0 INFO-VAX	Sat, 10 Jan 2004	Volume 2004 : Issue 18      Contents: Re: 7.3-1 to 7.3-2 Re: 7.3-1 to 7.3-2 Re: Anonymous Read-Only ! Re: Could OpenVMS-Athlon ever be? ! Re: Could OpenVMS-Athlon ever be? ! Re: Could OpenVMS-Athlon ever be? ! Re: Could OpenVMS-Athlon ever be?  Re: CSWS V2 and ODS5 Re: ds10l blowout  Re: ds10l blowout ( Experienced OpenVMS Admin/Guru available Re: GNU build tools ??P Re: Hewlett-Packard plans to EOL the HP AlphaServer Tru64. Find out how Sun can * Re: Hobbyist Layered Product license PAKs?* Re: Hobbyist Layered Product license PAKs?' Re: how to flush a file opened by DCL ? ' Re: how to flush a file opened by DCL ?   Re: HP TCPIP SMTP Queues (again)  Re: HP TCPIP SMTP Queues (again)  Re: HP TCPIP SMTP Queues (again)  Re: HP TCPIP SMTP Queues (again)  Re: HP TCPIP SMTP Queues (again)% Re: I just scored an AlphaServer 800! 2 Re: Kerberos installation fails with 7.3-2 upgrade2 Re: Kerberos installation fails with 7.3-2 upgrade2 Re: Kerberos installation fails with 7.3-2 upgrade2 Re: Kerberos installation fails with 7.3-2 upgrade Re: Multnet Documentation V4.0? # MX - create outbound mail with FQDN  Re: MySQL 4.0.17 for OpenVMS3 Re: Non-WS Process Quotas not in Performance Manual 3 Re: Non-WS Process Quotas not in Performance Manual 3 Re: Non-WS Process Quotas not in Performance Manual  Re: OpenVMS CertificationsP Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per minute usP Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per minute us  Re: Rally VAX to Alpha Migration  Re: Rally VAX to Alpha Migration8 Re: Secure Web Server Version 2.0 available for download8 Re: Secure Web Server Version 2.0 available for download' Silly Users with Password as "Password" + Re: Silly Users with Password as "Password" + Re: Silly Users with Password as "Password" + Re: Silly Users with Password as "Password" + Re: Silly Users with Password as "Password" + Re: Silly Users with Password as "Password" + Re: Silly Users with Password as "Password" / SKHPC: First Batch of 2003 SKHPC PDFs Available 3 Re: SKHPC: First Batch of 2003 SKHPC PDFs Available 3 Re: SKHPC: First Batch of 2003 SKHPC PDFs Available 3 Re: SKHPC: First Batch of 2003 SKHPC PDFs Available # Re: To all "patriotic" americans... 2 Ultra 3 drives in ES40 systems with EV6 processors VAXstation question  Re: VAXstation question  Re: VAXstation question  Re: Vaxstation some question... , Re: VMS naming conventions for disk volumes?, Re: VMS naming conventions for disk volumes?, Re: VMS naming conventions for disk volumes?, Re: VMS naming conventions for disk volumes?, Re: VMS naming conventions for disk volumes?, Re: VMS naming conventions for disk volumes?, Re: VMS naming conventions for disk volumes?, Re: VMS naming conventions for disk volumes?, Re: VMS naming conventions for disk volumes?$ Re: VMS to VMS copying between Nodes$ Re: VMS to VMS copying between NodesD Re: Well Andrew, "3" count them "3" security patches for VMS in fiveD Re: Well Andrew, "3" count them "3" security patches for VMS in five7 Re: Where to get layered product software for hobbyist? 7 Re: Where to get layered product software for hobbyist?   F ----------------------------------------------------------------------   Date: 9 Jan 2004 11:56:40 -0800 0 From: keith.cayemberg@conti.de (Keith Cayemberg) Subject: Re: 7.3-1 to 7.3-2 < Message-ID: <3a65a5c8.0401091156.467548e@posting.google.com>   You can find some info about    :  HP OpenVMS Alpha Version 7.3-2 new features and benefits;  HP OpenVMS Alpha Version 7.3-2 frequently asked questions    here...   6 http://h71000.www7.hp.com/openvms/operatingsystem.html   ------------------------------  % Date: Fri, 09 Jan 2004 19:11:08 -0600 @ From: "David J. Dachtera" <djesys.nospam@NeOaSrPtAhMlNiOnWk.net> Subject: Re: 7.3-1 to 7.3-2 6 Message-ID: <3FFF512C.7E13F3FC@NeOaSrPtAhMlNiOnWk.net>   Bhushan Narkhede wrote:  > 
 > Hi Guys, > D >   I am looking at the efforts that might be involved in moving our2 > system applications from OpenVMS 7.3-1 to 7.3-2. > / > can anyone guide on how to proceed on this???    Boot up and run?  E Seriously, if it runs on V7.3-1, the -2 upgreade probably won't break E it. If you're in an industry that demands validation, be sure to do a 
 thorough job.   H > What all changes have gone in from 7.3-1 to 7.3-2 ??? or point to some" > kind of documentation on it ????  1 Look for the New Features or release Notes doc.'s   G > I am trying to get the release notes for 7.3-2, but looks like the HP H > site is undergoing some kind of work or due to some reason not able to  > get any thing from their site.   Try again...   --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Sat, 10 Jan 2004 05:13:48 +0100 2 From: martin@radiogaga.harz.de (Martin Vorlaender)  Subject: Re: Anonymous Read-Only; Message-ID: <3fff7bfc.524144494f47414741@radiogaga.harz.de>   7 Michael D. Ober (obermd-.@.-alum-mit-edu-nospam) wrote: D > Does anyone know how to get NFS (GuardianOS Linux kernal) to allowJ > unlimited read-only access to an NFS share.  