anything else on the degree to which clock-rate is.' >>the bottleneck in the current scores.  >> >  > 	 > 	Right.s >  > K >>For example, take the two dual-processor Xeon ProLiant ML530G2T2P systems N >>tested.  One was a 2.4 GHz (512 KB cache) system with 8 GB of RAM, the otherH >>a 2.8 GHz (512 KB cache) system with 16 GB of RAM.  The 2.4 GHz systemK >>scored 34473, the 2.8 GHz system scored 38386:  only 11% higher, a figurenI >>which could be explained entirely by the increase in RAM based on other J >>evidence in the chart, which if true would mean that the 17% increase inK >>clock speed contributed - nothing.  And while I think that interpretationiM >>would be extreme, it remains clear that the 17% clock-rate increase yieldedtN >>*at best* an 11% better score even if you go to the other extreme and assumeM >>that doubling the RAM contributed nothing; if the truth is somewhere in thesM >>middle, one might infer that each change contributed equally, in which case.L >>extrapolating to the MP situation would imply that a 40% increase in clockK >>rate would yield only about a 13% better TPC-C score if the RAM were heldn >>constant at 32 GB. >> >  >  > 	Bad analysis , see below. >  > N >>Not that I'd expect you to have bothered to examine such data before jumpingM >>to the conclusion that your preferred platform will just naturally wipe out/L >>all competition:  that's always the attitude you had about Alpha, and thenM >>more recently Itanic (at Alpha's expense), and now you seem to be extending_M >>it to all things Intel with the same uncritical (and uninformed) enthusiasmnJ >>you've always shown.  If you didn't have such an obvious agenda I'd justJ >>consider you something of a fool; as it is, I find you fairly obnoxious. >> >  > E > 	Well, that is sweet.  But back to your analysis.  You're comparingiB > 	the same hardware is a nice touch.  But they might as well haveC > 	been running two different flavors of Linux running two separateeB > 	versions of Oracle.  It would have been as relevent.  Why?  The@ > 	two you chose to compare are running two different OSes, two  > 	differing versions of SQL.  o > G > 	Secondly, the 38,386 tpmC number was *obviously* pushed out the dooraH > 	as an initial .NET number.  It was submitted January 8, 2003.  A much- > 	better number came along in 3 months time.J > E > 	Take a 2 processor Xeon config, decrease its memory by 4 GBytes togF > 	12 GBytes and Windows Enterprise running the same database version,F > 	bump the clock from 2.8 GHz to 3.06 GHz and that 38,386 number goesA > 	to 44,943.  So by decreasing the memory by 1/4, increasing thee@ > 	clock by 9%, the tpmC number increases by 17%.  Now obviously@ > 	tpmC tuning comes into play here.  But to take 2 configs that< > 	are nearly a year apart, differing versions of the OS andB > 	database and extrapolate frequency scaling from there is a poor > 	comparison at best. >  > L >>Since as noted above Opteron at 1.8 GHz should be at the very least a fullK >>match for the current 3.06 GHz dual Xeon box, and while the Xeon clock iseJ >>only going up 10% by the end of the year while Opteron's is scheduled toC >>increase 33%, expecting anything like a 'shutout' for Xeon soundsD8 >>suspiciously like more of your characteristic hot air. >> >  > 
 > 	We'll see.t > 	 > 				Rob) >    ------------------------------  % Date: Mon, 23 Jun 2003 12:09:00 -0400 * From: "Bill Todd" <billtodd@metrocast.net>& Subject: Re: Portents of Itanium death2 Message-ID: <3AmdnanEQtS8u2qjXTWJjw@metrocast.net>  @ "Robert Klute" <robert_klute_removethis@hp.com> wrote in message2 news:ir7efvg2t7k8ip9ki7kbamksq1dqafcpqh@4ax.com...G > On Mon, 23 Jun 2003 13:44:14 +0100, Andrew Harrison SUNUK Consultancy-0 > <Andrew_No.Harrison_No@nospamn.sun.com> wrote: >3 >:B > >The maximum memory you can address with a 32 bit OS in a singleA > >flat model is 4GB, generally for OS reasons its slightly lower9 > >than this 3.75 GB.l >bC > This is usually hardware reasons, not OS regions.  The ability tonJ > utililize only 3.75GB of 4GB real memory is a pretty good indicator of a% > system that uses memory-mapped I/O.d  L I think perhaps you should study this discussion a bit more carefully beforeJ presuming to comment upon it.  Or possibly consider taking an introductory course in computer science.o   - bill   ------------------------------  % Date: Mon, 23 Jun 2003 13:39:14 -0400i* From: "Bill Todd" <billtodd@metrocast.net>& Subject: Re: Portents of Itanium death2 Message-ID: <iIOcnaz24LPXpmqjXTWJjQ@metrocast.net>  8 "Rob Young" <young_r@encompasserve.org> wrote in message- news:iSIBjuqet6fu@eisner.encompasserve.org...-@ > In article <cbSdnVukpIk2jG2jXTWJgQ@metrocast.net>, "Bill Todd"  <billtodd@metrocast.net> writes: >e > >.& > >>  In 4-processor CPU tpmC metrics? > > H > > Er, no, Rob:  do you really think that TPC-C is bandwidth-intensive? > >C >r > Yes.   Figures - confused as ever.o   ...i  L > > seemed rather inconsistent with your other observation that AMD was justK > > 'teasing' people by publishing 4-processor Opteron results in April for  aHL > > system not available until July (not to mention the fact that Intel doesJ > > exactly the same things, depending upon which way the wind seems to be
 > > blowing).e > >  >1; > The benchmark in April for hardware available in October.a  H Just like the first 64-processor SuperDome TPC-C submission (and the one2 that just replaced it, though it was done in May).   ...I  E > > For example, take the two dual-processor Xeon ProLiant ML530G2T2P  systemseJ > > tested.  One was a 2.4 GHz (512 KB cache) system with 8 GB of RAM, the otheruJ > > a 2.8 GHz (512 KB cache) system with 16 GB of RAM.  The 2.4 GHz systemF > > scored 34473, the 2.8 GHz system scored 38386:  only 11% higher, a figureK > > which could be explained entirely by the increase in RAM based on otherbL > > evidence in the chart, which if true would mean that the 17% increase in> > > clock speed contributed - nothing.  And while I think that interpretationG > > would be extreme, it remains clear that the 17% clock-rate increase  yielded I > > *at best* an 11% better score even if you go to the other extreme and( assumeK > > that doubling the RAM contributed nothing; if the truth is somewhere ino thelJ > > middle, one might infer that each change contributed equally, in which caseH > > extrapolating to the MP situation would imply that a 40% increase in clock H > > rate would yield only about a 13% better TPC-C score if the RAM were held > > constant at 32 GB. > >D >0 > Bad analysis  % Coming from you, that's a real laugh.o   ...      You're comparing$ > the same hardware is a nice touch.  E Too bad you didn't do the same below in your attempt to spin away the ( apparently poor scaling with clock rate.     But they might as well havetB > been running two different flavors of Linux running two separateA > versions of Oracle.  It would have been as relevent.  Why?  The.> > two you chose to compare are running two different OSes, two > differing versions of SQL.  F Your usual shoddy research, Rob.  The SQL versions were the same, justE different patch levels.  The OSs were the same family, just different 	 releases.    >eF > Secondly, the 38,386 tpmC number was *obviously* pushed out the door > as an initial .NET number.  B While the erstwhile .NET system may offer slightly increased TPC-CI performance over its predecessor, it doesn't change tuning significantly.sL And HP has been doing TPC-C tests for long enough that a few months wouldn'tJ affect their ability to tune on hardware and software with which they were already well-acquainted.  +   It was submitted January 8, 2003.  A muchc, > better number came along in 3 months time.  J On different hardware, idiot.  While I didn't find any submission 3 monthsH after January 8th, aside from the fact that none were ML530G2T2P systemsK another defining characteristic of all the faster submissions was that theyiJ used PC2100 memory rather than PC1600.  And even then the only submissionsL that improved performance significantly were also running faster processors.   >rD > Take a 2 processor Xeon config, decrease its memory by 4 GBytes toE > 12 GBytes and Windows Enterprise running the same database version,nE > bump the clock from 2.8 GHz to 3.06 GHz and that 38,386 number goesc@ > to 44,943.  So by decreasing the memory by 1/4, increasing the0 > clock by 9%, the tpmC number increases by 17%.  J As I noted above, you neglected the small matter of 25% (or 33%, depending- on how you want to measure it) faster memory.l     Now obviously ? > tpmC tuning comes into play here.  But to take 2 configs thatn; > are nearly a year apart, differing versions of the OS andaA > database and extrapolate frequency scaling from there is a poord > comparison at best.m  F That's right:  it gives an unfair advantage to the later system, whichL should be faster by virtue of whatever OS and database improvements occurredJ plus an additional year's tuning experience on that hardware.  And *still*J the 17% faster clock rate (plus twice the memory) on the later system only boosted the TPC-C score by 11%.e  H I guess you don't even pay much attention to your own observations, Rob.   - bill   ------------------------------  % Date: Mon, 23 Jun 2003 13:01:44 -0700-* From: "Jack Peacock" <peacock@simconv.com>& Subject: Re: Portents of Itanium death2 Message-ID: <uy2dna_mcK11wWqjXTWcpA@mpowercom.net>  @ "Robert Klute" <robert_klute_removethis@hp.com> wrote in message2 news:ir7efvg2t7k8ip9ki7kbamksq1dqafcpqh@4ax.com...C > This is usually hardware reasons, not OS regions.  The ability tonJ > utililize only 3.75GB of 4GB real memory is a pretty good indicator of a% > system that uses memory-mapped I/O.y > D I'm a little rusty on Pentium hardware but the hole at the top is byH software design, not hardware mapping.  Memory space is one of the areasK where NT shows it's design influence from VAX VMS, in that the hi-order bitrH separates user from system space.  When Microsoft hit the 2GB user spaceK barrier they moved the system boundary up to 3GB user and 1GB system.  WhenwJ that wasn't enough Intel was forced to add some hardware (CPU and chipset)G to kludge in a sort of bank switch to move physical memory past the 4GBhI limit (up to 36 bit address, if memory serves).  Problem is very few appshH support the address extension, and it isn't all that programmer friendlyK (compared to a flat hi address space like the Alpha).  That was part of therJ motivation for AMD to extend the x86 CPU out to a full 64-bit flat address space.  J Interesting to note that while Intel claims no one needs the extra addressI space Intel had to find a way past the 4GB limit on x86 architecture too.fF Shame MS never expanded SQL Server on the Alpha to use the full 64-bitG memory addressing.  It will be interesting to see AMD benchmarks for K8 J database servers running 64GB (and what x86 Intel server will they compareG it to, since it's Intel policy not to compare AMD to anything Itanium).n    Jack Peacockd   ------------------------------  % Date: Mon, 23 Jun 2003 13:32:06 -0700)3 From: Robert Klute <robert_klute_removethis@hp.com>e& Subject: Re: Portents of Itanium death8 Message-ID: <lhoefvcvgrk6bknt2vv3h20msrcs6gonn6@4ax.com>  H On Mon, 23 Jun 2003 12:09:00 -0400, "Bill Todd" <billtodd@metrocast.net> wrote:   > A >"Robert Klute" <robert_klute_removethis@hp.com> wrote in messagep3 >news:ir7efvg2t7k8ip9ki7kbamksq1dqafcpqh@4ax.com...sH >> On Mon, 23 Jun 2003 13:44:14 +0100, Andrew Harrison SUNUK Consultancy1 >> <Andrew_No.Harrison_No@nospamn.sun.com> wrote:a >> >>C >> >The maximum memory you can address with a 32 bit OS in a single B >> >flat model is 4GB, generally for OS reasons its slightly lower >> >than this 3.75 GB. >>D >> This is usually hardware reasons, not OS regions.  The ability toK >> utililize only 3.75GB of 4GB real memory is a pretty good indicator of ab& >> system that uses memory-mapped I/O. >nM >I think perhaps you should study this discussion a bit more carefully before-K >presuming to comment upon it.  Or possibly consider taking an introductoryv >course in computer science.  E Forgive my ignorance, I may be little out of date.  Perhaps you couldrH make life easier for me by suggesting a text and which chapters I shouldE focus on.   I am afraid the last time I taught OS design I used Shaw.u   ------------------------------  % Date: Mon, 23 Jun 2003 14:16:22 -0700r3 From: Robert Klute <robert_klute_removethis@hp.com> & Subject: Re: Portents of Itanium death8 Message-ID: <v1qefvobic6knje1niligg3uot2luj6ptm@4ax.com>  H On Mon, 23 Jun 2003 13:01:44 -0700, "Jack Peacock" <peacock@simconv.com> wrote:  A >"Robert Klute" <robert_klute_removethis@hp.com> wrote in messagei3 >news:ir7efvg2t7k8ip9ki7kbamksq1dqafcpqh@4ax.com...dD >> This is usually hardware reasons, not OS regions.  The ability toK >> utililize only 3.75GB of 4GB real memory is a pretty good indicator of a & >> system that uses memory-mapped I/O. >>E >I'm a little rusty on Pentium hardware but the hole at the top is by8) >software design, not hardware mapping.     F There is a 'hole' from F0000H to FE000H.  It is space reserved for theA BIOS and, on really old PCs, ROM-BASIC. Some would say that is OS H related, I tend to view it as system related since unlike the PDP-8s and? 12s PCs don't have switches to toggle in the bootstrap routine.     & > ... Memory space is one of the areasL >where NT shows it's design influence from VAX VMS, in that the hi-order bitI >separates user from system space.  When Microsoft hit the 2GB user space L >barrier they moved the system boundary up to 3GB user and 1GB system.  WhenK >that wasn't enough Intel was forced to add some hardware (CPU and chipset)iH >to kludge in a sort of bank switch to move physical memory past the 4GBJ >limit (up to 36 bit address, if memory serves).  Problem is very few appsI >support the address extension, and it isn't all that programmer friendlyTL >(compared to a flat hi address space like the Alpha).  That was part of theK >motivation for AMD to extend the x86 CPU out to a full 64-bit flat addresso >space.y >t  A Yes, it is called PAE (Physical Address Extensions) and to use ithF Windows must be enabled to use it and the application must know how toC use it.  It is similar in concept to LIM-EMS (Lotus-Intel-Microsofti9 Expanded Memory Specification) used on 8088 and 8086 PCs.-    K >Interesting to note that while Intel claims no one needs the extra addressoJ >space Intel had to find a way past the 4GB limit on x86 architecture too.G >Shame MS never expanded SQL Server on the Alpha to use the full 64-bitoH >memory addressing.  It will be interesting to see AMD benchmarks for K8K >database servers running 64GB (and what x86 Intel server will they compareeH >it to, since it's Intel policy not to compare AMD to anything Itanium). >   Jack Peacock  A There will be comparisons available, whether it be SPEC or TPC orxG whatever, as vendors produce numbers to tout their hardware.  Are therem/ any numbers yet for AMD that are pure 64 bit?      ------------------------------  % Date: Tue, 24 Jun 2003 13:23:24 -0700o* From: "Jack Peacock" <peacock@simconv.com>& Subject: Re: Portents of Itanium death2 Message-ID: <wxadnR6H3p8JL2WjXTWcpw@mpowercom.net>  @ "Robert Klute" <robert_klute_removethis@hp.com> wrote in message2 news:v1qefvobic6knje1niligg3uot2luj6ptm@4ax.com...H > There is a 'hole' from F0000H to FE000H.  It is space reserved for theC > BIOS and, on really old PCs, ROM-BASIC. Some would say that is OStJ > related, I tend to view it as system related since unlike the PDP-8s andA > 12s PCs don't have switches to toggle in the bootstrap routine.e >nD That hole is purely a PC design convention and isn't part of the CPUK architecture.   I believe many x86 boards don't bother with it anymore.  IndE fact I have a mid-80's 80286 S-100 system that has real mode RAM from F 00000-EFFFF (960KB) and a 64KB ROM board above that (FF000-FFFFF), and" another 64KB from (100000-10FFFF). >  >  Are there/ > any numbers yet for AMD that are pure 64 bit?u >fJ I seem to remember some early tests on Linux with beta K8 CPUs, which wereK so-so.  Looks like we won't see Windows 64 tests until the end of the year.iL Which begs the same question, how will Intel avoid comparing K8s to ItaniumsJ when there is no Xeon that will run Windows 64?  I expect the Itanium willJ show better numbers, but when cost is factored in the spread won't look so bad.  I Reading the news today I see Apple has done some Bad Things to make their K new G5 benchmarks comparable to high end X86 CPUs.  I hope AMD doesn't give.4 in to the Dark Side and follow in Apple's footsteps.    Jack Peacocko    Jack Peacockr   ------------------------------  % Date: Tue, 24 Jun 2003 19:29:32 -0700e3 From: Robert Klute <robert_klute_removethis@hp.com>l& Subject: Re: Portents of Itanium death8 Message-ID: <ki1ifv8bshktdub46mjhfkinih6mdsum5b@4ax.com>  H On Tue, 24 Jun 2003 13:23:24 -0700, "Jack Peacock" <peacock@simconv.com> wrote:  A >"Robert Klute" <robert_klute_removethis@hp.com> wrote in messagev3 >news:v1qefvobic6knje1niligg3uot2luj6ptm@4ax.com...gI >> There is a 'hole' from F0000H to FE000H.  It is space reserved for theCD >> BIOS and, on really old PCs, ROM-BASIC. Some would say that is OSK >> related, I tend to view it as system related since unlike the PDP-8s andsB >> 12s PCs don't have switches to toggle in the bootstrap routine. >>E >That hole is purely a PC design convention and isn't part of the CPUeL >architecture.   I believe many x86 boards don't bother with it anymore.  InF >fact I have a mid-80's 80286 S-100 system that has real mode RAM fromG >00000-EFFFF (960KB) and a 64KB ROM board above that (FF000-FFFFF), anda# >another 64KB from (100000-10FFFF).   ? Yes, it definitely is a convention.  At the same time it is thehG 'standard' as set by IBM when they released the first IBM PC.  Graphics H Video memory started at A0000, C0000 was a common starting point for the	 disk biosr  H The 80286 had a 24-bit address bus, but you had to run in protected mode the get to the extended memory..   ------------------------------    Date: 23 Jun 2003 09:42:28 -07000 From: dave.baxter@bannerhealth.com (Dave Baxter)+ Subject: Re: Problem with SAN attached SDLTg= Message-ID: <a3c44af1.0306230842.74a62a30@posting.google.com>t  ` "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message news:<3EF3CF77.F7995228@fsi.net>... > Dave Baxter wrote: > > 	 > > Help!i > > G > >     I have a cluster of ES40's (OVMS 7.3-1, with 7.3-1 Backup V0100HE > > patch applied) connected to an HP SAN (Brocade switches, HSG80's,,G > > etc.) and I use an MSL5026 mini-library connected to the SAN SwitchhI > > via an MDR FC/SCSI bridge.     I am not sure what version the Library>1 > > is at, however the SDLT is at F/W Version 46.sD > >      The Library/Robot functions fine.    The VMS INIT and MOUNTE > > command appear to work fine, however when I execute even the mosteF > > basic backup command VMS tells me that there are problems with the$ > > label on the tape.   As follows; > > - > > MICKEY$SYSMGR>> robot load slot 3 drive 0r  > > MICKEY$SYSMGR>> show dev mga > > H > > Device                  Device           Error    Volume        Free
 > > Trans MntfI > > Name                   Status           Count     Label        Blocks 
 > > Count CntA2 > > $2$MGA2:      (MICKEY)  Online               1 > > & > > MICKEY$SYSMGR>> init $2$mga2: test& > > MICKEY$SYSMGR>> mount/for $2$mga2:8 > > %MOUNT-I-MOUNTED, TEST mounted on _$2$MGA2: (MICKEY)4 > > MICKEY$SYSMGR>> backup/log *.* $2$mga2:save1.bckD > > %BACKUP-W-MOUNTERR, volume 1 on _$2$MGA2 was not mounted because. > > its label does not match the one requested > > E > >       Has anyone seen a problem like this.     I have tried every E > > qualifier that I thought might be related without any significantt > > success. > H > Add /IGNORE=LABEL. It's expecting to find a volume labelled SAVE1, but > you have TEST MOUNTed.   Thank you David,F        It kind of surprised me that this worked since this was the theE first thing that I tried after the initial failure.     I realise nownE that although the initial failure had dismounted the drive, there waslF some internal issue that was persisting which made subsequent attemptsE fail.    Today I unloaded and reloaded the tape (using the robot) and E repeated the commands above, (adding the /IGNORE=LABEL qualifier) ande it worked fine. C        My remaining question is this, is it now necessary to alwaysrE specify /IGNORE=LABEL or append /LABEL=<label> for the backup to work C from disk to tape.    Or is this something that is required because   the Tape Drive is SAN attached??F        Although with older versions of VMS and on other systems I have@ tended to use the /IGNORE=(LABEL,INTERLOCK) as a kind of routineE default qualifier, it isn't clear from the BACKUP help that this is ai requiremente   Thanks Dave.   ------------------------------  # Date: Mon, 23 Jun 2003 12:00:35 GMTs+ From: Jeff Cameron <JCam90502@jcameron.com>o+ Subject: Re: Problem with SAN attached SDLTa2 Message-ID: <BB1C39F2.9DDC%JCam90502@jcameron.com>   On 6/20/03 2:33 PM, in article> a3c44af1.0306201333.72f24045@posting.google.com, "Dave Baxter"% <dave.baxter@bannerhealth.com> wrote:    > Help!  > C >   I have a cluster of ES40's (OVMS 7.3-1, with 7.3-1 Backup V0100eC > patch applied) connected to an HP SAN (Brocade switches, HSG80's,tE > etc.) and I use an MSL5026 mini-library connected to the SAN Switch>G > via an MDR FC/SCSI bridge.     I am not sure what version the Libraryo/ > is at, however the SDLT is at F/W Version 46.a@ >    The Library/Robot functions fine.    The VMS INIT and MOUNTC > command appear to work fine, however when I execute even the mostnD > basic backup command VMS tells me that there are problems with the" > label on the tape.   As follows; > + > MICKEY$SYSMGR>> robot load slot 3 drive 0f > MICKEY$SYSMGR>> show dev mga > F > Device                  Device           Error    Volume        Free > Trans MnttG > Name                   Status           Count     Label        Blocksi > Count Cntc0 > $2$MGA2:      (MICKEY)  Online               1 > $ > MICKEY$SYSMGR>> init $2$mga2: test$ > MICKEY$SYSMGR>> mount/for $2$mga2:6 > %MOUNT-I-MOUNTED, TEST mounted on _$2$MGA2: (MICKEY)2 > MICKEY$SYSMGR>> backup/log *.* $2$mga2:save1.bckB > %BACKUP-W-MOUNTERR, volume 1 on _$2$MGA2 was not mounted because, > its label does not match the one requested >  > A >     Has anyone seen a problem like this.     I have tried every C > qualifier that I thought might be related without any significantt
 > success. > B >    This was not a problem when the systems were first installed,F > with OVMS 7.3 and Tape firmware Vers 45.    The upgrade to OVMS onlyE > happened last week and the problem predated that (I was hoping thatrC > the VMS upgrade might resolve it.)     I have a suspicion that iteF > might have started a couple of months ago when the SDLT Firmware was3 > upgraded to Version 46, although I can't be sure.a > ( > Any help would be gratefully received. >  > Dave Baxter.  :-(5K  Using the /LABEL=TEST on the backup command will allow the backup to write L to the volume. Normally Backup would want to write to a volume with the same label as the saveset name.   ------------------------------  % Date: Mon, 23 Jun 2003 21:35:17 -0500n1 From: "David J. Dachtera" <djesys.nospam@fsi.net> + Subject: Re: Problem with SAN attached SDLT ' Message-ID: <3EF7B8E5.6DB894B9@fsi.net>t   Dave Baxter wrote: > b > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message news:<3EF3CF77.F7995228@fsi.net>... > > Dave Baxter wrote: > > >  > > > Help!i > > > [snip] > > > ( > > > MICKEY$SYSMGR>> init $2$mga2: test( > > > MICKEY$SYSMGR>> mount/for $2$mga2:: > > > %MOUNT-I-MOUNTED, TEST mounted on _$2$MGA2: (MICKEY)6 > > > MICKEY$SYSMGR>> backup/log *.* $2$mga2:save1.bckF > > > %BACKUP-W-MOUNTERR, volume 1 on _$2$MGA2 was not mounted because0 > > > its label does not match the one requested > > >tG > > >       Has anyone seen a problem like this.     I have tried every G > > > qualifier that I thought might be related without any significant  > > > success. > >>J > > Add /IGNORE=LABEL. It's expecting to find a volume labelled SAVE1, but > > you have TEST MOUNTed. >  > Thank you David,H >        It kind of surprised me that this worked since this was the theG > first thing that I tried after the initial failure.     I realise nowtG > that although the initial failure had dismounted the drive, there was H > some internal issue that was persisting which made subsequent attemptsG > fail.    Today I unloaded and reloaded the tape (using the robot) andlG > repeated the commands above, (adding the /IGNORE=LABEL qualifier) ande > it worked fine. E >        My remaining question is this, is it now necessary to alwaysuG > specify /IGNORE=LABEL or append /LABEL=<label> for the backup to work E > from disk to tape.    Or is this something that is required becauses" > the Tape Drive is SAN attached??  F This has been documented behavior since ages. Check the doc.'s (System Management Utilities manual?).  H >        Although with older versions of VMS and on other systems I haveB > tended to use the /IGNORE=(LABEL,INTERLOCK) as a kind of routineG > default qualifier, it isn't clear from the BACKUP help that this is aB
 > requirement   H That does seem to be a deficiency in the on-line HELP, AFAICT, yes. I'veG just bumped my head on it often enough to have learned the lesson well.P   -- n David J. Dachteran dba DJE Systemss http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/o   ------------------------------  # Date: Tue, 24 Jun 2003 00:36:30 GMTh1 From: Michael Austin <maustin@firstdbasource.com>i+ Subject: Re: Problem with SAN attached SDLTo2 Message-ID: <3EF79B45.AA91BD2B@firstdbasource.com>   Dave Baxter wrote: > b > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message news:<3EF3CF77.F7995228@fsi.net>... > > Dave Baxter wrote: > > >h > > > Help!t > > >cI > > >     I have a cluster of ES40's (OVMS 7.3-1, with 7.3-1 Backup V0100 G > > > patch applied) connected to an HP SAN (Brocade switches, HSG80's, I > > > etc.) and I use an MSL5026 mini-library connected to the SAN Switch K > > > via an MDR FC/SCSI bridge.     I am not sure what version the Library 3 > > > is at, however the SDLT is at F/W Version 46.cF > > >      The Library/Robot functions fine.    The VMS INIT and MOUNTG > > > command appear to work fine, however when I execute even the mosthH > > > basic backup command VMS tells me that there are problems with the& > > > label on the tape.   As follows; > > > / > > > MICKEY$SYSMGR>> robot load slot 3 drive 0t" > > > MICKEY$SYSMGR>> show dev mga > > >dJ > > > Device                  Device           Error    Volume        Free > > > Trans MntaK > > > Name                   Status           Count     Label        Blocks  > > > Count Cntj4 > > > $2$MGA2:      (MICKEY)  Online               1 > > > ( > > > MICKEY$SYSMGR>> init $2$mga2: test( > > > MICKEY$SYSMGR>> mount/for $2$mga2:: > > > %MOUNT-I-MOUNTED, TEST mounted on _$2$MGA2: (MICKEY)6 > > > MICKEY$SYSMGR>> backup/log *.* $2$mga2:save1.bckF > > > %BACKUP-W-MOUNTERR, volume 1 on _$2$MGA2 was not mounted because0 > > > its label does not match the one requested > > > G > > >       Has anyone seen a problem like this.     I have tried every G > > > qualifier that I thought might be related without any significantr > > > success. > >mJ > > Add /IGNORE=LABEL. It's expecting to find a volume labelled SAVE1, but > > you have TEST MOUNTed. >  > Thank you David,H >        It kind of surprised me that this worked since this was the theG > first thing that I tried after the initial failure.     I realise now G > that although the initial failure had dismounted the drive, there wastH > some internal issue that was persisting which made subsequent attemptsG > fail.    Today I unloaded and reloaded the tape (using the robot) andsG > repeated the commands above, (adding the /IGNORE=LABEL qualifier) ando > it worked fine.nE >        My remaining question is this, is it now necessary to alwaysaG > specify /IGNORE=LABEL or append /LABEL=<label> for the backup to work/E > from disk to tape.    Or is this something that is required becausew" > the Tape Drive is SAN attached??H >        Although with older versions of VMS and on other systems I haveB > tended to use the /IGNORE=(LABEL,INTERLOCK) as a kind of routineG > default qualifier, it isn't clear from the BACKUP help that this is ae
 > requirement  >  > Thanks Dave.   You might try:   $mylabel = "label"0 $BACK/INIT/LABEL='mylabel'/ignore=interlock *.*  $2$MGA2:'mylabel'.bck/savesett     --   Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163y7 Sr. Consultant            http://www.firstdbasource.coms   ------------------------------    Date: 23 Jun 2003 09:58:11 -0700# From: fernando.vallarino@oca.com.uyK% Subject: Re: Problems with SYS$GETRMI = Message-ID: <5da5a1ff.0306230858.4d95993f@posting.google.com>k  Q Jonas Lindholm <jlhm@usa.net> wrote in message news:<3EF4CEFC.9030102@usa.net>...o > Hi,  > 1 > None of the specific CPU modes return any data.rJ > Instead you need to use the RMI$_MODES item to return an array of data, 3 > I believe it is 8 or 9 items for each active CPU.h > & > fernando.vallarino@oca.com.uy wrote: > G > >>None of these RMI$_ codes are valid when used with SYS$GETRMI. See:7 > >>P > >>  <http://h71000.www7.hp.com/doc/731FINAL/4527/4527pro_063.html#index_x_616> > >> > >>for valid RMI_ codes.- > >> > >> > > N > > That's right, but the RMI$_LEF returns the exact number of process in that
 > > state.R > > What can I do to obtain the cpu usage with the codes defined in that document? > > Which of them? > >    Good, I like that.4 I'm trying this, but I'm not understand some things.E I'm obtaining 10 values in a server with 1 cpu, 18 values in a serverC with- 2 cpus and 26 values in a server with 3 cpus.hD In all cases the first value agrees with the number of cpus (1, 2 or 3).  The cpu modes I know are: 
 -Interrupt -Synchronization -Kernel 
 -Executive -Supervisor  -Usere -Compatibility -Idle F So for eight values by cpu I understand 9, 17 or 25 for 1, 2 or 3 cpus2 because the first value is for the number of cpus.= Soo, i still have one value which I don't were it comes from.h  In which order are they comming?
