1 INFO-VAX	Fri, 15 Dec 2000	Volume 2000 : Issue 698       Contents:+ Re: "process crash" vs. "application crash" + Re: "process crash" vs. "application crash" ? %INSTALL-I-NONSHRADR, image installed ignoring '/SHARE=ADDRESS' , A wonderful message from an antivirus vendor0 Re: A wonderful message from an antivirus vendor Advance Server and NT2000  Re: Advance Server and NT2000  RE: Advance Server and NT2000   Re: Boing boing boing... splash?! Re: Cluster advertising in the UK  Re: Decserver 700  Re: Decserver 700  RE: EV68  was Re: Sun Cluster  Re: EV68  was Re: Sun Cluster  Help porting socket routines  Re: Help porting socket routines& Re: IBM and Linux ... purfect together& Re: IBM and Linux ... purfect together Mozilla 0.6  Re: Old VAX machine  Re: Old VAX machine  Re: Sun Cluster  Re: Sun Cluster  Re: Sun Cluster  Re: Sun Cluster  Re: Sun Cluster  Re: Sun Cluster  Re: Sun Cluster  Re: Sun Cluster  Re: Sun Cluster  TCP Bug - MULTINET + UCX- VAX - ALPHA Fortran common block differences? 1 Re: VAX - ALPHA Fortran common block differences? 1 Re: VAX - ALPHA Fortran common block differences? 1 Re: VAX - ALPHA Fortran common block differences? 1 Re: VAX - ALPHA Fortran common block differences? $ RE: vax/vms system as a DHCP client?  Re: Virtual I/O Cache and Oracle Re: VMS "froze"  Re: VMS "froze"  Re: Weird I/O problem   F ----------------------------------------------------------------------  # Date: Thu, 14 Dec 2000 19:21:31 GMT + From: John Santos <john.santos@verizon.net> 4 Subject: Re: "process crash" vs. "application crash"> Message-ID: <MPG.14a2e7ede0f3949998969f@news.bellatlantic.net>  D In article <91an79$qgb$1@nnrp1.deja.com>, david_dawkins@my-deja.com  says... A > Thanks to everyone for your help. We've now done the following:  > & > - given the process *all* privileges? > - doubled the quotas beyond the known-to-be-sufficient values ) > - turned on detached process accounting & >   ($ SET ACCOUNTING/ENABLE=DETACHED)6 > - run the process detached using LOGINOUT ("heavy"):9 >   $ RUN/DETACH/AUTHORIZE/ACCOU/INPUT=R.COM LOGINOUT.EXE   I You aren't specifying an output or error file (/output=r.log/error=r.err) G on the RUN command.  I don't know if this is relevant, especially since G you reassign sys$output and sys$error later in R.COM, but any port in a 	 storm ;-)   D I have found a couple of problems (bugs? features?) with starting upD detached processes.  One is the /error file gets appended to, so youF can't tell if the crash dump is current or ancient.  We usually renameF the .err file (our convention for the /error file type) to .err_old inH our startup procedure) and create a new .err file and write a time-stamp& to it so we can be sure it is current.  B If you don't write anything to the /output file (.log file in our E convention), sometimes the process dies and leaves it empty.  We make G the program write something, usually a program version number/date/time B banner right when the program starts up.  I think the issue is theG program doesn't actually open sys$output (assign a channel to it) until F it needs to print something, and sometimes when the program dies it isE due to a resource problem that prevents it from opening any files and E thus prevents it from printing any embedded debugging messages in the D program.  I don't know if this is a VAX/DEC BASIC issue or a genericF VMS issue.  I'm also not sure if it applies only to running a detachedE program directly or also to running LOGINOUT and using a .com file to  start the program.  B If you run the program directly from the "run/detach" command, youD can't read the /output file.  Apparently read-sharing isn't enabled.E If you run LOGINOUT, you can read the /output file.  In the 1st case, 7 you can sometimes see information in the file by using  B BACKUP/IGNORE=INTERLOCK to copy it, then SET FILE/END on the copy.E This depends on what has actually made it to disk and is not still in  the output buffers.   I For that matter, it may be that you don't see anything in the log because F info was written, but the file wasn't closed properly (file attributes1 not updated).  SET FILE/END might help here, too.    >   with R.COM being >   $ SET NOON* >   $ DEFINE SYS$OUTPUT APPHOME:OUTPUT.LOG) >   $ DEFINE SYS$ERROR  APPHOME:ERROR.LOG  >   $ SET PROC/DUMP / >   $ RUN/NODEBUG APPHOME:APPLICATION_DEBUG.EXE  >   >   (APPHOME is a group logical) > > > - Used ANAL/SYS to check that the running process's DUMP bit >   in the PCB is set. > C > - Installed an "exit" handler using "atexit" to just be sure that H >   the program isn't doing a crafty exit down in some library somewhereC >   (the handler would write to the log file and then call abort())  > E > - Installed a handler using $establish() to fclose(stderr) and then ( >   rethrow (by returning SS$_RESIGNAL). > < > Despite all this, it crashed again yesterday, and still noH > accounting entry, no dump file and no stack dump in APPHOME:ERROR.LOG. > ' > I'm going to take up chicken farming.  >  > Regards, and thanks again. > David Dawkins    Hope this helps!   --   John Santos    ------------------------------  % Date: Thu, 14 Dec 2000 23:55:41 -0500 , From: Howard S Shubs <hshubs@mindspring.com>4 Subject: Re: "process crash" vs. "application crash"> Message-ID: <hshubs-F70D90.23554114122000@news.mindspring.com>  J In article <91an79$qgb$1@nnrp1.deja.com>, david_dawkins@my-deja.com wrote:  % >- given the process *all* privileges    ARRRRRRGGHHHHH!!!!!!!!!!    ; >Despite all this, it crashed again yesterday, and still no G >accounting entry, no dump file and no stack dump in APPHOME:ERROR.LOG.   M Yes, debug it for real, perhaps with help from Compaq if you've got software   support. --   Howard S ShubsD "Run in circles, scream and shout!"  "I hope you have good backups!"   ------------------------------   Date: 15 Dec 2000 03:32:22 GMT* From: helbig@astro.rug.nl (Phillip Helbig)H Subject: %INSTALL-I-NONSHRADR, image installed ignoring '/SHARE=ADDRESS'. Message-ID: <91c3c6$7a7$1@info.service.rug.nl>  ! Is this something to worry about?   ? %ADX-I-INSTALLDONE, OpenVMS Alpha Debug V7.2X installation done N %VMSINSTAL-I-MOVEFILES, Files will now be moved to their target directories...P %INSTALL-I-NONSHRADR, image installed ignoring '/SHARE=ADDRESS' DISK$ALPHASYS:<S YS0.SYSCOMMON.SYSLIB>TRACE.EXEG -SYSTEM-F-NOPRIV, insufficient privilege or object protection violation    ------------------------------  % Date: Fri, 15 Dec 2000 04:05:35 +0100   From: Paul Sture <paul@sture.ch>5 Subject: A wonderful message from an antivirus vendor + Message-ID: <VA.000001dc.0383e61c@sture.ch>    Security ahoy!  F This message comes from a would be antivirus vendor, www.zonelabs.com.  @ >This is a multi-part message in MIME format. If you are reading> >this text, you may want to consider changing to a mail readerB >or gateway that understands how to properly handle MIME multipart
 >messages.   Are they kidding, or what? ___ 