 Thanks Jonas.(   ------------------------------   Date: 23 Jun 2003 23:19 CDTr' From: carl@gerg.tamu.edu (Carl Perkins)r% Subject: Re: Problems with SYS$GETRMIm- Message-ID: <23JUN200323194297@gerg.tamu.edu>o  ' Jonas Lindholm <jlhm@usa.net> writes...oC }All values are in cpu tickets but I always get the same value for b* }interrupts as for idle, I don't know why.  D I don't know why they'd be identical, but the CPU tick counters have@ always included the null time in the interrupt time counter. You8 normally subtract it out to get the real interrupt time.   And no, I don't know why.m   --- Carl   ------------------------------  # Date: Tue, 24 Jun 2003 03:25:48 GMTb# From: Jonas Lindholm <jlhm@usa.net>t% Subject: Re: Problems with SYS$GETRMIt& Message-ID: <3EF7C4E1.3010002@usa.net>   Hi!e  ( Tye structure looks something like this:   #pragma nomember_alignment   struct modes { 	int			cpucnt;	 	struct {  		char		cpuid; 		int unsigned	interrupts; 		int unsigned	mpsync; 		int unsigned	kernela 		int unsigned	executive;  		int unsigned	super;n 		int unsigned	user;5 		int unsigned	comat;  <== this is always 0 on alpha.t 		int unsigned	idle; 	} cpus[32]; }b    B All values are in cpu tickets but I always get the same value for ) interrupts as for idle, I don't know why.e    $ fernando.vallarino@oca.com.uy wrote:  S > Jonas Lindholm <jlhm@usa.net> wrote in message news:<3EF4CEFC.9030102@usa.net>...o >  >>Hi,a >>1 >>None of the specific CPU modes return any data. J >>Instead you need to use the RMI$_MODES item to return an array of data, 3 >>I believe it is 8 or 9 items for each active CPU.i >>& >>fernando.vallarino@oca.com.uy wrote: >> >>G >>>>None of these RMI$_ codes are valid when used with SYS$GETRMI. See:m >>>>O >>>> <http://h71000.www7.hp.com/doc/731FINAL/4527/4527pro_063.html#index_x_616>> >>>> >>>>for valid RMI_ codes.c >>>> >>>> >>>>M >>>That's right, but the RMI$_LEF returns the exact number of process in thati	 >>>state. Q >>>What can I do to obtain the cpu usage with the codes defined in that document?e >>>Which of them?e >>>  >>>u >  > Good, I like that.6 > I'm trying this, but I'm not understand some things.G > I'm obtaining 10 values in a server with 1 cpu, 18 values in a server( > with/ > 2 cpus and 26 values in a server with 3 cpus. F > In all cases the first value agrees with the number of cpus (1, 2 or > 3).3 > The cpu modes I know are:8 > -Interrupt > -Synchronization	 > -Kernell > -Executive
 > -Supervisore > -User  > -Compatibility > -IdlevH > So for eight values by cpu I understand 9, 17 or 25 for 1, 2 or 3 cpus4 > because the first value is for the number of cpus.? > Soo, i still have one value which I don't were it comes from.h" > In which order are they comming? > Thanks Jonas.% >    ------------------------------    Date: 24 Jun 2003 10:20:03 -0700# From: fernando.vallarino@oca.com.uyh% Subject: Re: Problems with SYS$GETRMI = Message-ID: <5da5a1ff.0306240920.29037c02@posting.google.com>c  $ Jonas and Carl, thank you very much. All works fine now.4   ------------------------------    Date: 23 Jun 2003 20:54:24 -07000 From: Jason.Armistead@otis.com (Jason Armistead)R Subject: Secondhand MicroVAX / AlphaStation systems/parts resellers in Australia ?= Message-ID: <48ac369b.0306231954.7a9136d8@posting.google.com>s   Hi  D Can anyone out there in comp.os.vms land recommend a good secondhandB MicroVAX / AlphaStation reseller in Australia, and preferrably New
 South Wales ?n  A I'm looking for a MicroVAX 3100/76 or similar, or an AlphaStationl
 circa 266 MHzw  F Or, I'm after someone with a working AlphaStation 500/266 power supplyB module (DEC P/N 30-45491-01 Rev B02) which was a MagneTek MaverickC 3617-32-100 unit.  After a few mysterious power-offs overnight, the F AlphaStation now is completely dead with no apparent signs of damage. @ The P/S doesn't fire up and its fan stays dead still.  I've seenF plenty of bagging of the Italian MagneTek units as used by IBM / Apple@ on other newsgroups and suspect that mine has gone the same way.   Cheers   Jason Armisteado   ------------------------------  + Date: Tue, 24 Jun 2003 22:57:08 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)7 Subject: SET HOST/LAT and cut and paste with DECwindows $ Message-ID: <bdal04$vjk$2@online.de>  B Why can I only cut and paste two lines into a DECterm which has a I process which logged in via SET HOST/LAT?  If I log in via TELNET, I can e cut and paste much more.   ------------------------------  % Date: Mon, 23 Jun 2003 09:54:50 -040091 From: Bob Supnik <bsupnik@nauticusnet.nospam.com>  Subject: SIMH V3.0-0 releasedn8 Message-ID: <le1efvk7n9hvgjros2f7n0vanq956t26ef@4ax.com>  6 SIMH V3.0 was released yesterday on the SIMH web site.   http://simh.trailing-edge.como  D Despite the major version number change, the visible differences areD few. There have been substantial changes to internal data structuresA and definitions.  Also, there have been bug fixes in almost everyh= simulated system.  Please see the ReadMe for further details.   > In addition to sources, the web site has a zip file of Windows? executables; the Windows executables for VAX and PDP-11 are now  compiled with Ethernet support.    /Bob Supnika   ------------------------------  % Date: Mon, 23 Jun 2003 16:06:08 +0200-$ From: Michael Unger <unger@decus.de>! Subject: Re: SIMH V3.0-0 releasedm5 Message-ID: <bd7234$q2bdn$1@ID-152801.news.dfncis.de>r  ' On 23-Jun-2003 15:54, Bob Supnik wrote:n  8 > SIMH V3.0 was released yesterday on the SIMH web site. >  > http://simh.trailing-edge.come >  > [...]g > @ > In addition to sources, the web site has a zip file of WindowsA > executables; the Windows executables for VAX and PDP-11 are nowo! > compiled with Ethernet support.   F That's great! (I'm not able to compile and link myself because I don't, have any compiler available on that system.)   Michael9   -- a  @ Please do *not* send "Security Patch Notifications" or "SecurityA Updates"; this system isn't running a Micro$oft operating system.i= And don't annoy me <mailto:postmaster@[127.0.0.1]> please ;-)d   ------------------------------  # Date: Mon, 23 Jun 2003 23:39:06 GMTe0 From: "Noone Special" <kckckckckc2003@yahoo.com>! Subject: Re: SIMH V3.0-0 releasedoF Message-ID: <3d1a8b97042248d4fdc56f171350ff8f@unlimited.ultrafeed.com>  L When I try to run pdp11.exe or vax.exe I get a message indicating that it isI looking for a dll. I have already installed winpcap. What is this dll andP where can I get it?.   Thanks    > "Bob Supnik" <bsupnik@nauticusnet.nospam.com> wrote in message2 news:le1efvk7n9hvgjros2f7n0vanq956t26ef@4ax.com...8 > SIMH V3.0 was released yesterday on the SIMH web site. >e > http://simh.trailing-edge.com  > F > Despite the major version number change, the visible differences areF > few. There have been substantial changes to internal data structuresC > and definitions.  Also, there have been bug fixes in almost everyr? > simulated system.  Please see the ReadMe for further details.l >i@ > In addition to sources, the web site has a zip file of WindowsA > executables; the Windows executables for VAX and PDP-11 are nowx! > compiled with Ethernet support.o >h
 > /Bob Supnikd   ------------------------------  # Date: Tue, 24 Jun 2003 02:32:14 GMTw0 From: "Noone Special" <kckckckckc2003@yahoo.com>! Subject: Re: SIMH V3.0-0 releasedsF Message-ID: <2db378d77188e13369a18313bef88aa1@unlimited.ultrafeed.com>  K Ok, it is looking for the packet.dll, but I have that dll installed... What 
 is happening?e   Thanks  ; "Noone Special" <kckckckckc2003@yahoo.com> wrote in messagee@ news:3d1a8b97042248d4fdc56f171350ff8f@unlimited.ultrafeed.com...K > When I try to run pdp11.exe or vax.exe I get a message indicating that itg isK > looking for a dll. I have already installed winpcap. What is this dll andd > where can I get it?e >  > Thanks >i >i@ > "Bob Supnik" <bsupnik@nauticusnet.nospam.com> wrote in message4 > news:le1efvk7n9hvgjros2f7n0vanq956t26ef@4ax.com...: > > SIMH V3.0 was released yesterday on the SIMH web site. > >e! > > http://simh.trailing-edge.com  > >%H > > Despite the major version number change, the visible differences areH > > few. There have been substantial changes to internal data structuresE > > and definitions.  Also, there have been bug fixes in almost everyiA > > simulated system.  Please see the ReadMe for further details.4 > >cB > > In addition to sources, the web site has a zip file of WindowsC > > executables; the Windows executables for VAX and PDP-11 are now # > > compiled with Ethernet support.  > >l > > /Bob Supnik  >o >e   ------------------------------    Date: 23 Jun 2003 21:01:11 -07000 From: Jason.Armistead@otis.com (Jason Armistead)! Subject: Re: SIMH V3.0-0 releasedo= Message-ID: <48ac369b.0306232001.2fc2e851@posting.google.com>d  a Michael Unger <unger@decus.de> wrote in message news:<bd7234$q2bdn$1@ID-152801.news.dfncis.de>....) > On 23-Jun-2003 15:54, Bob Supnik wrote:e > : > > SIMH V3.0 was released yesterday on the SIMH web site. > > ! > > http://simh.trailing-edge.comQ > > 	 > > [...]h > > B > > In addition to sources, the web site has a zip file of WindowsC > > executables; the Windows executables for VAX and PDP-11 are now*# > > compiled with Ethernet support.a > H > That's great! (I'm not able to compile and link myself because I don't. > have any compiler available on that system.)   Bobe  E The FAQ needs to be updated, specifically question 2.2 about Ethernet D support on Windows requiring you to recompile from sources (which is not true for V3.0 SIMH)i  D PS: Congratulations on a very wonderful piece of software.  It was a9 real buzz for me to see OpenVMS 6.2 VAX fire up on my PC.s   Cheers   Jasone   ------------------------------    Date: 23 Jun 2003 21:49:30 -07000 From: Jason.Armistead@otis.com (Jason Armistead)! Subject: Re: SIMH V3.0-0 releasedg= Message-ID: <48ac369b.0306232049.5a4feb55@posting.google.com>s  ~ "Noone Special" <kckckckckc2003@yahoo.com> wrote in message news:<2db378d77188e13369a18313bef88aa1@unlimited.ultrafeed.com>...M > Ok, it is looking for the packet.dll, but I have that dll installed... Whatr > is happening?   1 Not QUITE, it's looking for a DLL called "packet"a  D For some reason the binaries of VAX.EXE (and probably PDP11.EXE) are6 looking for the "packet" file WITHOUT the .DLL suffix.  F If you dump the VAX.EXE with your favourite HEX dump program, and then= search for ".dll" you'll see a reference to WSOCK32.DLL, thenoB WPCAP.DLL, but in between these is one for just plain old "packet" (minus the .dll suffix).  D Go back and look at an old build of SIMH and you'll see it has these but "packet.dll"  D The easiest work-around is to copy PACKET.DLL from C:\WINDOWS\SYSTEM> or C:\WINNT\SYSTEM into your SIMH directory and rename it from PACKET.DLL to PACKET  D That'll get you going, though we really need a proper build put ontoE the SIMH web site.  I suspect someone goofed up when specifying the Cu) compiler dependence on the packet driver.l   Cheers  E Jason (who just discovered and fixed the same problem not more than as few minutes ago).      > Thanks > = > "Noone Special" <kckckckckc2003@yahoo.com> wrote in messageoB > news:3d1a8b97042248d4fdc56f171350ff8f@unlimited.ultrafeed.com...M > > When I try to run pdp11.exe or vax.exe I get a message indicating that itP >  is M > > looking for a dll. I have already installed winpcap. What is this dll and  > > where can I get it?v > >@
 > > Thanks > >m > >eB > > "Bob Supnik" <bsupnik@nauticusnet.nospam.com> wrote in message6 > > news:le1efvk7n9hvgjros2f7n0vanq956t26ef@4ax.com...< > > > SIMH V3.0 was released yesterday on the SIMH web site. > > > # > > > http://simh.trailing-edge.comd > > > J > > > Despite the major version number change, the visible differences areJ > > > few. There have been substantial changes to internal data structuresG > > > and definitions.  Also, there have been bug fixes in almost every C > > > simulated system.  Please see the ReadMe for further details.a > > >rD > > > In addition to sources, the web site has a zip file of WindowsE > > > executables; the Windows executables for VAX and PDP-11 are nowh% > > > compiled with Ethernet support.t > > >o > > > /Bob Supnikp > >h > >h   ------------------------------    Date: 24 Jun 2003 09:40:45 +0200C From: vaxinf@chclu.chemie.uni-konstanz.de (Eberhard Heuser-Hofmann)e! Subject: Re: SIMH V3.0-0 releaseda- Message-ID: <3ef8007d$1@news.uni-konstanz.de>   C In article <le1efvk7n9hvgjros2f7n0vanq956t26ef@4ax.com>, Bob Supnikt( <bsupnik@nauticusnet.nospam.com> writes:8 |>SIMH V3.0 was released yesterday on the SIMH web site. |> |>http://simh.trailing-edge.comB |>F |>Despite the major version number change, the visible differences areF |>few. There have been substantial changes to internal data structuresC |>and definitions.  Also, there have been bug fixes in almost everya? |>simulated system.  Please see the ReadMe for further details.i |>@ |>In addition to sources, the web site has a zip file of WindowsA |>executables; the Windows executables for VAX and PDP-11 are now)! |>compiled with Ethernet support.r |>
 |>/Bob Supnik  |>  & Did someone  test VAX-AXP.EXE on V7.3?   Here's my result:n  
 sim> boot cpun     KA655-B V5.3, VMB 2.7n Performing normal system tests. @ 40..39..38..37..36..35..34..33..32..31..30..29..28..27..26..25..@ 24..23..22..21..20..19..18..17..16..15..14..13..12..11..10..09.. 08..1 Exception in interrupt or exception, PC: 2004FF574. %SYSTEM-W-UNWIND, unwind currently in progress   eberhard   ------------------------------    Date: 23 Jun 2003 13:56:24 -05002 From: newton_l@encompasserve.org (Lawrence Newton) Subject: Slow FTP Gets.y3 Message-ID: <OHyo0zoCrMCa@eisner.encompasserve.org>5  L I have just configured a DS10 with VMS 7.3-1 with patches, but no patch for M TCPIP.  I am using the DE500 built in DE500 ethernet ports. Every thing seems>K to work except I can do an ftp get from other systems and the transfer runshH at about a megabyte/second. If I do an ftp get from the DS10 to the sameM systems, it runs at less than 10 kilobytes/second. I did a short ftp put from - a PC and it appeared to run at a fast speed. t  N I probably need to run some more tests, but has anyone had a similar problem??   Lawrence   ------------------------------  # Date: Mon, 23 Jun 2003 19:06:31 GMTu6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER) Subject: Re: Slow FTP Gets.o4 Message-ID: <XcIJa.78811$RM6.1090856@news.chello.at>  h In article <OHyo0zoCrMCa@eisner.encompasserve.org>, newton_l@encompasserve.org (Lawrence Newton) writes:M >I have just configured a DS10 with VMS 7.3-1 with patches, but no patch for uN >TCPIP.  I am using the DE500 built in DE500 ethernet ports. Every thing seemsL >to work except I can do an ftp get from other systems and the transfer runsI >at about a megabyte/second. If I do an ftp get from the DS10 to the sametN >systems, it runs at less than 10 kilobytes/second. I did a short ftp put from. >a PC and it appeared to run at a fast speed.   9 1.) Install ECOs. They fix problems. And TCPIP has a lot.rG 2.) Not digging too much into, I almost bet that this a autonegotiationa4 problem of the ethernet hardware (DE500, Switchport)   --   Peter "EPLAN" LANGSTOEGERa% Network and OpenVMS system specialists E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------    Date: 23 Jun 2003 18:15:23 -0700# From: dooleys@snowy.net.au (dooley)  Subject: Re: Slow FTP Gets.o< Message-ID: <1ca82fc6.0306231715.ca078ed@posting.google.com>  m newton_l@encompasserve.org (Lawrence Newton) wrote in message news:<OHyo0zoCrMCa@eisner.encompasserve.org>...cN > I have just configured a DS10 with VMS 7.3-1 with patches, but no patch for O > TCPIP.  I am using the DE500 built in DE500 ethernet ports. Every thing seemssM > to work except I can do an ftp get from other systems and the transfer runspJ > at about a megabyte/second. If I do an ftp get from the DS10 to the sameO > systems, it runs at less than 10 kilobytes/second. I did a short ftp put fromy/ > a PC and it appeared to run at a fast speed. x > P > I probably need to run some more tests, but has anyone had a similar problem?? > 
 > Lawrence( I did have a similar problem a while ago& though it was not a vms/tcpip problem,3 it was actually due to misconfiguration of a switchr- To track it down I used ttcp.exe to get some e stats between various nodese4 (you can also get ttcpw for testing from/to your pc) Phil   ------------------------------  # Date: Tue, 24 Jun 2003 02:04:58 GMTl; From: "John Gemignani, Jr." <jon-nope@thiswontworkossc.net>A Subject: Re: Slow FTP Gets.s= Message-ID: <elOJa.15951$Jw6.6779435@news1.news.adelphia.net>p  ? "Lawrence Newton" <newton_l@encompasserve.org> wrote in messageg- news:OHyo0zoCrMCa@eisner.encompasserve.org...oI > I have just configured a DS10 with VMS 7.3-1 with patches, but no patcho fornI > TCPIP.  I am using the DE500 built in DE500 ethernet ports. Every thinge seems4H > to work except I can do an ftp get from other systems and the transfer runsJ > at about a megabyte/second. If I do an ftp get from the DS10 to the sameJ > systems, it runs at less than 10 kilobytes/second. I did a short ftp put from. > a PC and it appeared to run at a fast speed. > F > I probably need to run some more tests, but has anyone had a similar	 problem??  >F
 > Lawrence  I Check your default route.  When I converted my network I forgot to changedJ the default route and while incoming data was coming in off a T1, the ACKs were going over a 56K modem.   John [Red faced]   ------------------------------  # Date: Tue, 24 Jun 2003 00:42:08 GMTo1 From: Michael Austin <maustin@firstdbasource.com>  Subject: Re: Slow FTP Gets.'2 Message-ID: <3EF79C97.935199CF@firstdbasource.com>   Lawrence Newton wrote: > M > I have just configured a DS10 with VMS 7.3-1 with patches, but no patch for0O > TCPIP.  I am using the DE500 built in DE500 ethernet ports. Every thing seems M > to work except I can do an ftp get from other systems and the transfer runsLJ > at about a megabyte/second. If I do an ftp get from the DS10 to the sameO > systems, it runs at less than 10 kilobytes/second. I did a short ftp put froml. > a PC and it appeared to run at a fast speed. > P > I probably need to run some more tests, but has anyone had a similar problem?? > 
 > Lawrence   at the console P00>>  set ew*mode FASTFDe  or is that an EI device?A  9 and if your switch supports it, force it 100 Full Duplex.k   -- r Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163    ------------------------------  % Date: Mon, 23 Jun 2003 11:53:50 -05003+ From: "Irving F. Snurd" <dacadc0@lycos.com>k Subject: SSH2 ON VMS?l( Message-ID: <3EF7309E.3060601@lycos.com>  F We're using SSH2 on our Linux servers, and we run Putty on our PCs to  connect to them, but VMSG doesn't fit this pic, as it only supports SSH1.  Is there a way to get _ SSH2 on VMS?    8 Is SSH2 coming?  Do I already have it and don't know it?  #         Thanx, Bill Manwaring, IUCFo   ------------------------------  % Date: Mon, 23 Jun 2003 16:26:53 -0400o% From: "AwkEmacs" <awkemacs@yahoo.com>i Subject: Re: SSH2 ON VMS? 6 Message-ID: <pan.2003.06.23.20.26.53.337558@yahoo.com>  : On Mon, 23 Jun 2003 11:53:50 -0500, Irving F. Snurd wrote:   > H > We're using SSH2 on our Linux servers, and we run Putty on our PCs to  > connect to them, but VMSI > doesn't fit this pic, as it only supports SSH1.  Is there a way to get   > SSH2 on VMS?   > : > Is SSH2 coming?  Do I already have it and don't know it? > % >         Thanx, Bill Manwaring, IUCFo  9 Process Software sells a SSH package that supports SSH2. o   Awkn   ------------------------------   Date: 23 JUN 2003 19:15:02 GMT4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher) Subject: Re: SSH2 ON VMS?c6 Message-ID: <23JUN03.19150271@thuria.waisman.wisc.edu>  C In a previous article, "Irving F. Snurd" <dacadc0@lycos.com> wrote:o -> eH ->We're using SSH2 on our Linux servers, and we run Putty on our PCs to  ->connect to them, but VMSI ->doesn't fit this pic, as it only supports SSH1.  Is there a way to get S ->SSH2 on VMS?    - See:	http://h71000.www7.hp.com/new/index.htmlt  6 This is the second release of the early adopter's kit.  F Tip: if you're running the first EAK release (from March or so) you'll need to do:   "  $ @sys$manager:tcpip$ssh_shutdown!  $ @sys$manager:tcpip$ssh_startupH  H After installing the June 20 release in order to install the new images.   --G -- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madison 6 --               karcher.nomorespxm@waisman.wisc.edu     ------------------------------    Date: 23 Jun 2003 14:31:00 -0700( From: bob@instantwhip.com (Bob Ceculski) Subject: Re: SSH2 ON VMS?o= Message-ID: <d7791aa1.0306231331.34daac99@posting.google.com>e  [ "Irving F. Snurd" <dacadc0@lycos.com> wrote in message news:<3EF7309E.3060601@lycos.com>....H > We're using SSH2 on our Linux servers, and we run Putty on our PCs to  > connect to them, but VMSI > doesn't fit this pic, as it only supports SSH1.  Is there a way to get   > SSH2 on VMS?    4 Multinet and TCPware have had it for a while now ...   ------------------------------  % Date: Mon, 23 Jun 2003 17:38:13 -0400 . From: Mike Bartman <omni@foolie.omniphile.com> Subject: Re: SSH2 ON VMS?S8 Message-ID: <bksefvsl857ad1lmvc7n8e184lg3rrsf5i@4ax.com>  C On Mon, 23 Jun 2003 16:26:53 -0400, "AwkEmacs" <awkemacs@yahoo.com>v wrote:; >On Mon, 23 Jun 2003 11:53:50 -0500, Irving F. Snurd wrote:b  ; >> Is SSH2 coming?  Do I already have it and don't know it?  >>  & >>         Thanx, Bill Manwaring, IUCF >I: >Process Software sells a SSH package that supports SSH2.   B Correct.  It works with MultiNet, TCPware...and TCPIP Services For OpenVMS (UCX :^).  See:   % http://www.process.com/tcpip/ssh.htmlS   for more info.  F Disclaimer:  I work for Process Software, but not on SSH specifically.   -- Mike Bartmanh@ ----------------------------------------------------------------=   To reply via e-mail, remove the 'foolie.' from the address. %   I'm getting sick of all the SPAM...i@ ----------------------------------------------------------------   ------------------------------  # Date: Tue, 24 Jun 2003 00:39:42 GMT(1 From: Michael Austin <maustin@firstdbasource.com>  Subject: Re: SSH2 ON VMS? 2 Message-ID: <3EF79C06.3914B119@firstdbasource.com>   "Irving F. Snurd" wrote: > G > We're using SSH2 on our Linux servers, and we run Putty on our PCs tom > connect to them, but VMSH > doesn't fit this pic, as it only supports SSH1.  Is there a way to get > SSH2 on VMS? > : > Is SSH2 coming?  Do I already have it and don't know it? > % >         Thanx, Bill Manwaring, IUCFi  E IMHO  TERATERM with SSH applied works much better than Putty.   quicksE google for teraterm and ssh should find what you are looking for.. if0H you are looking for it....  I am not sure if Reflections10 has it yet or not... -- t Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #2611630   ------------------------------  % Date: Mon, 23 Jun 2003 19:36:25 +0200  From: Dirk Munk <munk@home.nl> Subject: Re: SSH2 ON VMS?12 Message-ID: <bd7e72$uj6$1@news2.tilbu1.nb.home.nl>  M There is a SSH2 Early Adpter Kit for TCP/IP V5.3 available now. (EAK is less i then Beta !)Q However the full functional SSH2 kit will be part of TCP/IP V5.4, that will be a hO part of OpenVMS 7.3-2, and that will be available later this year (if all goes b well).N The OpenVMS SSH will be 'real' SSH, not OpenSSH. The difference is in the way  the keys are administrated etc.i   Irving F. Snurd wrote: > H > We're using SSH2 on our Linux servers, and we run Putty on our PCs to  > connect to them, but VMSI > doesn't fit this pic, as it only supports SSH1.  Is there a way to get   > SSH2 on VMS?  : > Is SSH2 coming?  Do I already have it and don't know it? > $ >        Thanx, Bill Manwaring, IUCF >    ------------------------------    Date: 24 Jun 2003 07:23:23 -0500+ From: young_r@encompasserve.org (Rob Young)e Subject: Re: SSH2 ON VMS?/3 Message-ID: <+gGDaWMKGKL$@eisner.encompasserve.org>$  f In article <3EF79C06.3914B119@firstdbasource.com>, Michael Austin <maustin@firstdbasource.com> writes: > "Irving F. Snurd" wrote: >> tH >> We're using SSH2 on our Linux servers, and we run Putty on our PCs to >> connect to them, but VMS I >> doesn't fit this pic, as it only supports SSH1.  Is there a way to getf >> SSH2 on VMS?. >>  ; >> Is SSH2 coming?  Do I already have it and don't know it?l >>  & >>         Thanx, Bill Manwaring, IUCF > G > IMHO  TERATERM with SSH applied works much better than Putty.   quick G > google for teraterm and ssh should find what you are looking for.. if J > you are looking for it....  I am not sure if Reflections10 has it yet or > not...   	Yes, Reflections 10 does:  : http://www.wrq.com/products/reflection/win/ruov/specs.html  / Security (with Reflection Security Components1)e   SSL/TLS Telnet encryption K Integrated SSL 3.0 and TLS 1.0 Telnet encryption through Microsoft securityn APIs2i Reflection OpenSSH Client 0 Connections to ssh-1 and ssh-2 protocol servers 8 Port forwarding for TCP/IP ports through secure channel / Password, user key and Kerberos authentication  < Key generation and conversion (RSA-1, RSA, and DSA) utility % scp and sftp file transfer utilities  / Standard SSH configurations for easy deploymentf   			Rob   ------------------------------  % Date: Mon, 23 Jun 2003 12:46:03 +0100 O From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>e$ Subject: Re: Sun to be the next DEC!0 Message-ID: <bd6pa0$fan$1@new-usenet.uk.sun.com>   Heinz Getzler wrote:o > bob@instantwhip.com (Bob Ceculski) wrote in message news:<d7791aa1.0306160915.37f9047b@posting.google.com>...b > 5 >>gee Andrew, I hope this means we willn't be hearing 5 >>from you anymore :) ... there is still time to join ! >>the OpenVMS Itanium sales team!f >>M >>http://techupdate.zdnet.com/techupdate/stories/main/0,14179,2914032,00.html  >  > F > In many ways Sun's problems are similar to HP's. Sun is primarilly aH > hardware  company. In the past few years Sun has made some attempts to@ > become more of a software company with it's Sun One products &< > Java.... Sun's big problem is that it is still paranoid ofF > Linux............IBM on the otherhand is not afraid of Linux and hasG > embraced it more than any other company. If Sun were smart they woulda  > push Java to the Linux market.   Humm we have done.  ; The majority of the Linux systems in financial services areE8 running Java, we provided the source to allow that to be	 possible.s  8 And hidden in the recent announcements about Sun selling6 Red Hat ES/AS/WS was a reciprical deal with Red Hat to redistribute Java with RH.  6 As for IBM embracing Linux, that seems in the light of6 the SCO suit to be a very mixed blessing for the Linux