 Paul Sture Switzerland    ------------------------------  % Date: Thu, 14 Dec 2000 19:31:06 -0800 ! From: Shane.F.Smith@Healthnet.com 9 Subject: Re: A wonderful message from an antivirus vendor D Message-ID: <OF5252494A.E1E4D649-ON882569B6.00133829@foundation.com>  I Heh, their text sucks, but their personal firewall product "ZoneAlarm" is  really good. Pity it's PC only.    Shane           4 Paul Sture <paul@sture.ch> on 12/14/2000 07:05:35 PM   Please respond to paul@sture.ch    To:   Info-VAX@Mvb.Saic.Com  cc:   6 Subject:  A wonderful message from an antivirus vendor     Security ahoy!  F This message comes from a would be antivirus vendor, www.zonelabs.com.  @ >This is a multi-part message in MIME format. If you are reading> >this text, you may want to consider changing to a mail readerB >or gateway that understands how to properly handle MIME multipart
 >messages.   Are they kidding, or what? ___ 
 Paul Sture Switzerland    ------------------------------  # Date: Thu, 14 Dec 2000 20:13:06 GMT  From: hornjt@my-deja.com" Subject: Advance Server and NT2000) Message-ID: <91b9kb$bmn$1@nnrp1.deja.com>   H Does anyone know if Advance Server on Openvms is compatible with NT2000.G We are looking at setting up Advance Server on OpenVms 7.2-1 to connect D to an nt4 domain, and were wondering if when we upgrade NT to NT2000 will advance server still work. # Any assistance will be appreciated.      Sent via Deja.com  http://www.deja.com/   ------------------------------    Date: 14 Dec 2000 21:34:04 +0100* From: eplan@kapsch.net (Peter LANGSTOEGER)& Subject: Re: Advance Server and NT2000* Message-ID: <3a392ebc$1@news.kapsch.co.at>  D In article <91b9kb$bmn$1@nnrp1.deja.com>, hornjt@my-deja.com writes:I >Does anyone know if Advance Server on Openvms is compatible with NT2000. H >We are looking at setting up Advance Server on OpenVms 7.2-1 to connectE >to an nt4 domain, and were wondering if when we upgrade NT to NT2000   >will advance server still work.  . 	http://www.openvms.compaq.com/advancedserver/= 	http://www.openvms.compaq.com/pathworks/pw_win2000_stat.html    --  < Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888 < <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------  % Date: Thu, 14 Dec 2000 16:04:33 -0500 - From: "Mitchell, David R." <mitchell@WPI.EDU> & Subject: RE: Advance Server and NT2000B Message-ID: <FE1835D68492D311BF7900508B5BEB0D0DECEC@petra.wpi.edu>  K The "official" support for Windows 2000 wont come until Advanced Server 7.3 H due out real-soon-now.  But, they will provide "best effort" support for( Advanced Server 7.2A and Pathworks 6.0C.  L As for real world experience, we recently upgraded our NT 4.0 Primary DomainK Controller to Windows 2000 and had some initial problems.  Most were solved E by applying ECO3 for Advanced Server.  The rest were fixed by a patch C supplied by Compaq service just about a week ago.  Since that patch ! everything has worked flawlessly.    David    > -----Original Message-----6 > From: hornjt@my-deja.com [mailto:hornjt@my-deja.com]+ > Sent: Thursday, December 14, 2000 3:13 PM  > To: Info-VAX@Mvb.Saic.Com $ > Subject: Advance Server and NT2000 >  > > > Does anyone know if Advance Server on Openvms is compatible  > with NT2000.? > We are looking at setting up Advance Server on OpenVms 7.2-1   > to connectF > to an nt4 domain, and were wondering if when we upgrade NT to NT2000! > will advance server still work. % > Any assistance will be appreciated.  >  >  > Sent via Deja.com  > http://www.deja.com/ >    ------------------------------  % Date: Thu, 14 Dec 2000 14:17:33 -0500 % From: "Brian Tillman" <tillman_brian> ) Subject: Re: Boing boing boing... splash? $ Message-ID: <3a391c0b$1@news.si.com>  I >Score is two balls (one I got in Vienna, the other through the post) and  >they both work.K >Three posters (which are still in the tubes until I find somewhere to hang # >them, but I've not forgotten them) % >At  least one postcard (maybe more).    I've gotten none of these! :-( --  A Brian Tillman                   Internet: tillman_brian at si.com A Smiths Industries, Inc.                   tillman at swdev.si.com = 3290 Patterson Ave. SE, MS      Addresses modified to prevent < Grand Rapids, MI 49512-1991     SPAM.  Replace "at" with "@"8        This opinion doesn't represent that of my company   ------------------------------    Date: 15 Dec 2000 01:14:31 +0800, From: Paul Repacholi <prep@prep.synonet.com>* Subject: Re: Cluster advertising in the UK0 Message-ID: <877l53dix4.fsf@k9.prep.synonet.com>  / Tim Llewellyn <tim.llewellyn@bbc.co.uk> writes:   G > I've also got a DHL package from Compaq to pickup. Lucky me, I wonder  > if it will bounce.  D I had one yesterday. Wow, it was only the week before I ordered! Sue# sorted that one well. Thanks folks.   B I quick glance stopped my eye at the security section. BZZT I hate= to bring this up, but the is NO VMS listing on the ITSEC list < for C2 or B1, not event on the pending list. And that is theF November updated list. So when will VMS be at least 'under evaluation' just like NT ;)    --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------    Date: 14 Dec 2000 14:25:07 -05002 From: malmberg@eisner.decus.org (John E. Malmberg) Subject: Re: Decserver 700+ Message-ID: <eBsP$bBKW84x@eisner.decus.org>   Y In article <PJ5_5.32$ys.2191@news4.atl>, "Robert Greene" <rgreene@atc.skytel.com> writes: N > Is this is a good place to ask a question about configuring a decserver 700? >  > K > How do I configure the decserver 700 to request it's load (WWENG2) from a  > specific IP node?  > N > I just bought the thing a couple of months ago and now the network admin hasJ > put it behind a router that only transfers IP. It requests a load from aI > node (IP) but the decserver never gets the load file... the LED display K > cycles from 4-2 over and over and the decserver just keeps trying to load  > the file from the same node. >  > < > If I could specify what node contains the load file,  then > maybe I could get it to load.  >  > I > If I hook the server up to another router (bad me, not allowed) then it  > works fine.... > N > I am fairly familiar with decserver in a MOP or LAT environment but not sure > how to handle this one...  >  >  >  >  >  >    ------------------------------    Date: 14 Dec 2000 14:36:51 -05002 From: malmberg@eisner.decus.org (John E. Malmberg) Subject: Re: Decserver 700+ Message-ID: <BZkBCW57oKGz@eisner.decus.org>   ( In article <PJ5_5.32$ys.2191@news4.atl>,4 "Robert Greene" <rgreene@atc.skytel.company> writes:? > Is this is a good place to ask a question about configuring a  > decserver 700?   Why not?  ; > How do I configure the decserver 700 to request it's load # > (WWENG2) from a specific IP node?   B I do not know exactly.  I would guess that the load is using BOOTP6 followed by TFTP after the host validates the request.  ? I have little if any memorable experience with a DECserver 700.   < > I just bought the thing a couple of months ago and now the> > network admin has put it behind a router that only transfersA > IP. It requests a load from a node (IP) but the decserver never @ > gets the load file... the LED display cycles from 4-2 over and@ > over and the decserver just keeps trying to load the file from > the same node.  @ Routers generally block BOOTP and DHCP requests by default.  Use8 caution when enabling a router to do "BOOTP forwarding".  < > If I could specify what node contains the load file,  then > maybe I could get it to load.   < Are you also able to specify it's IP address and the Router?  A > If I hook the server up to another router (bad me, not allowed)  > then it works fine....  . That router may have BOOTP forwarding enabled.  A > I am fairly familiar with decserver in a MOP or LAT environment ( > but not sure how to handle this one...  C The CSC may be able to assist with explaining how the load sequence  works.   -John  wb8tyw@qsl.network Representing only myself   ------------------------------  % Date: Thu, 14 Dec 2000 19:05:03 -0500 / From: "McCarthy Kevin P." <McCarthyKP@BWSC.ORG> & Subject: RE: EV68  was Re: Sun Cluster: Message-ID: <4C519CCC638BD411A4270000F8CD1D82075B62@NTSV2>  F Is the 833MHz es40 a ev68 part ??  any idea when the 833 MHz ES40 will be available??   Kevin   F (trying to convince my boss that Terry's newsletter is a good purchase+ although it does not pass the "bulk" test.)    -----Original Message-----9 From: Terry C. Shannon [mailto:terryshannon@mediaone.net] . Posted At: Thursday, December 14, 2000 6:44 PM Posted To: vms Conversation: Sun Cluster  Subject: Re: Sun Cluster      9 "Paul Repacholi" <prep@prep.synonet.com> wrote in message * news:87d7evdjb0.fsf@k9.prep.synonet.com...* > Rudolf Wingert <win@fom.fgan.de> writes: > E > > Last year did we hear about a 1.6Ghz Alpha. But we do see nothing  about B > > a real product till today. When can we see Alpha produced in a 13microV0 > > technology? I do see nothing on the horizon.  G EV69 in 0.13 by 2002 according to info gathered yesterday. EV68 in 0.18h/ exists today; I have one of these 1GHz wonders.    ------------------------------  # Date: Fri, 15 Dec 2000 04:25:42 GMTn4 From: "Terry C. Shannon" <terryshannon@mediaone.net>& Subject: Re: EV68  was Re: Sun Cluster; Message-ID: <a3h_5.5352$1M.1136411@typhoon.ne.mediaone.net>   : "McCarthy Kevin P." <McCarthyKP@BWSC.ORG> wrote in message4 news:4C519CCC638BD411A4270000F8CD1D82075B62@NTSV2...H > Is the 833MHz es40 a ev68 part ??  any idea when the 833 MHz ES40 will > be available??  F The 833MHz ES40 should be available in 1Q01. Benchmarks reflecting theF 833MHz part first went public in June. I believe this part has the 8MBJ cache. Later in the year comes the ES45 "Privateer," which'll have a ~1GHzL EV68 part, new memory subsystem, 4 PCI slots at 66MHz vice 2 at 33MHz on theG incumbent ES40. I have more details on Privateer, including info on thet0 bandwidth increase, in my "low-bulk" newsletter.   cheers,n   terry se >c > Kevin  > H > (trying to convince my boss that Terry's newsletter is a good purchase- > although it does not pass the "bulk" test.)o >n > -----Original Message-----; > From: Terry C. Shannon [mailto:terryshannon@mediaone.net]e0 > Posted At: Thursday, December 14, 2000 6:44 PM > Posted To: vms > Conversation: Sun Cluster- > Subject: Re: Sun Cluster >- >- > ; > "Paul Repacholi" <prep@prep.synonet.com> wrote in message., > news:87d7evdjb0.fsf@k9.prep.synonet.com..., > > Rudolf Wingert <win@fom.fgan.de> writes: > >eG > > > Last year did we hear about a 1.6Ghz Alpha. But we do see nothing1 > about>D > > > a real product till today. When can we see Alpha produced in a	 > 13micro 2 > > > technology? I do see nothing on the horizon. >lI > EV69 in 0.13 by 2002 according to info gathered yesterday. EV68 in 0.18e1 > exists today; I have one of these 1GHz wonders.  >S >A >U   ------------------------------  # Date: Thu, 14 Dec 2000 23:28:52 GMTv& From: trevor <trevor_deja@my-deja.com>% Subject: Help porting socket routiness* Message-ID: <3A3965F2.6F5FBF0@my-deja.com>   Hi  G I'm trying to port an application written for VMS & multinet to VMS andaD UCX.  Along the way I've run into problems with some of the C socket3 routines.  For example ioctl(s, SIOCSIFADDR, &ifr).f  G What I really need is documentation that describes the socket routines,hG the function codes and the arguments.  I know the C socket routines areuC described in the C manuals, but this is at a relatively superficialoG level.  What I need is an I/O User Reference Manual type description of H the socket routines which gives more detail.  Can anyone suggest where I will find such information.  Tan   Regards Trevor   trevor_deja@my-deja.comv   ------------------------------  # Date: Fri, 15 Dec 2000 01:32:48 GMTr  From: cstranslations@my-deja.com) Subject: Re: Help porting socket routinesg) Message-ID: <91bsc1$r02$1@nnrp1.deja.com>   ? Have you looked at DIGITAL TCP/IP Services for OpenVMS C SocketaG Programming? It's part of the TCP/IP Services for OpenVMS. You can findr it on line at:  3 http://www.openvms.compaq.com:8000/index.html#tcpipn  C The part number use to be AA-LU51K-TE. It's out of date and I don't . have the new one in front of me at the moment.   JoeA    * In article <3A3965F2.6F5FBF0@my-deja.com>,)   trevor <trevor_deja@my-deja.com> wrote:  > Hi >dE > I'm trying to port an application written for VMS & multinet to VMS  andrF > UCX.  Along the way I've run into problems with some of the C socket5 > routines.  For example ioctl(s, SIOCSIFADDR, &ifr).r >-? > What I really need is documentation that describes the sockett	 routines,IE > the function codes and the arguments.  I know the C socket routinesg areAE > described in the C manuals, but this is at a relatively superficialdF > level.  What I need is an I/O User Reference Manual type description ofB > the socket routines which gives more detail.  Can anyone suggest where Io! > will find such information.  Tah >i > Regards Trevor >e > trevor_deja@my-deja.comp >e     Sent via Deja.com  http://www.deja.com/   ------------------------------  % Date: Thu, 14 Dec 2000 21:49:31 +0100L" From: "Hans Vlems" <hvlems@iae.nl>/ Subject: Re: IBM and Linux ... purfect togetherU( Message-ID: <91bbm8$hnk$1@news.IAEhv.nl>  8 IBM's VM in no way compares to what a Galaxy implements.C VM allows you to run several guests OS's, irrespective the hardware G resources. In fact you can run VM as a guest itself. The host (primary) B VM translates hardware devices to the software devices used by its various guests.uA A galaxy can run several instances of VMS but they just share the C hardware and each instance uses at least one, privately owned, cpu. K What comes close to that concept on IBM 370/390 systems is something calleddK PR/SM ("prism") which partitions the hardware. That'll allow you to run twoM VSEOK instances on the same machine. While (a lot) less flexible than running VM,kI it is also considerably cheaper than a VM license that needs renewel each- year.-! VM is nice, but rather expensive.   