 community.  8 RH and all the other Linux distro companies could end up< going through all the source with a fine tooth comb removing9 offending sections (roughly what IBM mad the x86 ROM BIOSO' clone vendors do in the early PC days).f  : This would not be a good thing for OpenSource and sadly it9 isn't as if IBMs contributions to the OpenSouce communityi; ammount to much anyway, a so so Journaling Filesystem whichn8 there are alternatives for and some kernel work that may3 have to be ripped out at grat cost, nothing else of  note springs to mind.p  5 Sun on the otherhand hasn't bothered wasting money on-3 foolishness like Linux on a mainframe or Power, why06 decommoditise something that relies almost exclusively- on commoditity hardware as its sales benefit.d  / Instead we have donated things that are usefullc$ Grid, Java, OpenOffice, Mozilla etc.  H > Another big problems is that a most of their solutions are designed toH > support the Solaris platform only. N-1 is Sun's data center management? > solution that is designed to support Sun hardware only. IBM'suG > On-demand stategy is a solution designed to support multiple OS's andH > hardware.  >   6 Not entirely accurate N1 is multi-platform while IBM's8 offering is mostly a thin technology wrapper for a thick services offering.   Regards  Andrew Harrisonr   ------------------------------    Date: 24 Jun 2003 12:59:27 -0700' From: timasmith@hotmail.com (Tim Smith)  Subject: system disk on ODS-5w< Message-ID: <a7234bb1.0306241159.3968ae2@posting.google.com>  Q I hear this is not recommended - what is the current status/issues for vms 7.3-1?;  < How big a deal is it to convert a system disk back to ODS-2?   thanks   Tim    ------------------------------  % Date: Tue, 24 Jun 2003 21:28:31 +0100 * From: "John Travell" <john@travell.uk.net>! Subject: Re: system disk on ODS-5 5 Message-ID: <bdac9i$r1ogl$1@ID-120847.news.dfncis.de>   4 "Tim Smith" <timasmith@hotmail.com> wrote in message6 news:a7234bb1.0306241159.3968ae2@posting.google.com...L > I hear this is not recommended - what is the current status/issues for vms 7.3-1? >E> > How big a deal is it to convert a system disk back to ODS-2? >B $$$ Backup 'disk 'tape $$$ init/structure=2 $$$ backup/noinit 'tape 'diske   -- John Travell  VMS crashdump expertise for hire john@travell.uk.netr http://www.travell.uk.net/       --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.489 / Virus Database: 288 - Release Date: 10/06/2003,   ------------------------------    Date: 24 Jun 2003 15:32:06 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) ! Subject: Re: system disk on ODS-5e3 Message-ID: <GCJCVxSvbg$X@eisner.encompasserve.org>   f In article <a7234bb1.0306241159.3968ae2@posting.google.com>, timasmith@hotmail.com (Tim Smith) writes:S > I hear this is not recommended - what is the current status/issues for vms 7.3-1?e > > > How big a deal is it to convert a system disk back to ODS-2?      Easy, but time consuming:  B    1)  Restore the last image backup from when the system disk was       ODS-2.C    2)  Reapply all the changes you made since then except the ODS-5        conversion..  H    In short, there is no reverse-conversion, you gotta back up and start     over from where you left off.  F    While my OS is too out of date to support ODS-5 system disks, I runE    all my user disks ODS-5 and have never encountered a problem others7    than scripts that assumed unzip did .->_ conversion.m   ------------------------------  # Date: Tue, 24 Jun 2003 20:58:54 GMTN3 From: hammond@not@peek.ssr.hp.com (Charlie Hammond)l! Subject: Re: system disk on ODS-5d0 Message-ID: <iY2Ka.3287$gT.199@news.cpqcorp.net>  = In article <a7234bb1.0306241159.3968ae2@posting.google.com>, 6) timasmith@hotmail.com (Tim Smith) writes:r  8 >I hear [that ODS-5 system disks are] not recommended - 1 >what is the current status/issues for vms 7.3-1?o  : ODS-5 system disks are supported for OpenVMS Alpha V7.3-1.  B Some utilities and Layered Products (HP and third-party) may have  limitations related to ODS-5.  e  H The POLYCENTER Software Installation (PCSI) utility can tolerate ODS-5, N but it does not like to encounter file or directory names that exceed ODS-2.  K Neither in the kit nor on the installation target.  (These PCSI constraints = may be relaxed or eliminated in a future version of OpenVMS.)   > My personal recommendation is that if you do not want or need @ the additional features provided by ODS-5, you should use ODS-2.  = >How big a deal is it to convert a system disk back to ODS-2?   . Not trivial.  Requires a backup and restore.  ; May be issues with ODS-5 filenames that are reset to ODS-2 o compliant names.  < You should read the related documentation before you decide.8 The V7.3-1 Upgrade and Installation manual points to the documenation you should review.n  ? One supposes that at some future time, ODS-2 will be as rare as ? ODS-1 is today.  There is no reason to avoid it if you need it.f8 If you do not need ODS-5, ODS-2 continues to serve well.   --  J       Charlie Hammond -- Hewlett-Packard Company -- Ft Lauderdale  FL  USAF           (hammond@not@peek.ssr.hp.com -- remove "@not" when replying)J       All opinions expressed are my own and not necessarily my employer's.   ------------------------------  % Date: Tue, 24 Jun 2003 22:55:16 +0200r From: Dirk Munk <munk@home.nl>! Subject: Re: system disk on ODS-5-2 Message-ID: <bdae8h$cb0$1@news2.tilbu1.nb.home.nl>   Tim Smith wrote:S > I hear this is not recommended - what is the current status/issues for vms 7.3-1?3 > > > How big a deal is it to convert a system disk back to ODS-2? >  > thanks >  > Timt    N We have converted all (system)disks to ODS-5, no real problems. Sometimes you M run into some minor problems when a utility doesn't like lowercase filenames l% etc.. PCSI is one of those utilities.r  Q Some more recent applications (java etc.) may require a ODS-5 disk, they can not n be used with ODS-2.y  D  From VMS 7.3-1 upwards you can transform your system disk to ODS-5.   ------------------------------  % Date: Mon, 23 Jun 2003 16:36:51 -0500m+ From: "Irving F. Snurd" <dacadc0@lycos.com>0% Subject: Re: TCPIP problem with 7.3-1o( Message-ID: <3EF772F3.5010809@lycos.com>  D We're running the 5.3 stack, and our systems guy did the "mandatory H TCPIP update" as part of the upgrade to VMS 7.3-1.  We're running on an H alpha DS10. We aren't sure what the stack was under 7.3.  Our alpha can @ talk to PLCs and other machines.  Our systems guy broke down theB connect packets and they look the same as the packets coming from @ machines that succeed in talking to the motor controller.  He's E scratching his head; I was hoping some resourceful puzzle-solver sys iG admin out there would say, "Oh, yeah, we've seen that; just diddle the o @#@ parameter...">           Bill Manwaring, IUCF   Thomas H. Pauli wrote:   > Hi,d > E > I think it's not VMS you have problems with, but the TCP/IP stack.  H > Which version do you run now ($ TCPIP SHOW VERSION) and which did you J > run before ? Perhaps you've got a TCP/IP upgrade wioth your VMS upgrade! >. > Thomas > 6 > On Mon, 23 Jun 2003 11:46:31 -0500, Irving F. Snurd  > <dacadc0@lycos.com> wrote: >>F >> We upgraded from 7.3 to 7.3-1 and now cannot talk to a TCPIP motor F >> controller (Parker  CompuMotor 6k).  We cannot telnet or "connect" F >> (socket code), whereas we could do both with 7.3.  We tried Linux, 4 >> and vms 7.2, and they're both able to talk to it. >>G >> The motor controller is sort  of stupid; it does not respond to ARP nI >> broadcasts, so we bind the MAC address to the IP number in vms before sG >> we connect, but the source and destination MAC addresses are in the aG >> connect packet.  The motor controller just does not want to respond e
 >> to 7.3. >>H >> Has anyone else had tcpip trouble?  I guess we'll have to go back to  >> 7.3.i >>  >> regards, Bill Manwaring, IUCF >> >> >u >r >n   ------------------------------  % Date: Mon, 23 Jun 2003 11:46:31 -0500e+ From: "Irving F. Snurd" <dacadc0@lycos.com>c! Subject: TCPIP problem with 7.3-1n( Message-ID: <3EF72EE7.2030307@lycos.com>  C We upgraded from 7.3 to 7.3-1 and now cannot talk to a TCPIP motor :C controller (Parker  CompuMotor 6k).  We cannot telnet or "connect"  G (socket code), whereas we could do both with 7.3.  We tried Linux, and e- vms 7.2, and they're both able to talk to it.   D The motor controller is sort  of stupid; it does not respond to ARP I broadcasts, so we bind the MAC address to the IP number in vms before we cI connect, but the source and destination MAC addresses are in the connect ?C packet.  The motor controller just does not want to respond to 7.3.i  I Has anyone else had tcpip trouble?  I guess we'll have to go back to 7.3.r  %         regards, Bill Manwaring, IUCF,   ------------------------------  % Date: Mon, 23 Jun 2003 20:49:45 +0200o. From: "Thomas H. Pauli" <thomaspauli@arcor.de>% Subject: Re: TCPIP problem with 7.3-1A, Message-ID: <oprq8g87aploujp4@news.arcor.de>   Hi,4  I I think it's not VMS you have problems with, but the TCP/IP stack. Which wK version do you run now ($ TCPIP SHOW VERSION) and which did you run before i= ? Perhaps you've got a TCP/IP upgrade wioth your VMS upgrade!t   Thomas  H On Mon, 23 Jun 2003 11:46:31 -0500, Irving F. Snurd <dacadc0@lycos.com>  wrote:  E > We upgraded from 7.3 to 7.3-1 and now cannot talk to a TCPIP motor  E > controller (Parker  CompuMotor 6k).  We cannot telnet or "connect"  I > (socket code), whereas we could do both with 7.3.  We tried Linux, and E/ > vms 7.2, and they're both able to talk to it.) >uF > The motor controller is sort  of stupid; it does not respond to ARP K > broadcasts, so we bind the MAC address to the IP number in vms before we  K > connect, but the source and destination MAC addresses are in the connect EE > packet.  The motor controller just does not want to respond to 7.3.  >aK > Has anyone else had tcpip trouble?  I guess we'll have to go back to 7.3.o >m > regards, Bill Manwaring, IUCFa >n >e       -- l Thomas H. Paulif Hammersteinstr. 19 14199 Berlin Germanye   ------------------------------  # Date: Mon, 23 Jun 2003 19:00:12 GMT 6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)% Subject: Re: TCPIP problem with 7.3-1b4 Message-ID: <07IJa.78771$RM6.1090856@news.chello.at>  V In article <3EF72EE7.2030307@lycos.com>, "Irving F. Snurd" <dacadc0@lycos.com> writes:D >We upgraded from 7.3 to 7.3-1 and now cannot talk to a TCPIP motor D >controller (Parker  CompuMotor 6k).  We cannot telnet or "connect" H >(socket code), whereas we could do both with 7.3.  We tried Linux, and . >vms 7.2, and they're both able to talk to it. > E >The motor controller is sort  of stupid; it does not respond to ARP :J >broadcasts, so we bind the MAC address to the IP number in vms before we J >connect, but the source and destination MAC addresses are in the connect D >packet.  The motor controller just does not want to respond to 7.3. > J >Has anyone else had tcpip trouble?  I guess we'll have to go back to 7.3.   Not likely.t8 TCPIP communication is done via the TCPIP addon product.- What is the version of your installed stack ?g. Did you (re)install it after the VMS upgrade ?G Do you know that TCPIP V5.3 (ECO2, with ECO3 almost here) and TCPIP_SSH 5 (which updates core components of TCPIP) is current ?t@ Do you know that other IP stacks (TCPware, Multinet) do also runI (sometimes much better and sometimes much cheaper than TCPIP) on V7.3-1 ?eB Did you try to reconfig TCPIP after the VMS and/or TCPIP upgrade ?   -- 0 Peter "EPLAN" LANGSTOEGERl% Network and OpenVMS system specialistR E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  # Date: Tue, 24 Jun 2003 02:13:16 GMTs; From: "John Gemignani, Jr." <jon-nope@thiswontworkossc.net>s% Subject: Re: TCPIP problem with 7.3-1 = Message-ID: <0tOJa.15952$Jw6.6781618@news1.news.adelphia.net>   6 "Irving F. Snurd" <dacadc0@lycos.com> wrote in message" news:3EF772F3.5010809@lycos.com...E > We're running the 5.3 stack, and our systems guy did the "mandatory I > TCPIP update" as part of the upgrade to VMS 7.3-1.  We're running on an.I > alpha DS10. We aren't sure what the stack was under 7.3.  Our alpha caniB > talk to PLCs and other machines.  Our systems guy broke down theC > connect packets and they look the same as the packets coming fromdA > machines that succeed in talking to the motor controller.  He'sbF > scratching his head; I was hoping some resourceful puzzle-solver sysH > admin out there would say, "Oh, yeah, we've seen that; just diddle the > @#@ parameter..."  >3 >         Bill Manwaring, IUCF >l  K Try showing some of the counters from the TCPIP utility and see if anything2F is being dropped.  You aren't running the two stacks at the same time,J right?  It's possible that you have turned something on/off in the upgradeL that doesn't match your previous configuration.  If the controller is really9 dumb, is it possible that it has checksumming turned off?.   -John      > Thomas H. Pauli wrote: >x > > Hi,t > >sF > > I think it's not VMS you have problems with, but the TCP/IP stack.I > > Which version do you run now ($ TCPIP SHOW VERSION) and which did youiL > > run before ? Perhaps you've got a TCP/IP upgrade wioth your VMS upgrade! > >f
 > > Thomas > >v7 > > On Mon, 23 Jun 2003 11:46:31 -0500, Irving F. Snurdd > > <dacadc0@lycos.com> wrote: > >tG > >> We upgraded from 7.3 to 7.3-1 and now cannot talk to a TCPIP motorcG > >> controller (Parker  CompuMotor 6k).  We cannot telnet or "connect"cG > >> (socket code), whereas we could do both with 7.3.  We tried Linux,e6 > >> and vms 7.2, and they're both able to talk to it. > >>H > >> The motor controller is sort  of stupid; it does not respond to ARPJ > >> broadcasts, so we bind the MAC address to the IP number in vms beforeH > >> we connect, but the source and destination MAC addresses are in theH > >> connect packet.  The motor controller just does not want to respond > >> to 7.3. > >>I > >> Has anyone else had tcpip trouble?  I guess we'll have to go back to3	 > >> 7.3.e > >>" > >> regards, Bill Manwaring, IUCF > >> > >> > >e > >d > >i >e   ------------------------------  # Date: Tue, 24 Jun 2003 00:45:32 GMTn1 From: Michael Austin <maustin@firstdbasource.com>n% Subject: Re: TCPIP problem with 7.3-1s2 Message-ID: <3EF79D63.4F8DFAA7@firstdbasource.com>   "Irving F. Snurd" wrote: > D > We upgraded from 7.3 to 7.3-1 and now cannot talk to a TCPIP motorD > controller (Parker  CompuMotor 6k).  We cannot telnet or "connect"H > (socket code), whereas we could do both with 7.3.  We tried Linux, and/ > vms 7.2, and they're both able to talk to it.I > E > The motor controller is sort  of stupid; it does not respond to ARPeJ > broadcasts, so we bind the MAC address to the IP number in vms before weJ > connect, but the source and destination MAC addresses are in the connectE > packet.  The motor controller just does not want to respond to 7.3.w > K > Has anyone else had tcpip trouble?  I guess we'll have to go back to 7.3.o > ' >         regards, Bill Manwaring, IUCF   B sounds like a routing issue and/or a port speed issue.  If you areA funning 5.3 make you install ECO2 --or you will have more strange3H problems than just this one... Been there, done that --- and didn't like it at all!!!   -- e Regards,  6 Michael Austin            OpenVMS User since June 19847 First DBA Source, Inc.    Registered Linux User #261163t7 Sr. Consultant            http://www.firstdbasource.com,   ------------------------------  % Date: Tue, 24 Jun 2003 15:36:03 +0200 $ From: "Peter Flunger" <p-i-b@gmx.at>% Subject: Re: TCPIP problem with 7.3-1h0 Message-ID: <bd9k44$2e8$1@newsreader1.netway.at>  + "Irving F. Snurd" <dacadc0@lycos.com> wrotef >"K > Has anyone else had tcpip trouble?  I guess we'll have to go back to 7.3.  >s  = Yes, we had a lot of problems after upgrading to VMS7.3 LAN03  and VMS 7.3-1 LAN02 ECOs.b< If you have a DE6xxx series card in your DS10 you might find" the following information usefull:  : HP implemented a new autosensing algorithm for their DE6xx( series ethernet HBAs. ( EIDRIVER Image )H If you are connected to a port ( switch/hub ) that supports autosensing,A you will have to change the ethernet Card on console level to uses autosenseing as well.l  ( set ewx0_mode auto_negotiate )L If you set the device to use FastFD and the Port on the switch is configured for I autonegotiation you will see the oddest things happen. ( huge packet loss@ rates,J no communication at all, or communication working but veeeeeer slow, aso )H So either set both sites to a fixed speed or use autonegotiation on both sides. HtH , Peter    ------------------------------   Date: 24 JUN 2003 14:27:13 GMT4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher)% Subject: Re: TCPIP problem with 7.3-1s6 Message-ID: <24JUN03.14271398@thuria.waisman.wisc.edu>  C In a previous article, "Irving F. Snurd" <dacadc0@lycos.com> wrote:eF ->We're running the 5.3 stack, and our systems guy did the "mandatory J ->TCPIP update" as part of the upgrade to VMS 7.3-1.  We're running on an J ->alpha DS10. We aren't sure what the stack was under 7.3.  Our alpha can B ->talk to PLCs and other machines.  Our systems guy broke down theD ->connect packets and they look the same as the packets coming from B ->machines that succeed in talking to the motor controller.  He's G ->scratching his head; I was hoping some resourceful puzzle-solver sys aI ->admin out there would say, "Oh, yeah, we've seen that; just diddle the   ->@#@ parameter..."i  E TCPIP 5.3 did change quite a few stack settings to be consistent with4C TCPIP on TRUE64. One of the changes that affected us was increasingrG tcp_keepidle from 150 (75 seconds) to 14400 (2 hours). You can set thate back to the 5.1 value with:t     $ @TCPIP$DEFINE_COMMANDS&   $ sysconfig -r inet tcp_keepidle=150   To list all inet settings:     $ sysconfig -q inete  B Others were changed too. The trouble is without a TCPIP 5.1 systemF around you won't know what the 5.1 values were as I've never found any documentation of the changes.n   --G -- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madisonr7 --                karcher.nomorespxm@waisman.wisc.edu  m   ------------------------------  % Date: Tue, 24 Jun 2003 21:15:49 +0100s* From: "John Travell" <john@travell.uk.net>% Subject: Re: TCPIP problem with 7.3-1r5 Message-ID: <bdabq3$r9khj$1@ID-120847.news.dfncis.de>o  A "Carl Karcher" <karcher@thuria.waisman.wisc.edu> wrote in message?0 news:24JUN03.14271398@thuria.waisman.wisc.edu...E > In a previous article, "Irving F. Snurd" <dacadc0@lycos.com> wrote:6G > ->We're running the 5.3 stack, and our systems guy did the "mandatory8K > ->TCPIP update" as part of the upgrade to VMS 7.3-1.  We're running on antK > ->alpha DS10. We aren't sure what the stack was under 7.3.  Our alpha candD > ->talk to PLCs and other machines.  Our systems guy broke down theE > ->connect packets and they look the same as the packets coming from C > ->machines that succeed in talking to the motor controller.  He'seH > ->scratching his head; I was hoping some resourceful puzzle-solver sysJ > ->admin out there would say, "Oh, yeah, we've seen that; just diddle the > ->@#@ parameter..."d >nG > TCPIP 5.3 did change quite a few stack settings to be consistent witheE > TCPIP on TRUE64. One of the changes that affected us was increasing I > tcp_keepidle from 150 (75 seconds) to 14400 (2 hours). You can set thate > back to the 5.1 value with:  >  >   $ @TCPIP$DEFINE_COMMANDS( >   $ sysconfig -r inet tcp_keepidle=150 >?  J You need to know why you want to set tcp_keepidle to any particular value.  L What tcp_keepidle does is control the elapsed time between successive checksF to see if the application initiating the IP connection is still alive. Example problem:J I used to have a dial-back-on-demand ISDN connection. This was intended toK be connected only when activity actually required it to be, and remain as aoK 'virtual' connection when idle. The problem was that I was getting the lineeF re-connected after only a minute's idle time even when I was at lunch.K After eliminating all of the suspected Windoze trash, the problem turned toe9 be tcpip on VMS, caused by a tcp_keepidle setting of 150.yD My ISDN disconnect timer was set to 90 seconds, so when the line wasL connected to do a keepalive check, it was still alive for the next keepaliveI 75 seconds later. The line was dropped at 90 seconds, then a minute later - was picked up again for the next keepalive...pI Setting tcp_keepidle to 14400 saved over a thousand pounds a year in ISDNl
 phone bills !e  L Before you set tcp_keepidle back to 150, you really need to know exactly WHY9 this is the right way to solve whatever problem you have.n   -- John Travell  VMS crashdump expertise for hire john@travell.uk.neto http://www.travell.uk.net/       ---6& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).A Version: 6.0.489 / Virus Database: 288 - Release Date: 10/06/2003f   ------------------------------   Date: 24 JUN 2003 21:45:23 GMT4 From: karcher@thuria.waisman.wisc.edu (Carl Karcher)% Subject: Re: TCPIP problem with 7.3-1I6 Message-ID: <24JUN03.21452326@thuria.waisman.wisc.edu>  B In a previous article, "John Travell" <john@travell.uk.net> wrote:  N ->Before you set tcp_keepidle back to 150, you really need to know exactly WHY; ->this is the right way to solve whatever problem you have.   G That's why I wish they had documented this major change. This is a big pE difference between TCPIP 5.1 and 5.3 which is related to this thread.uD I don't know what other parameters were changed but am told this was8 not the only one (I have no 5.1 system to compare with).  K > Setting tcp_keepidle to 14400 saved over a thousand pounds a year in ISDNm > phone bills !l  E In our case the change to 14400 resulted in leaving a file open afterl= a PC application (Eudora with mailboxes on a pathworks share) D abnormally terminated. This essentially caused the user to be lockedF out of the application for 2 hours. Goes to show how "your mileage may vary".   --G -- Carl Karcher, Waisman Computing Services, Waisman Center, UW-Madisong7 --                karcher.nomorespxm@waisman.wisc.edu  o   ------------------------------  % Date: Tue, 24 Jun 2003 18:59:02 -0500n( From: Michael Rice <marice@whiteice.com>% Subject: Re: TCPIP problem with 7.3-1f/ Message-ID: <vfhpeb7udj1oeb@corp.supernews.com>n  ) On 6/24/2003 4:45 PM, Carl Karcher wrote:,D > In a previous article, "John Travell" <john@travell.uk.net> wrote: > P > ->Before you set tcp_keepidle back to 150, you really need to know exactly WHY= > ->this is the right way to solve whatever problem you have.n > I > That's why I wish they had documented this major change. This is a big dG > difference between TCPIP 5.1 and 5.3 which is related to this thread.CF > I don't know what other parameters were changed but am told this was: > not the only one (I have no 5.1 system to compare with). >  > K >>Setting tcp_keepidle to 14400 saved over a thousand pounds a year in ISDN- >>phone bills !1 >  > G > In our case the change to 14400 resulted in leaving a file open aftera? > a PC application (Eudora with mailboxes on a pathworks share) F > abnormally terminated. This essentially caused the user to be lockedH > out of the application for 2 hours. Goes to show how "your mileage may > vary". >   G The 14400 value is the normal setting for most platforms.  The setting ,B is only used when TCP keepalive packets are being sent (i.e., the + SO_KEEPALIVE option is set for the socket).p  I I'm surprised that setting it back to 75 seconds was sufficient for you. lI   I've always seen (on other operating systems) the tcp_keepidle used in  H combination with other parameters such as tcp_keepcnt (usual default is H 8) and tcp_keepintvl (usual default is 75 seconds).  I wonder how these I parameters are used in VMS TCPIP.  IIRC, under "default" conditions, the  H stack will wait for 2 hours (tcp_keepidle) of inactivity before issuing H the first keepalive.  It will then try 8 times (tcp_keepcnt), issuing 1 D keepalive every 75 seconds (tcp_keepintvl).  This gives a "default" E keepalive timeout of 7800 seconds.  Just changing tcp_keepidle to 75 o9 seconds should have resulted in a timeout of 675 seconds.a   ------------------------------  # Date: Tue, 24 Jun 2003 23:37:39 GMTo0 From: HARANGOZO CSABA   <spameater@spam.invalid>% Subject: Re: TCPIP problem with 7.3-1 6 Message-ID: <7h5Ka.7747$eE.75929@nasal.pacific.net.au>  5 Carl Karcher <karcher@thuria.waisman.wisc.edu> wrote:g [...snip...]   > To list all inet settings: >  >  $ sysconfig -q inet > D > Others were changed too. The trouble is without a TCPIP 5.1 systemH > around you won't know what the 5.1 values were as I've never found any > documentation of the changes.s  A 	The following are the inet default values in TCPIP 5.1, *except*H 	tcp-recvspace & tcp_sendspace.n   $tcpip show ver   ?   Compaq TCP/IP Services for OpenVMS Alpha Version V5.1 - ECO 3a6   on a AlphaServer 8200 5/300 running OpenVMS V7.3-1     $sysconfig -q inet inet:  inifaddr_hsize = 32G