 Hans Vlems  ' Nigel Arnot heeft geschreven in berichtn. <009F488C.7195157E.31@maxwell.ph.kcl.ac.uk>... >> Why do we need software VM ?h >>' >> Is Galaxy itself not a hardware VM ?o >>K >> A GS320 with a 2-node VMS cluster with each 8 CPU's and 16 Linux systems 9 >> with 1 CPU each could be a nice system for some sites.! >> >> Arne  >s> >I'm thinking server farms run by ISPs (or by computer science& >departments and other training orgs). > I >What these have in common is that you want two levels of management. ThemK >organisation manages the real machine, carving it up into virtual machines,F >and determining each VM's share of the physical resource. Each VM hasG >a manager with full privileges on that VM, but no privilege elsewhere.nF >Furthermore, in general no VM trusts any other VM, and the management- >of the physical machine trusts none of them.@ >oG >VMS is fine if you have a conventional (homogeneous) structure, with aO systemJ >management and unprivileged (or partly privileged) users, and some degreeC >of mutual trust (such as that coming from all working for the samem
 employer). >gD >If one can sensibly use Galaxy software and hardware to implement aJ >number of VMs an order of magnitude greater than the number of CPUs, thenJ >I've misunderstood Galaxy, and it's far more directly equivalent to IBM's VM.m >a > Yours,
 > Nigel Arnot  > NRA@MAXWELL.PH.KCL.AC.UK >o7 > "In the beginning there was nothing, which exploded."@   ------------------------------  % Date: Thu, 14 Dec 2000 17:23:49 -0500s3 From: "paul hanrahan" <hanrahan_paul@ameritech.net>e/ Subject: Re: IBM and Linux ... purfect togethere@ Message-ID: <XLb_5.622$4Z5.63100@nntp0.chicago.il.ameritech.net>   Hi:u  E VM should be a lot less expensive these days.  Back 1991-2 the IBM VMkE development labs in Kingston, New York was closed and merged with IBM-K Endicott's programming lab.  We went from 1200 to 900 to 600 to 300 people.lJ That's about the point when I got cut from IBM's VM development staff.  At1 one point things had sunk to as few as 50 people.u  K The answer was to out source new development to places like Carnegie MelloneG and then have former developers who were still hanging on in hopes of auL pension man the Level II trouble shooting queues.  I've written thousands ofK lines of code that's ended up in VM systems.  Probably obsolete since I wasmL trashed back in 1994.  Anyway it was a cool career and a cool system to work on.v   Regards,  
 Paul Hanrahan. Former VM developerp- "Hans Vlems" <hvlems@iae.nl> wrote in message " news:91bbm8$hnk$1@news.IAEhv.nl...: > IBM's VM in no way compares to what a Galaxy implements.E > VM allows you to run several guests OS's, irrespective the hardwarewI > resources. In fact you can run VM as a guest itself. The host (primary)aD > VM translates hardware devices to the software devices used by its > various guests. C > A galaxy can run several instances of VMS but they just share thewE > hardware and each instance uses at least one, privately owned, cpu.gF > What comes close to that concept on IBM 370/390 systems is something calledI > PR/SM ("prism") which partitions the hardware. That'll allow you to run  two  > VSEsI > instances on the same machine. While (a lot) less flexible than running/ VM,-K > it is also considerably cheaper than a VM license that needs renewel each@ > year.t# > VM is nice, but rather expensive.v >e > Hans Vlems >0) > Nigel Arnot heeft geschreven in bericht>0 > <009F488C.7195157E.31@maxwell.ph.kcl.ac.uk>...! > >> Why do we need software VM ?e > >>) > >> Is Galaxy itself not a hardware VM ?w > >>E > >> A GS320 with a 2-node VMS cluster with each 8 CPU's and 16 Linuxo systems ; > >> with 1 CPU each could be a nice system for some sites.o > >>	 > >> Arned > > @ > >I'm thinking server farms run by ISPs (or by computer science( > >departments and other training orgs). > >_K > >What these have in common is that you want two levels of management. The D > >organisation manages the real machine, carving it up into virtual machinesH > >and determining each VM's share of the physical resource. Each VM hasI > >a manager with full privileges on that VM, but no privilege elsewhere.-H > >Furthermore, in general no VM trusts any other VM, and the management/ > >of the physical machine trusts none of them.e > >eI > >VMS is fine if you have a conventional (homogeneous) structure, with a. > systemL > >management and unprivileged (or partly privileged) users, and some degreeE > >of mutual trust (such as that coming from all working for the same  > employer). > >AF > >If one can sensibly use Galaxy software and hardware to implement aL > >number of VMs an order of magnitude greater than the number of CPUs, thenL > >I've misunderstood Galaxy, and it's far more directly equivalent to IBM's > VM.x > > 
 > > Yours, > > Nigel Arnots > > NRA@MAXWELL.PH.KCL.AC.UK > >e9 > > "In the beginning there was nothing, which exploded."  >  >-   ------------------------------  % Date: Thu, 14 Dec 2000 23:50:32 -0700o1 From: Vance Haemmerle <vance@toyvax.Tucson.AZ.US>  Subject: Mozilla 0.62 Message-ID: <3A395CC8.D87969B@toyvax.Tucson.AZ.US>  = Anyone had a chance to run Mozilla 0.6 on VMS yet?  I had M18 B running, but 0.6 just seems to hang before displaying any windows.   -- Vance Haemmerlet vance@toyvax.Tucson.AZ.USe   ------------------------------  # Date: Thu, 14 Dec 2000 19:39:49 GMT + From: John Santos <john.santos@verizon.net>  Subject: Re: Old VAX machine> Message-ID: <MPG.14a2ec384e4a2f3e9896a0@news.bellatlantic.net>  A In article <LM6_5.556$4Z5.49002@nntp0.chicago.il.ameritech.net>, m# hanrahan_paul@ameritech.net says...b > Hi:  > N > I recently picked up a truck load of old computers for free.  I had to leaveL > behind an old vax machine.  Huge things.  Like a stand up freezer and runs' > VMS or something.  Anyone interested?o > 
 > Regards, >t > Paul Hanrahan ! > www.geocities.com\hanrahan_pauls > (216)228-2726)   Hi, Paul  G Someone might be, and you have asked in the right place (might also tryiG comp.sys.dec), but most people would need to know two things:  1) where F is it?  (216 area code seems to be Ohio).  And what model?  Most VAXesC say in large letters on the front "VAX 11/780" or "VAX 6000-660" or"E "MicroVAX II" or some such.  This would really help someone decide if-D they are interested.  If they are, then they might ask more detailedC questions or they might be willing to take the whole thing off your  hands for one part!8  H Most of the old, big VAXes have limited value, especially for hobbyists,H because of the size, power, and cooling requirements, but sometimes theyD are very valuable for spare parts or a standby (backup) system for a? business.  Sometimes one small board inside might be worth moresB than the entire rest of the machine.  Also, sometimes the softwareC licensed to the machine is worth much more than the machine itself.l   -- o John Santoso   ------------------------------  % Date: Thu, 14 Dec 2000 17:10:22 -0500n3 From: "paul hanrahan" <hanrahan_paul@ameritech.net>T Subject: Re: Old VAX machine@ Message-ID: <nzb_5.621$4Z5.62117@nntp0.chicago.il.ameritech.net>   John:   I Thanks for replying.  The machine is air cooled but it's 210 power.  I'll K find out the model.  It was as Case Western Reserve University and used for-G a library catalog system.  Some folks I've consulted with recently havesI legacy VAX cobol systems.  I simply don't have enough space for the thinge= since I run out of a home office environment in my apartment.n  L Get back to .... ah wait a minute I have the inventory sheet from when I didA the pick up.  Desyste  5500 within H9702-AA rack.  Listed here as  operational.  Hope that helps.   Regards,   Paul8 "John Santos" <john.santos@verizon.net> wrote in message8 news:MPG.14a2ec384e4a2f3e9896a0@news.bellatlantic.net...B > In article <LM6_5.556$4Z5.49002@nntp0.chicago.il.ameritech.net>,% > hanrahan_paul@ameritech.net says...u > > Hi:s > > J > > I recently picked up a truck load of old computers for free.  I had to leavehI > > behind an old vax machine.  Huge things.  Like a stand up freezer and  runs) > > VMS or something.  Anyone interested?o > >l > > Regards, > >  > > Paul Hanrahand# > > www.geocities.com\hanrahan_paul  > > (216)228-2726b >m