 ipdefttl = 64e ipdirected_broadcast = 0 ipforwarding = 0 ipfragttl = 60
 ipgateway = 0g ipport_userreserved = 65535a ipport_userreserved_min = 49152u ipqmaxlen = 1024 ipqs = 1 ipsendredirects = 1w ipsrcroute = 1 pmtu_decrease_intvl = 1200 pmtu_enabled = 1 pmtu_increase_intvl = 240> pmtu_rt_check_intvl = 20 subnetsarelocal = 1c tcbhashnum = 1 tcbhashsize = 512r tcbquicklisten = 1 tcp_compat_42 = 1e tcp_cwnd_segments = 2p tcp_dont_winscale = 0  tcp_keepalive_default = 0o tcp_keepcnt = 8e tcp_keepidle = 150 tcp_keepinit = 150 tcp_keepintvl = 150m tcp_msl = 60 tcp_mssdflt = 536  tcpnodelack = 1e tcp_recvspace = 49152$ tcp_rexmit_interval_min = 2u tcp_rexmtmax = 128 tcprexmtthresh = 3 tcp_rttdflt = 3s tcp_sendspace = 49152  tcp_ttl = 60 tcptwreorder = 0 tcp_urgent_42 = 1i udpcksum = 1 udp_recvspace = 42080  udp_sendspace = 9216 udp_ttl = 30 ovms_nobroadcastcheck = 0- $- 					Cheers,   Csaba  J  -------------------------------------------------------------------------H   CSABA I. HARANGOZO  |d|i|g|i|t|a|l|  csabah(at)zipworld(dot)com(dot)auJ  -------------------------------------------------------------------------;    EARTH::AUSTRALIA:[SYDNEY]HARANGOZO.CSABA;1, delete? [N]:v    Wood's Axiom :tI   As soon as a still-to-be-finished computer task becomes a life-or-death %           situation, the power fails.g   ------------------------------    Date: 23 Jun 2003 07:23:06 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)u/ Subject: Re: Tracing  EXE -> OBJ -> SRC modulesf3 Message-ID: <8Tl2xeHfICm7@eisner.encompasserve.org>s  _ In article <CIEJLCMNHNNDLLOOGNJIMEALHHAA.tom@kednos.com>, "Tom Linden" <tom@kednos.com> writes:cK > ANAL/IMAGE will tell me which linker was used to link the OBJ, but how do 4 > I uniquely determine which object module was used? > L > ANAL/OBJ will tel me which compiler was used, but how do I determine which > source module was used?a > L > Assume there are no log files.  IOW, given only an image can you determine > the OBJ and SRC modules?  C    Some compilers support the ident feature.  Macro and Fortran I'm.D    fairly sure of.  This is usefull if you actually change the ident    sometime.   ------------------------------   Date: 24 Jun 2003 07:18:28 GMT/ From: "Dave Weatherall" <djweath@attglobal.net>i/ Subject: Re: Tracing  EXE -> OBJ -> SRC modules 5 Message-ID: <DTiotGxQ0bj6-pn2-5C7hyXD7bqh4@localhost>-  " On Mon, 23 Jun 2003 12:23:06 UTC, < koehler@eisner.nospam.encompasserve.org (Bob Koehler) wrote:  a > In article <CIEJLCMNHNNDLLOOGNJIMEALHHAA.tom@kednos.com>, "Tom Linden" <tom@kednos.com> writes:9M > > ANAL/IMAGE will tell me which linker was used to link the OBJ, but how doo6 > > I uniquely determine which object module was used? > > N > > ANAL/OBJ will tel me which compiler was used, but how do I determine which > > source module was used?  > > N > > Assume there are no log files.  IOW, given only an image can you determine > > the OBJ and SRC modules? > E >    Some compilers support the ident feature.  Macro and Fortran I'mrF >    fairly sure of.  This is usefull if you actually change the ident >    sometime.  E This is the technique we use to trace/manage/check changes. Makes it  F straighforward to show our Q/A man that the proverbial one line changeD in one module is the only difference between one image and another. E Similarly all our objects are in libraries so the idents then become oF the only obvious differentiator between versions - unless one wants toF ANAL /OBJ /MHD on all the library modules. I've done that in times of  confusion too.   -- i Cheers - Dave.   ------------------------------  % Date: Mon, 23 Jun 2003 19:24:47 -0400 : From: Daniel T. Burrows <danburrows@nospam.mindspring.com># Subject: VAX - VMS license transfer 8 Message-ID: <jn2ffvormno24bjt9sj7mig9j5qa3o8vt6@4ax.com>  B How does a person go about transferring a commercial use VAX - VMS/ license to a PC running VAX emulation software?e  = If it is possible what are the costs and ordering procedures?S   Thanks Dan   E PS:   If this has been addressed recently please accept my appologiesM( as I have been off the list for a while.   ------------------------------    Date: 23 Jun 2003 19:02:22 -0500- From: Kilgallen@SpamCop.net (Larry Kilgallen)n' Subject: Re: VAX - VMS license transferp3 Message-ID: <ISCK2d0HJtYJ@eisner.encompasserve.org>   u In article <jn2ffvormno24bjt9sj7mig9j5qa3o8vt6@4ax.com>, Daniel T. Burrows <danburrows@nospam.mindspring.com> writes: D > How does a person go about transferring a commercial use VAX - VMS1 > license to a PC running VAX emulation software?r > ? > If it is possible what are the costs and ordering procedures?t  @ There is a web page describing it, that should be accessible via  http://www.openvms.compaq.com/ .   ------------------------------  % Date: Mon, 23 Jun 2003 21:58:42 -0400n. From: "Rob Lyons" <rob.lyons@resilientsys.com>' Subject: Re: VAX - VMS license transfer + Message-ID: <bd8b66$d6f$1@bob.news.rcn.net>h  ; > Thanks but the only related thing I could find there was:m@ > http://h71000.www7.hp.com/openvms/sri-charon-vax-emulator.html >n0 > There was nothing for any other VAX emulators.  G That's it, HP doesn't license their software on any other VAX emulator.   	 Rob Lyonsn Resilientsys.com   ------------------------------  % Date: Mon, 23 Jun 2003 20:46:58 -0400 : From: Daniel T. Burrows <danburrows@nospam.mindspring.com>' Subject: Re: VAX - VMS license transfer 8 Message-ID: <am7ffvsqj0fld1fcjch3kedvjqlbgia937@4ax.com>  F On 23 Jun 2003 19:02:22 -0500, Kilgallen@SpamCop.net (Larry Kilgallen) wrote:  v >In article <jn2ffvormno24bjt9sj7mig9j5qa3o8vt6@4ax.com>, Daniel T. Burrows <danburrows@nospam.mindspring.com> writes:E >> How does a person go about transferring a commercial use VAX - VMSm2 >> license to a PC running VAX emulation software? >> o@ >> If it is possible what are the costs and ordering procedures? > A >There is a web page describing it, that should be accessible via@! >http://www.openvms.compaq.com/ .e    9 Thanks but the only related thing I could find there was:o> http://h71000.www7.hp.com/openvms/sri-charon-vax-emulator.html  . There was nothing for any other VAX emulators.   ------------------------------  % Date: Tue, 24 Jun 2003 18:08:04 -0400.* From: "Stanley F. Quayle" <stan@stanq.com>' Subject: Re: VAX - VMS license transfero. Message-ID: <3EF89384.22405.8F5E81C@localhost>  G > >>>How does a person go about transferring a commercial use VAX - VMS 4 > >>>license to a PC running VAX emulation software?  A > >http://h71000.www7.hp.com/openvms/sri-charon-vax-emulator.htmlf > >i1 > >There was nothing for any other VAX emulators.f  F CHARON-VAX is a product that was developed by SRI in cooperation with @ HP (Compaq at the time).  Notice also on the same web page that F CHARON-VAX systems are eligible for OpenVMS software support from HP. E Many commercial users require software support, and so CHARON-VAX is n# the only choice for commercial use.a      
 --Stan Quayle  Quayle Consulting Inc.  
 ----------C Stanley F. Quayle, P.E. N8SQ  +1 614-868-1363  Fax: +1 614 868-1671x1 8572 North Spring Ct. NW, Pickerington, OH  43147v= Preferred address:  stan@stanq.com       http://www.stanq.como   ------------------------------    Date: 23 Jun 2003 07:40:56 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) / Subject: Re: Volume Locking on ANAL/DISK/REPAIRi3 Message-ID: <8QVj+4se0nVU@eisner.encompasserve.org>i  [ In article <3EF270F6.8994AFBE@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:o > B > ...but is unsupported. If it should trash your disk, you're SOL.  D    Not in my shop.  If my disks get trashed I have backups.  I don't    care what trashed them.   ------------------------------  % Date: Mon, 23 Jun 2003 21:25:59 -0500o1 From: "David J. Dachtera" <djesys.nospam@fsi.net> / Subject: Re: Volume Locking on ANAL/DISK/REPAIRe' Message-ID: <3EF7B6B7.157EAA47@fsi.net>s   Bob Koehler wrote: > ] > In article <3EF270F6.8994AFBE@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:i > >bD > > ...but is unsupported. If it should trash your disk, you're SOL. > F >    Not in my shop.  If my disks get trashed I have backups.  I don't >    care what trashed them.  G I do care - I don't want to have downtime for the same reason more thanuB once without knowing what happened and how to fix and/or avoid it.   -- t David J. Dachteraf dba DJE Systemsa http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/-   ------------------------------    Date: 24 Jun 2003 07:44:25 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)e/ Subject: Re: Volume Locking on ANAL/DISK/REPAIRo3 Message-ID: <tXiA9e6zE1Qz@eisner.encompasserve.org>t  [ In article <3EF7B6B7.157EAA47@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:t > Bob Koehler wrote: >>  ^ >> In article <3EF270F6.8994AFBE@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes: >> >E >> > ...but is unsupported. If it should trash your disk, you're SOL.9 >> gG >>    Not in my shop.  If my disks get trashed I have backups.  I don'tt >>    care what trashed them.6 > I > I do care - I don't want to have downtime for the same reason more thansD > once without knowing what happened and how to fix and/or avoid it.  H    I care, too.  But generally I equate the initials SOL with being much    deeper in S.$   ------------------------------  # Date: Mon, 23 Jun 2003 12:02:20 GMT + From: Jeff Cameron <JCam90502@jcameron.com>a% Subject: Re: Who can you trust, then? 2 Message-ID: <BB1C3A5B.9DDD%JCam90502@jcameron.com>  G On 6/21/03 6:20 PM, in article vfa15847nrop5e@corp.supernews.com, "GregR# Cagle" <gregc@gregcagle.com> wrote:-   > Bill Todd wrote: >  >>> I wrote: >>> A >>> So your basic argument is that these people can't be trusted.l >>   >> c >> Got it in one.o > ; > I've got an idea. There seems to be a lot of talk in this 9 > group about how you can't trust Compaq and by extension,< > HP. As a change, I was thinking it might be interesting to: > hear about vendors who CAN be trusted. Who would you bet6 > your career and your future (and your company's) on?   Digital Equipment Corporationt   ------------------------------    Date: 24 Jun 2003 14:30:48 -0500+ From: young_r@encompasserve.org (Rob Young)y Subject: WRITEMAP.SYSd3 Message-ID: <X0yK7ib9D7wg@eisner.encompasserve.org>   -         I have an idea that might be obvious.g  D         Since host-based mini-merge is coming to a future version ofK         VMS, I thought why not combine the tracking of "writes in progress"s         on to a WRITEMAP.SYS?e  %         The purpose would be twofold:t  M                 1)  Allowing a BACKUP/MODIFIED_BLOCKS backing up only blocks uH                         that have been written to since last full IMAGE.  K                         (Let's face it, many of us are backing up databaseswN                         every night and in some/many of these databases little0                         changes night to night).  H                 2)  Memory required to track "writes in  progress" could5                         be reduced and fixed in size.e    H         The layout of WRITEMAP.SYS would be an index file each data nodeG         would track 127 blocks at a time.  Each 127 block to be trackedY          would be a 9 byte entry.  F         Byte1  Byte2  Byte3  Byte4   Byte5  Byte6  Byte7  Byte8  Byte9G         |   Seconds of last write  | Number of writes         |  Misc |l  J         Bytes 1-4 would keep the time of the last write in seconds, offsetF         from an arbitrary base date - Jan 1, 2000  (future versions ofK         VMS would pick a higher base date if getting close to date overflowsK         68 years from now).  Byte5-8 would track how many writes took place M         between mounts.  Clear this field on DISMOUNT (what kind of overhead?mF         What if it isn't cleared because of crash?  Some background to8         clear if inconsistent or mounting after crash:  +         SET VOLUME/CLEAR=WRITEMAPCOUNTERS).e  E         Byte9 would be bits to play with.  Zero bit 1 when full IMAGE2J         backups occur.  flip Byte9 bit 1 on every write - no test and set.  B 	By tracking number of writes, you could find out very hot regions? 	and perhaps dump those blocks (or more techniques) to identifyu- 	what is going on with a program/data region.s  E         The number of blocks used by WRITEMAP.SYS wouuld be less thaneN         4000 blocks for a 9 GByte drive (9 bytes * 145134 entries + 1200 block7         index overhead - more overhead? less overhead?)m  %         A home block would look like:   F         Byte1  Byte2  Byte3  Byte4   Byte5  Byte6  Byte7  Byte8  Byte9D         | %   |    Reserved                                        |  I         Byte1 would track total percent that have been modifed by lookingkG         at bit1 of Byte9.  That way you could write a DCL test wrapper:i  P         $ percent_modifed =  f$getdvi("''Target'","PERCENT_MODIFED")!since image$         $ if percent_modifed .gt. 50         $ then i6         $       backup_quals = backup_quals + "/IMAGE"         $ elseK         $       backup_quals = backup_quals + "/MODIFED_BLOCKS/SINCE=IMAGE".         $ endifa  A 	The reason here is it makes little sense to do a modified backup-@ 	if 90% of your disk has changed.  Might as well do a full image< 	and have that to fall back to.  Conversely, it makes little> 	sense to do a full image if 5% of the disk has changed.  When@ 	doing 50, 100 GByte disks this can be a huge time savings - the
 	whole point.e  E         Now the idea is whoever is mastering the "writes in progress"wF         for the host-based minimerge feature would flip the associated&         bits for the 127 block region.  C         By tracking seconds since last write, you could alternatelyh>         backup modified blocks to the prior image and build upE         associated DCL to support that infrastructure.   Backup mightd         look like:  9         $ BACKUP/MODIFIED_BLOCKS/SINC="15-JUN-2003:03:00"e  = 	This would give the manager the comfort of knowing you could:? 	restore 1 or 2 images and still apply the differentials to gett 	back a sane volume.  C         Likewise, the mini-copy could keep the memory used small by I         tracking when the shadow was broken out and added back and simply!J         trundle down through looking at seconds since last write OR betterE         yet, flip bit2 on byte9 when shadow members are broken out onnF         every write, then shadow mini-copy would simply look at bit 2., 	If bit 2 is flipped, copy that region back.  #                                 Rob.   ------------------------------    Date: 24 Jun 2003 15:03:40 -0500+ From: young_r@encompasserve.org (Rob Young)t Subject: Re: WRITEMAP.SYSo3 Message-ID: <8wRw9hfa0ckX@eisner.encompasserve.org>t  a In article <X0yK7ib9D7wg@eisner.encompasserve.org>, young_r@encompasserve.org (Rob Young) writes:a >  > J >                 2)  Memory required to track "writes in  progress" could7 >                         be reduced and fixed in size.' >    	Make that:t  7 		Memory required to track blocks modifed for mini-copy % 		could be reduced and fixed in size.-   				Robe   ------------------------------    Date: 24 Jun 2003 13:05:56 -0500+ From: young_r@encompasserve.org (Rob Young)n( Subject: XFC docs/SYSGEN inconsistencies3 Message-ID: <y$APAazsXu+x@eisner.encompasserve.org>    Looking at docs:  < http://h71000.www7.hp.com/doc/731FINAL/6017/6017pro_081.html  E For example, if your system has 128GB of memory, you may find that byaO dedicating 8GB of memory to XFC, your cache hit rates and overall response timeo# is consistently good. For example:         $ RUN SYS$SYSTEM:SYSGENe SYSGEN> USE CURRENTP SYSGEN> SET VCC_MAX_CACHE 8000 SYSGEN> WRITE ACTIVE $ RUN SYS$SYSTEM:SYSMANeE SYSMAN> RESERVED_MEMORY ADD VCC$MIN_CACHE_SIZE /SIZE=8000 /ALLOCATE -t6 _SYSMAN> /NOGLOBAL_SECTION /NOZERO /NOPAGE_TABLE/RAD=0         . 18.5.2 Controlling the Maximum Cached I/O SizeM The dynamic system parameter VCC_MAX_IO_SIZE controls the maximum size of I/OuJ that can be cached by XFC. This parameter specifies the size in blocks. By default, it is 128.    ---.  
 	Two things..   6 	1)  SYSGEN shows a MAX of 3700000 blocks under 7.3-1:   $ mcr sysgen show vcc_maxsizetH Parameter Name           Current    Default     Min.      Max.     Unit  Dynamic H --------------           -------    -------    -------   -------   ----  -------tK VCC_MAXSIZE               1024000       6400         0    3700000 Blocks   c   	Maybe that should be -1 or 0?  A 	Secondly, the docs say by default VCC_MAX_IO_SIZE is 128.  It isr 	127:r  ! $ mcr sysgen show vcc_max_io_sizerH Parameter Name           Current    Default     Min.      Max.     Unit  DynamicaH --------------           -------    -------    -------   -------   ----  -------iN VCC_MAX_IO_SIZE               127        127         0         -1 Blocks     D     				Robh   ------------------------------  % Date: Mon, 23 Jun 2003 17:11:12 +0100AO From: Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com>aC Subject: Re: [OT] SCO tells IBM no more AIX ... IBM better buy VMS!u0 Message-ID: <bd78r1$krd$1@new-usenet.uk.sun.com>   John Eisenschmidt wrote:  E > To be honest, and smarter people than me have speculated on this, InE > don't think IBM can buy SCO. The legal affect of one SYS V licensee / > buying it up...what about Sun and SGI and HP?A > @ > Unless they bought SCO and released the SYS V source into Open
 > Source.  >   > Sun would be unaffected, we bought out the SYS V IP in Solaris? from Novell for a one off fee before SCO bought it from Novell.   ? The case doesn't have any issues for Sun nor would the transfer > of the SYS V IP and licensing ownership to IBM (if they bought SCO) have any effect on Sun.  > I would doubt that SCO are likely to sue HP, HP's contribution> to Linux OpenSource has been negligable so it is unlikely that5 SCO IP has leaked to OpenSource through the HP route.   B SGI are a significant OpenSource contributor, XFS and a reasonable ammount of Linux kernel work.e  D However they wern't part of the ill fated project Monteray (IBM/SCO)0 which IBM pulled the plug on to SCO's annoyance.   Regardse Andrew Harrisong   ------------------------------  % Date: Tue, 24 Jun 2003 14:29:34 -0700d+ From: "Barry Treahy, Jr." <Treahy@MMaz.com>t2 Subject: Re: [SPAM] Re: VAX - VMS license transfer' Message-ID: <3EF8C2BE.4030401@MMaz.com>E   Daniel T. Burrows wrote:  G >On 23 Jun 2003 19:02:22 -0500, Kilgallen@SpamCop.net (Larry Kilgallen)y >wrote:h >o >  I >tw >>In article <jn2ffvormno24bjt9sj7mig9j5qa3o8vt6@4ax.com>, Daniel T. Burrows <danburrows@nospam.mindspring.com> writes:  >>     >>E >>>How does a person go about transferring a commercial use VAX - VMSA2 >>>license to a PC running VAX emulation software? >>>l@ >>>If it is possible what are the costs and ordering procedures?	 >>>      s >>>mB >>There is a web page describing it, that should be accessible via" >>http://www.openvms.compaq.com/ . >>     >> >n >@: >Thanks but the only related thing I could find there was:? >http://h71000.www7.hp.com/openvms/sri-charon-vax-emulator.html  > / >There was nothing for any other VAX emulators.s >e >  s >FG I do not believe any of the 'other' emulators have passed the suite of sI actual HP hardware diagnostics, as Charon has, so HP most likely doesn't -I view them as machines acceptable to run VMS, let alone transfer licenses n too for commercial purposes...   Barrya   -- r  @ Barry Treahy, Jr  *  Midwest Microwave  *  Vice President & CIO   A E-mail: Treahy@mmaz.com * Phone: 480/314-1320 * FAX: 480/661-7028    ------------------------------    Date: 23 Jun 2003 07:59:38 -0500; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)r! Subject: Re: __PAL_REMQUEQ from Ct3 Message-ID: <JEQ6+bWDjKLI@eisner.encompasserve.org>   w In article <JMtJa.60930$Io.5695755@newsread2.prod.itd.earthlink.net>, "Brian Kennedy" <kennedybg@earthlink.net> writes:pM > Has anyone used the __PAL_REMQUEQ function defined in the BUILTINS.H header L > file? This function differs from the other REMQUEx functions in that it isK > suppose to remove an entry from anywhere in a queue, not just the head oraK > the tail. However, I get an ACCVIO when I use the function. The followingxL > code will generate the ACCVIO on an Alpha running OpenVMS 7.2. Can someoneM > shed some light on why this is happening? Documentation on this function isi) > almost non-existent. Thanks in advance.   I    PAL REMQUE calls should emulate the 32 bit VAX REMQUE instruction.  I  I    don't know what the Q at the end of _PAL_REMQUEQ is supposed to stand :H    for (my architecture manual is in another building) but I see you're H    trying to manipulate 64 bit addresses.  Are you sure _PAL_REMQUEQ is     supposed to handle this?h  C    All the PAL calls are fully documented in the Alpha architecturenC    reference, and all the VAX instructions that are replaced by PALsB    calls on the Alpha are fully documented in the VAX architecture
    reference.d   ------------------------------  # Date: Mon, 23 Jun 2003 14:23:59 GMT & From: John Reagan <john.reagan@hp.com>! Subject: Re: __PAL_REMQUEQ from Co2 Message-ID: <34EJa.3134$X%6.1725@news.cpqcorp.net>   Brian Kennedy wrote:M > Has anyone used the __PAL_REMQUEQ function defined in the BUILTINS.H headerrL > file? This function differs from the other REMQUEx functions in that it isK > suppose to remove an entry from anywhere in a queue, not just the head ort > the tail.   D REMQUEQ removes an entry from a quadword absolute queue.  Just like 7 REMQUE removes an entry from a longword absolute queue.d     > . > int main( int arg_count, char* arg_array[] ) > {t > #define MAX_ENTRIES 32" >     Data*    array[MAX_ENTRIES]; >     Data*    queueitem;" >     uint64    vmsize;s >     unsigned long counter; >     unsigned int result; > : >     for ( counter =0; counter < MAX_ENTRIES; counter++ ) >     {o  >         vmsize = sizeof(Data);  H Since Data is a pointer, I suspect 'vmsize' is 4, not what you intended.  Look at it in the debugger, yes?  7 >         LIB$GET_VM_64( &vmsize, &array[counter], 0 );t* >         array[counter]->value = counter;? >         result = __PAL_INSQTIQ( &QueueHead, array[counter] );u >     }      -- o John Reagano' Compaq Pascal/{A|I}MACRO Project LeaderI Hewlett-Packard Companyt   ------------------------------  % Date: Mon, 23 Jun 2003 15:28:24 +0100a* From: "Richard Brodie" <R.Brodie@rl.ac.uk>! Subject: Re: __PAL_REMQUEQ from Ch, Message-ID: <bd72qa$11ha@newton.cc.rl.ac.uk>  : "Brian Kennedy" <kennedybg@earthlink.net> wrote in message? news:JMtJa.60930$Io.5695755@newsread2.prod.itd.earthlink.net...   M > Has anyone used the __PAL_REMQUEQ function defined in the BUILTINS.H header-L > file? This function differs from the other REMQUEx functions in that it isK > suppose to remove an entry from anywhere in a queue, not just the head or9 > the tail.s  H And also that it operates on absolute, rather than self-relative queues.  8 > Documentation on this function is almost non-existent.  O http://h71000.www7.hp.com/doc/73final/4515/4515pro_021.html#8_queueinstructions   N The best place for introductory information is to read the VAX instruction setC reference manual, together with the AARM  for how the Alpha PALcodetM differs. The addition of quadword queue operations is one obvious difference.f   ------------------------------  # Date: Tue, 24 Jun 2003 02:01:46 GMTo/ From: "Brian Kennedy" <kennedybg@earthlink.net>e! Subject: Re: __PAL_REMQUEQ from ClE Message-ID: <eiOJa.62162$Io.5843100@newsread2.prod.itd.earthlink.net>F  K The link was a huge help. I did try searching the net, including HP's site,tJ for information on REMQUEQ prior to posting. I guess I just didn't use the, right keywords. Anyway, thanks for the link.    5 "Richard Brodie" <R.Brodie@rl.ac.uk> wrote in messaget& news:bd72qa$11ha@newton.cc.rl.ac.uk... > < > "Brian Kennedy" <kennedybg@earthlink.net> wrote in messageA > news:JMtJa.60930$Io.5695755@newsread2.prod.itd.earthlink.net...: >:H > > Has anyone used the __PAL_REMQUEQ function defined in the BUILTINS.H headerK > > file? This function differs from the other REMQUEx functions in that ite isJ > > suppose to remove an entry from anywhere in a queue, not just the head or
 > > the tail.u >aJ > And also that it operates on absolute, rather than self-relative queues. >t: > > Documentation on this function is almost non-existent. >c >eL http://h71000.www7.hp.com/doc/73final/4515/4515pro_021.html#8_queueinstructi ons  >kL > The best place for introductory information is to read the VAX instruction set E > reference manual, together with the AARM  for how the Alpha PALcodesC > differs. The addition of quadword queue operations is one obviousd difference.# >a >e >. >i >    ------------------------------   End of INFO-VAX 2003.347 ************************