 > Hi, Paul > I > Someone might be, and you have asked in the right place (might also tryMI > comp.sys.dec), but most people would need to know two things:  1) where H > is it?  (216 area code seems to be Ohio).  And what model?  Most VAXesE > say in large letters on the front "VAX 11/780" or "VAX 6000-660" orRG > "MicroVAX II" or some such.  This would really help someone decide ifcF > they are interested.  If they are, then they might ask more detailedE > questions or they might be willing to take the whole thing off yourg > hands for one part!0 > J > Most of the old, big VAXes have limited value, especially for hobbyists,J > because of the size, power, and cooling requirements, but sometimes theyF > are very valuable for spare parts or a standby (backup) system for aA > business.  Sometimes one small board inside might be worth moreDD > than the entire rest of the machine.  Also, sometimes the softwareE > licensed to the machine is worth much more than the machine itself.s >  > --
 > John Santosd   ------------------------------  % Date: Thu, 14 Dec 2000 16:19:25 -0500 ' From: "Bill Todd" <billtodd@foo.mv.com>a Subject: Re: Sun Cluster( Message-ID: <91bdc7$rau$1@pyrite.mv.net>  3 Android <Andy.Warren@Teradyne.com> wrote in messaget& news:3A38D537.A6B7FC20@Teradyne.com... >  > Bill Todd wrote: >w2 > > Philip J. Lewis <phl@bwsc.dk> wrote in message- > > news:kmnX5.50$%N2.1014@news.get2net.dk... J > > > Very sad, very true.  Suntan Bob and his willing board have a lot to
 > > answer
 > > > for. > >oG > > Compaq now has well over two years' worth of its own negligible VMSeH > > promotion to answer for, and unlike BP are still in a position to be heldH > > accountable for the disparity between their encouragement of the VMSF > > faithful and near-total neglect of VMS in any wider context.  Some peopleI > > may feel they should be thankful that Compaq didn't actively kill the- workJ > > that was already in the pipeline when they took over - but I'm not one of	 > > them.G > >u
 > > - bill > I > I got an interesting call from a Compaq marketing type yesterday...  HeT spent J > 20 minutes or so asking me about our use of OpenVMS, Alpha's, and VAXen. HeF > was interested in our purchasing plans, especially any upgrades.  He	 wanted touK > make sure I was aware of any new product announcements or special offers.- I-G > haven't had any contacts like this from Compaq (or Digital) in years.T Again,F > this survey didn't ask a single question about NT, unix, or anything	 unrelated:	 > to VMS.r  E Now, that's more like it (though it *still* doesn't indicate outreach8H outside the existing customer base):  the last time I heard a story likeJ that, the punch line was at pitch to move to Tru64 or NT (I forget which).   - bill   > J > I get the impression they're trying to get a little customer feedback to help( > guide them in their promotion efforts. >  >V > -- >3 > Andy "Android" Warreno. > (25 years using Digital Equipment Computers) >    ------------------------------  % Date: Thu, 14 Dec 2000 16:23:45 -0500 ' From: "Bill Todd" <billtodd@foo.mv.com>r Subject: Re: Sun Cluster( Message-ID: <91bdka$rih$1@pyrite.mv.net>  0 Alan Greig <agreig@my-deja.com> wrote in message2 news:46oh3t4jvf9jmapacs8st676sbuhpaud9d@4ax.com...% > On Tue, 05 Dec 2000 07:57:16 -0300,s, > fabio_compaq@ep-bc.petrobras.com.br wrote: > L > >Theres a news in CNET about the new Sun Cluster, comparisons with HP and > >IBM, but no OpenVMS > >4 > >uK > >http://news.cnet.com/news/0-1003-200-3995209.html?tag=st.ne.1002.thed.nib > >n
 > >Regards > >t > >FCt >sB > This week's Computing (UK) has an article on  page 26 "Sun playsF > catch-up on server clusters". It quotes Martin Hingley of IDC saying/ > "DEC VAX clusters did it 10 or 15 years ago."r >eB > Although nice to see a mention again the implication is that VMSH > clusters are no longer around.. It should really say "DEC (now Compaq)G > VMS clusters have being doing it for 10 or 15 years and they're stillh > ahead of the pack."h  L Reminds me of another reference I read somewhere in the last couple of days,L saying that Sun was starting to do now what S/390 (Parallel Sysplex) and VMSK clusters (mentioned in that order) had been doing for 'several years' (truetI for the former, but hardly adequate homage to VMS's 17-year leadership in.
 the area).   - bill   > -- > Alan Greig   ------------------------------  % Date: Thu, 14 Dec 2000 16:37:36 -0500 ' From: "Bill Todd" <billtodd@foo.mv.com>e Subject: Re: Sun Cluster( Message-ID: <91bee9$s3p$1@pyrite.mv.net>  D Larry Kilgallen <Kilgallen@eisner.decus.org.nospam> wrote in message% news:2THcjDCOEuvz@eisner.decus.org...0L > In article <91adtv$5e9$1@pyrite.mv.net>, "Bill Todd" <billtodd@foo.mv.com> writes:h > >GH > > Larry Kilgallen <Kilgallen@eisner.decus.org.nospam> wrote in message) > > news:ydjYT94dZis2@eisner.decus.org...v9 > >> In article <916lc2$2cm$1@pyrite.mv.net>, "Bill Todd"s <billtodd@foo.mv.com>. > > writes:j > >> >- > >> > <kparris@my-deja.com> wrote in messagee' > > news:915j4t$jcc$1@nnrp1.deja.com...l > >>L > >> >> Richie Lary certainly comes to mind here; he has been very prominent inH > >> >> the storage field in general and most recently in SANs.  He justG > >> >> received a Lifetime Achievement Award from the I/O Technologieso Forum E > >> >> (see http://www.sresearch.com/itf/itf_2000/award/awards.htm).t > >> >H > >> > Since I was saddened that prior commitments made it impossible to attendF > >> > Richie's going-away party a few months ago, I'm not sure he's a suitableJ > >> > candidate to put forward any more - unless you also want to include Dave& > >> > Cutler, Gordon Bell, KO, et al. > >>G > >> "A few months ago" would have been before his recent LA DECUS talk  given K > >> under Compaq auspices.  He may be somewhat independent now, but it waseE > >> clear to me there was still a connection between Richie Lary anda Compaq.  > >P > > And your point is? >cL > You must not realize how hostile that turn of phrase is.  Or maybe you do.  I 'Fraid I occasionally get fed up with people (you just happened to be theaF most recent) who throw out only tenuously related information as if itK refuted a statement they aren't comfortable accepting.  Far too much of theiK discussion that goes on here reminds me of a bunch of lambs (to appropriate H a metaphor from another thread) vigorously disputing any possible futureK relationship to the veal cutlets their shepards enjoy, despite all evidencel suggesting otherwise.    >dK > >                     There's still a connection between *me* and DECpaq,k evenL > > after almost 14 years' separation, but I'm no longer a DECpaq engineer - andl > > neither is Richie. >hA > I did not see you presenting on Compaq's behalf at the LA DECUSh Conference.i( > Of course I might have just missed it.  G Then again, I'd been gone for considerably more than a few weeks at the H time, and didn't have any commitments made prior to my departure left to fulfill.  E The original comment was made after Keith Parris offered Richie as an H example of a person still at Compaq (in response to assertions that goodJ people were still being lost) whom non-Compaq people would flock to listenH to at some industry event.  My point, which remains as valid as ever, isL that Richie no longer fits that definition, though informal and perhaps evenL some formal ties with the company will doubtless persist for some time (it's  still hard to really leave DEC).   - bill   ------------------------------  # Date: Thu, 14 Dec 2000 21:58:34 GMTd= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)c Subject: Re: Sun Cluster0 Message-ID: <009F4977.8EA9AB9B@SendSpamHere.ORG>  R In article <91bee9$s3p$1@pyrite.mv.net>, "Bill Todd" <billtodd@foo.mv.com> writes: {...snip...}L >refuted a statement they aren't comfortable accepting.  Far too much of theL >discussion that goes on here reminds me of a bunch of lambs (to appropriateI >a metaphor from another thread) vigorously disputing any possible futurehL >relationship to the veal cutlets their shepards enjoy, despite all evidence >suggesting otherwise.  ; With one exception Bill, veal is not derived from lamb.  :)c     --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM             sO city, n., 1. a place where trees are cut down and streets are named after them.n   ------------------------------    Date: 15 Dec 2000 01:06:11 +0800, From: Paul Repacholi <prep@prep.synonet.com> Subject: Re: Sun Cluster0 Message-ID: <87d7evdjb0.fsf@k9.prep.synonet.com>  ( Rudolf Wingert <win@fom.fgan.de> writes:  I > Last year did we hear about a 1.6Ghz Alpha. But we do see nothing about H > a real product till today. When can we see Alpha produced in a 13micro. > technology? I do see nothing on the horizon.  < If you go to The Register, or Slashdot you will quickly find8 the article on the IBM 0.13u process. THEY are only just; transfering it from the lab, and will be fabbing the Power4m3 with it. Guess what else is due to go on that line.a  < **EVERYONE** is setting new Speclate95 numbers. Sparc, HPPA,= Alpha, IBM, and of course the Unobtainium. Sparcs where what,a9 about 15 months late? Near all of it due to fab hold-ups. 9 The 264 delay was I suspect, figuring out how to test thev" sucker well enought to let it out.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.-@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------  # Date: Thu, 14 Dec 2000 23:44:01 GMT14 From: "Terry C. Shannon" <terryshannon@mediaone.net> Subject: Re: Sun Cluster> Message-ID: <5Xc_5.108740$751.2638961@typhoon.ne.mediaone.net>  9 "Paul Repacholi" <prep@prep.synonet.com> wrote in messager* news:87d7evdjb0.fsf@k9.prep.synonet.com...* > Rudolf Wingert <win@fom.fgan.de> writes: > K > > Last year did we hear about a 1.6Ghz Alpha. But we do see nothing about J > > a real product till today. When can we see Alpha produced in a 13micro0 > > technology? I do see nothing on the horizon.  G EV69 in 0.13 by 2002 according to info gathered yesterday. EV68 in 0.18o/ exists today; I have one of these 1GHz wonders.    ------------------------------   Date: 15 Dec 2000 00:28:54 GMT2 From: mathog@seqaxp.bio.caltech.edu (David Mathog) Subject: Re: Sun Cluster, Message-ID: <91bok6$7uk@gap.cco.caltech.edu>  u In article <5Xc_5.108740$751.2638961@typhoon.ne.mediaone.net>, "Terry C. Shannon" <terryshannon@mediaone.net> writes:o >>L >> > Last year did we hear about a 1.6Ghz Alpha. But we do see nothing aboutK >> > a real product till today. When can we see Alpha produced in a 13microe1 >> > technology? I do see nothing on the horizon.d >eH >EV69 in 0.13 by 2002 according to info gathered yesterday. EV68 in 0.180 >exists today; I have one of these 1GHz wonders. >r  G Is this marvel in a GS class system or in something a bit more compact?u  M With Alphas the high end chips go into the big servers and take quite a whiledJ to percolate down into the realm where regular folks can afford them.  NotF that Alphas ever become really affordable, but at least a stripped 4K$H linux/alpha is occasionally within the budget.  For about 10 years AlphaG has almost always had fastest chip bragging rights, but an awful lot ofeI customers at every time point were stuck buying _new_ machines with chips9B as much as 2 years behind state of the art, which is not nearly soH comfortably ahead of AMD and Intel.   This is very different than in theK AMD/Intel world where pretty much anybody who wants one can buy the fastest L chip for a premium that doesn't totally break the bank.  (Or in other words,E I don't care that the Alpha is fastest if I can't afford to buy that  	 machine!)n  K When will we see this 1Ghz chip in a DS10 sized system?  If history is any nE guide, 2002 would be the earliest we should look for it.  Especially i1 because they only just put out the 600 Mhz model.t   Regards,   David Mathog mathog@seqaxp.bio.caltech.eduh> Manager, sequence analysis facility, biology division, Caltech   ------------------------------  % Date: Thu, 14 Dec 2000 22:07:41 -0500h' From: "Bill Todd" <billtodd@foo.mv.com>t Subject: Re: Sun Cluster( Message-ID: <91c1p6$bk2$1@pyrite.mv.net>  H Brian Schenkenberger, VAXman- <system@SendSpamHere.ORG> wrote in message* news:009F4977.8EA9AB9B@SendSpamHere.ORG...L > In article <91bee9$s3p$1@pyrite.mv.net>, "Bill Todd" <billtodd@foo.mv.com> writes:. > {...snip...}J > >refuted a statement they aren't comfortable accepting.  Far too much of theaB > >discussion that goes on here reminds me of a bunch of lambs (to appropriate K > >a metaphor from another thread) vigorously disputing any possible futurerE > >relationship to the veal cutlets their shepards enjoy, despite alle evidence > >suggesting otherwise. >e= > With one exception Bill, veal is not derived from lamb.  :)   H Clearly when my patience disappears, so also do my spelling and culinary insight.   - bill   >e >g > --4 > VAXman- OpenVMS APE certification number: AAA-0001 VAXman(at)TMESIS(dot)COM > K > city, n., 1. a place where trees are cut down and streets are named aftern them.w   ------------------------------  # Date: Fri, 15 Dec 2000 04:33:45 GMT-4 From: "Terry C. Shannon" <terryshannon@mediaone.net> Subject: Re: Sun Cluster; Message-ID: <Jah_5.5358$1M.1139918@typhoon.ne.mediaone.net>D  ? "David Mathog" <mathog@seqaxp.bio.caltech.edu> wrote in message<& news:91bok6$7uk@gap.cco.caltech.edu...J > In article <5Xc_5.108740$751.2638961@typhoon.ne.mediaone.net>, "Terry C., Shannon" <terryshannon@mediaone.net> writes: > >>H > >> > Last year did we hear about a 1.6Ghz Alpha. But we do see nothing abouteE > >> > a real product till today. When can we see Alpha produced in a  13microl3 > >> > technology? I do see nothing on the horizon.6 > >2J > >EV69 in 0.13 by 2002 according to info gathered yesterday. EV68 in 0.182 > >exists today; I have one of these 1GHz wonders. > >e >oI > Is this marvel in a GS class system or in something a bit more compact?d  F All I had was the chip itself, odds are real good it'll show up in theH GS-Series in 1Q or 2Q of next year, then in the ES45 by the end of 2001.  : As for MARVEL, well, it's a GS-followon with the EV7 chip.   >gI > With Alphas the high end chips go into the big servers and take quite at whilefL > to percolate down into the realm where regular folks can afford them.  NotH > that Alphas ever become really affordable, but at least a stripped 4K$J > linux/alpha is occasionally within the budget.  For about 10 years AlphaI > has almost always had fastest chip bragging rights, but an awful lot ofwK > customers at every time point were stuck buying _new_ machines with chipseD > as much as 2 years behind state of the art, which is not nearly soJ > comfortably ahead of AMD and Intel.   This is very different than in theE > AMD/Intel world where pretty much anybody who wants one can buy then fastestjG > chip for a premium that doesn't totally break the bank.  (Or in otheri words,F > I don't care that the Alpha is fastest if I can't afford to buy that > machine!)w  I A point well taken. Guess DECpaq is playing a balancing act. Heck, if youdL could buy an 8-pack of 1GHz Slates, Myrinet the puppies together, etc, would# you pay much, much more for a GS80?s  I I'm not saying that DECpaq couldn't execute the balancing act better (I'deJ argue for faster, better, sooner) but it's easy for me to tell those folksD what to do with their money and their chips. (Of course, using 20-20L hindsight, a larger investment in Alpha--to hurry things up--might have been1 money better spent than the CMGI investment, no?)t       >aL > When will we see this 1Ghz chip in a DS10 sized system?  If history is anyF > guide, 2002 would be the earliest we should look for it.  Especially3 > because they only just put out the 600 Mhz model.c  H I believe the real speed is 616MHz, have seen numbers and claims to that  effect. But hey, why quibble....  H DS10E is slated for 667MHz with 8MB cache, followed by 750MHz. UncertainJ about faster followons. May require a system redesign (a la ES45 Privateer# vs ES40) for the really fast stuff.    ------------------------------  # Date: Fri, 15 Dec 2000 01:52:42 GMT ' From: moi_is_me <moi_is_me@my-deja.com>s! Subject: TCP Bug - MULTINET + UCXn) Message-ID: <91bth8$rv0$1@nnrp1.deja.com>e  C Found what i consider a serious bug. It seems to be present in both C UCX (TCP/IP Services) 5.0 and Multinet 4.2 on a ALPHA running 7.2-1    Try the following...   Server4 1. Prepare a TCPIP Server, and Accept the connection 2. DONT read any data - yetl     Client 3.   Connect to the Server* 4.   Determine the Socket Read Quota - XXX= 5.   Send XXX bytes to the Server (closes the receive window)a 6.   Queue a timer - 1 sec 7.   Send 1000 Bytes to Server* 8.   When QIO Write completes cancel timer? 9.   Keep running total of number of bytes written (accepted byt      UCX/Multinet)0 10.  If timer routine goes off, cancel QIO write/ 11.  Repeat steps 6 thru 9 until step 10 occursl   Server 11   Queue a timer - 1 sec 12.  Issue Reads) 13.  When QIO Read completes cancel timer / 14.  Keep running total of number of bytes read-/ 15.  If timer routine goes off, cancel QIO read 1 16.  Repeat steps 11 thru 14 until step 15 occurs:    D Compare running totals in the Server and Client - More bytes arrivedC at Server than what the Client is aware of,  ie when Cancelling thedB Write within the Client, neither UCX or Multinet let you know thatD some bytes were actually transmitted, the iosb.nbytes always says 0.  A I have also tried the above with TcpWare - and have not found any 	 problems.   G I don't have a support contract, so can not report the problem directlyi. to ComPaq or Process ... anyone else care to ?   -Pierrel     Sent via Deja.como http://www.deja.com/   ------------------------------   Date: 14 Dec 2000 21:41:20 GMT# From: casinoop2@aol.com (CasinoOp2)c6 Subject: VAX - ALPHA Fortran common block differences?: Message-ID: <20001214164120.18072.00010276@ng-fi1.aol.com>  3 I need help with what should be a simple process...h  B I am trying to move a program from a VAX to an ALPAH machine (bothE running VMS v7.2). This program used a shared, memory resident commonHC block. It works fine on the VAX, but not on the ALPHA. A short testpC procedure shown below illustrates the idea. There is a common blockiE called Point that contains 15 integers. This short program will read,pG set, or just display the current contents. On the VAX, one user runninglG this program can set the values and another user will see what has beentD set. On the ALPHA, each user seems to see his own copy of this data.' What am I doing (or not doing) wrong???i      Clark Calkins, programmer    Schafer Corporation    ccalkins@schafercorp.comt    G = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =o@ c	Program to test VMS resident common block access from Fortran. ce@ c	Compile and link THIS file by (note default directory I use isE c	dia0:[ana.bank]). Note that is must be linked AFTER Point is linked 
 c	below... c	 $ fortran testc.for c	 $ link testc,testcb/options c-' c	and "Testcb.opt" contains one line...:! c	 dia0:[ana.bank]Point/shareablel cd) c	and "Point.for" contains three lines...e c	 integer *4 Pointers(15) c	 common /Point/Pointersg c	 end ca' c	and this is compiled and linked by...m c	 $ fortran point.for c	 $ link Point.obj/share2 c H c	then this is installed in memory with the following commands (requires c	CMKRNL privilege). c	 $ Install:==$Installc7 c	 $ Install dia0:[ana.bank]Point.exe/shared/open/writeu c A c	Lastly, a logical is defined (without this VMS wants to look inh! c	sys$library for some reason)...F* c	 $ define Point dia0:[ana.bank]point.exe ci 	program TestC 	integer*4 i,j c- 	integer*4 Pointers(15)t 	common /Point/Pointers  c  	j=1303 / 100	type *,'Current contents of /Point/ are...'e
 	do i=1,15 	 type 1000, i, Pointers(i)t 1000	format(i5,i15)t 	enddo cs/ 	type *,'Enter command (0=end, 1=set, 2=show):'" 	accept *,ia 	if (i .eq. 0) goto 200e 	if (i .eq. 2) goto 100 
 	do i=1,15 	 Pointers(i)=ran(j)*100.e 	enddo	 	goto 100  200	type *,'Ending'c 	endG = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =    ------------------------------   Date: 14 Dec 2000 23:27:46 GMT* From: helbig@astro.rug.nl (Phillip Helbig): Subject: Re: VAX - ALPHA Fortran common block differences?. Message-ID: <91bl1i$35c$1@info.service.rug.nl>  : In article <20001214164120.18072.00010276@ng-fi1.aol.com>,& casinoop2@aol.com (CasinoOp2) writes:   5 > I need help with what should be a simple process...C > D > I am trying to move a program from a VAX to an ALPAH machine (bothG > running VMS v7.2). This program used a shared, memory resident commondE > block. It works fine on the VAX, but not on the ALPHA. A short testiE > procedure shown below illustrates the idea. There is a common blockeG > called Point that contains 15 integers. This short program will read,oI > set, or just display the current contents. On the VAX, one user runningdI > this program can set the values and another user will see what has been F > set. On the ALPHA, each user seems to see his own copy of this data.) > What am I doing (or not doing) wrong???b  : I'm not sure, but could it be a 4-byte/8-byte thing?  See       $ HELP FORTRAN/ALIGNH   ------------------------------   Date: 14 Dec 2000 23:57:04 GMT# From: casinoop2@aol.com (CasinoOp2)t: Subject: Re: VAX - ALPHA Fortran common block differences?: Message-ID: <20001214185704.09486.00005483@ng-cg1.aol.com>  ; >I'm not sure, but could it be a 4-byte/8-byte thing?  See w >n >   $ HELP FORTRAN/ALIGN >   M I have tried ALIGN=(COMMON=STANDARD) and ALIGN=(COMMON=MULTILANGUAGE) with nou+ appearent differences. Still does not work.a  L When I do a "install/list/full point" I get the following info on the ALPHA:  , POINT;4       Open    Shared    Lnkbl    Wrt+        Entry access count               = 3>+        Current / Maximum shared     = 2 / 2w,        Global section count              = 1    J The VAX is the same EXCEPT the Global section count shows "2". A 32/64 bit thing?  $   Clark Calkins, (junior) programmer   Schafer Corp   ccalkins@schafercorp.com   ------------------------------  % Date: Thu, 14 Dec 2000 20:04:25 -0500   From: John Santos <JOHN@egh.com>: Subject: Re: VAX - ALPHA Fortran common block differences?4 Message-ID: <1001214200039.342A-100000@Ives.egh.com>    On 14 Dec 2000, CasinoOp2 wrote:  = > >I'm not sure, but could it be a 4-byte/8-byte thing?  See   > >c > >   $ HELP FORTRAN/ALIGN > >i > O > I have tried ALIGN=(COMMON=STANDARD) and ALIGN=(COMMON=MULTILANGUAGE) with no - > appearent differences. Still does not work.  > N > When I do a "install/list/full point" I get the following info on the ALPHA: > . > POINT;4       Open    Shared    Lnkbl    Wrt- >        Entry access count               = 3r- >        Current / Maximum shared     = 2 / 2.. >        Global section count              = 1 >  > L > The VAX is the same EXCEPT the Global section count shows "2". A 32/64 bit > thing?  H The default .PSECT attributes on COMMONS are different on VAX vs. ALPHA.  $ VAX defaults to SHR, Alpha to NOSHR.  G I think you need a PSECT_ATTRIBUTES statement in your link options fileo
 like this:  m= psect_attr=foo,   long,pic,ovr,rel,gbl,shr,noexe,rd,wrt,novecy  % Where foo is the name of your COMMON.s   -- R John Santose Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------  % Date: Fri, 15 Dec 2000 12:08:29 +1100e/ From: "Phil Howell" <phowell@snowyhydro.com.au>l: Subject: Re: VAX - ALPHA Fortran common block differences?1 Message-ID: <c8e_5.1719$%v1.55623@ozemail.com.au>n  . CasinoOp2 <casinoop2@aol.com> wrote in message4 news:20001214164120.18072.00010276@ng-fi1.aol.com...5 > I need help with what should be a simple process...< >lD > I am trying to move a program from a VAX to an ALPAH machine (bothG > running VMS v7.2). This program used a shared, memory resident commonhE > block. It works fine on the VAX, but not on the ALPHA. A short testoE > procedure shown below illustrates the idea. There is a common blockoG > called Point that contains 15 integers. This short program will read, I > set, or just display the current contents. On the VAX, one user running I > this program can set the values and another user will see what has beenEF > set. On the ALPHA, each user seems to see his own copy of this data.) > What am I doing (or not doing) wrong???- <snip>4 I think the default for common blocks on VAX was SHR the default on alpha is NOSHReE have a look at http://www5.compaq.com/fortran/docs/vms-um/dfum029.htm99 and search for references to symbol_vector and psect_attre. You may also want to ask in comp.lang.fortran. Phil   ------------------------------  % Date: Thu, 14 Dec 2000 14:24:50 -0500p# From: John Vottero <John@MVPSI.com>y- Subject: RE: vax/vms system as a DHCP client? D Message-ID: <C15945A9D9EFCF11BA8B08002BBF1CCC0CD920@berry.mvpsi.com>  K It depends upon which TCP/IP package you're running.  If it's Compaq TCP/IPeK Services (formerly known as UCX) then you have to wait for V5.1 which is in  field test now.e   > -----Original Message-----  > From: Dan [mailto:dan@vrx.net], > Sent: Thursday, December 14, 2000 12:47 PM > To: Info-VAX@Mvb.Saic.Com + > Subject: vax/vms system as a DHCP client?  >  > 2 > How do I setup my vax (vms) to be a DHCP client?/ > I'd like it to get an IP address dynamically.  >  > Dan. >    ------------------------------  % Date: Thu, 14 Dec 2000 21:01:16 -0500-" From: Dan Sugalski <dan@sidhe.org>) Subject: Re: Virtual I/O Cache and Oracle : Message-ID: <5.0.1.4.0.20001214210025.0214a4b0@24.8.96.48>  - At 10:45 AM 12/14/00 +0000, Alan Greig wrote: G >On Wed, 13 Dec 2000 23:05:51 -0800, Koloth <koloth@tmisnet.com> wrote:u >rO > >Most databases have their own caching systems.  Enabling VIOC will only slowl > >things down because you areK > >moving data from one cache to another on the same system.  You would be i > better2 > >off increasing the size of the cache in Oracle. > >z >pE >Turning on VIOC is hiighly unlikely to slow things down. Even thoughaC >you might be moving data from one cache to another the alternativegC >(assuming only difference is VIOC on or off) would be to move datae@ >from disk to Oracle SGA. And it's far easier for most people toA >understand and turn on VIOC than it is to muck about with Oracle:D >configuration. Plus turning on VIOC will improve performance of theA >entire system - not just Oracle. Given sufficient memory and cpua >resources.. >7C >I would turn it on and give it 200MB of memory for starters if younD >have sufficient. See how it goes. You might suddenly find disk read0 >hit rates of 90% depending on the applications.  K This is generally a bad idea with Oracle. The DB accesses don't go through lC the VIOC, so it's just memory wasted as far as Oracle is concerned.o   					Dan  I --------------------------------------"it's like this"-------------------o2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and eveni;                                       teddy bears get drunka   ------------------------------    Date: 14 Dec 2000 15:32:42 -0500, From: koehler@eisner.decus.org (Bob Koehler) Subject: Re: VMS "froze"+ Message-ID: <nnIJTNLKoYh3@eisner.decus.org>o  d In article <918a7l$83lm$1@lead.zk3.dec.com>, "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> writes: > F > In the future, we are addressing this on larger systems by having anC > external microprocessor that handles system management functions.a  B Like the LSI-11 on the front of my 11/780's?  It couldn't halt theH 11/780 in an IPL 31 loop even though the console TR was highest priorityC on the SBI (can't slip signals into the CPU if it isn't listening).   F ----------------------------------------------------------------------? Bob Koehler                     | Computer Sciences Corporatione= NASA GSFC Flight Software       | Federal Sector, Civil GroupsE                                 | please remove ".aspm" when replyingt   ------------------------------    Date: 15 Dec 2000 01:09:14 +0800, From: Paul Repacholi <prep@prep.synonet.com> Subject: Re: VMS "froze"0 Message-ID: <87ae9zdj5x.fsf@k9.prep.synonet.com>  ) carl@gerg.tamu.edu (Carl Perkins) writes:   I > Even the DS10 has a little processor on one of the serial ports that isyK > smart enough to handle things including dial-up connections even when the H > main processor is not functioning, allowing you to check the status ofD > various things and/or boot the system if necessary (and possible).   Do you mean the RSM?  C The Cobras had the consoles  and display on a uP board. Now if theya will just add a LAT port...h   -- i< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.r@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.   ------------------------------   Date: 14 Dec 2000 19:09:21 GMT2 From: mathog@seqaxp.bio.caltech.edu (David Mathog) Subject: Re: Weird I/O problem, Message-ID: <91b5t1$peg@gap.cco.caltech.edu>  p In article <2q8_5.765$yz3.27601@iguano.antw.online.be>, "Steven Driesen" <NOSPAMsteven.driesen@tijd.com> writes: >eH >This is what sometimes happens for a couple of runs and then spontanous >disappears:2 >- the open of the file is allways ok (IOSTAT = 0)M >- a read from the same file fails with IOSTAT=29 (FILE NOT FOUND !!!). Theret >is no close statement between> >  the open and the read. I checked, and the file does exist!!  I That's pretty strange unless maybe that's what Fortran does when you readtJ past the end of the file, and that's what you've done.  More likely thoughG that there's a bug in your program that's occasionally being triggered.sI When the program gets into that mode it goes nuts, and it just happens to . give this message as it spins out of control.    > ? >Another thing that also sometimes happens and then disappears:  >- open of file is okdM >- write to file fails with IOSTAT = 66 (RECORD OVERFLOW). I checked the data / >that is being written to the file and it looksh >  OK for the format statement.   C Any chance something else is creating a file with the same name andrI conflicting RMS values?  You might get a message such as the above if therD fortran file requires "large" records (as are set by default in yourF program) and something else has created a file with the same name withC smaller records.  Fortran would pick up the RMS parameters from thehK existing file (because you did not specify values in your program) and thendI your program could fail on the write much as you described.  If we could AI see the snippet of code involved it might be more evident what was going t wrong.  J Anyway, analyze the RMS values on the file in question and make sure that : they are what you (and your program) think they should be.  
 Good luck,   David Mathog mathog@seqaxp.bio.caltech.edud? Manager, sequence analysis facility, biology division, Caltech     ------------------------------   End of INFO-VAX 2000.698 ************************