All the access will be fromF > one client (VMS 7.3) , but the user accounts are highly variable, so7 > keeping a user list on the NFS server isn't feasible.   H If a NFS request comes in with a UID/GID that is not in the proxy table,G it will default to UID/GID = (-2,-2). Set up the proxy table containing H that UID/GID and give the mapped VMS user read-only access to the share.   cu,    Martin --  J One OS to rule them all       | Martin Vorlaender  |  VMS & WNT programmer7 One OS to find them           | work: mv@pdv-systeme.de J One OS to bring them all      |   http://www.pdv-systeme.de/users/martinv/> And in the Darkness bind them.| home: martin@radiogaga.harz.de   ------------------------------  % Date: Fri, 09 Jan 2004 15:18:14 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>* Subject: Re: Could OpenVMS-Athlon ever be?) Message-ID: <3FFF0C85.99FDFBFB@istop.com>   ! re: lack of registers in the 8086   N I can understand that in the early days of the 8086, there was no way to put aM serious OS on that chip. But when Intel transformed the toy into a 32 bit CPU J which no longer required those damned segment registers etc, didn't it add more registers ?  L And when AMD made the 64 bit version of the 8086, if it was eying serious OSA such as Solaris, Linux and possibly some proprietary ones such as K VMS/HP-UX/AIX, wouldn't they have put in sufficient numbers of registers to  make such serious OS possible ?    ------------------------------  $ Date: Fri, 9 Jan 2004 13:01:40 -0800* From: "Jack Peacock" <peacock@simconv.com>* Subject: Re: Could OpenVMS-Athlon ever be?2 Message-ID: <hoWdnYv_QYUoi2KiRVn-gg@mpowercom.net>  7 "JF Mezei" <jfmezei.spamnot@istop.com> wrote in message # news:3FFF0C85.99FDFBFB@istop.com... # > re: lack of registers in the 8086  > J > I can understand that in the early days of the 8086, there was no way to put a K > serious OS on that chip. But when Intel transformed the toy into a 32 bit  CPU L > which no longer required those damned segment registers etc, didn't it add > more registers ? > I The 80386 extended the registers to 32 bits but left the same number: AX, I BX, CX, DX GP semi-genreal purpose regs, SI and DI indexes, BP frame base L pointer, and the SP stack pointer.  Two additional seg registers were added.H The problem is that the instruction set doesn't have room for additionalJ register address bits.  Intel would have had to bolt on a whole new 32-bitJ instruction set to take advantage of additional general purpose registers.F Been a while since I looked at the AMD-64 instruction extensions but IK believe AMD did add in more registers that could only be addressed with new 
 instructions.   L Segment registers are still needed with 16 bit address instructions, even if0 runing the 386 in linear 32 bit addressing mode.  L I can't see how Intel would preserve upward compatibility if the instructionH subfields for addressing were enlarged.  Zilog handles this problem withI either a mode bit or a prefix byte when an instruction has to switch from : 16-bit to 24-bit addressing in the latest Z80 incarnation.   Jack Peacock   ------------------------------   Date: 9 Jan 2004 18:10:11 -0600 - From: Kilgallen@SpamCop.net (Larry Kilgallen) * Subject: Re: Could OpenVMS-Athlon ever be?3 Message-ID: <bZ$g0+3BOQou@eisner.encompasserve.org>   o In article <kbCLb.12020$Rt4.6069@news.cpqcorp.net>, "Fred Kleinsorge" <my-last-name@stardotzko.dec.com> writes:  > 1 > "L. Blunt" <someone@out.there> wrote in message ' > news:btmm5c02nti@news1.randori.com... K >> Ah, that's the thing, I thought Macro32 was the VAX assembler, which was B >> just compiled on Alphas. Not that it had been extended for that > architecture3 >> as well (that's what I thought Macro64 was for).  >> > J > The Alpha and IA64 assemblers are there for things you just can't do anyM > other way.  But there is a large amount of existing code that originated on K > VAX and is Macro-32.  *Very* often it was the case that it was simpler to L > make Alpha-specific changes to the Macro-32 code, than to rewrite the code > in, say C.  E Rewriting something from one language to another is always hazardous. H Up through VMS V7.3, for instance, the MAIL/OLD command was still there.A I rewrote some heavily used software from one language to another G because the existing implementation could not support new requirements. D The specification for the programming task was about 500 pages long,@ and it was very difficult to get people to thoroughly review it.   ------------------------------  % Date: Fri, 09 Jan 2004 23:35:58 -0600 / From: Chris Scheers <chris@applied-synergy.com> * Subject: Re: Could OpenVMS-Athlon ever be?3 Message-ID: <3FFF8F3E.5350AD14@applied-synergy.com>    John Reagan wrote: > H > - Porting Macro-32 to 8086 would be very difficult.  The biggest issueI > the lack of sufficient registers on 8086.  Think of Macro32 source code H > has having several 'predeclared variables' named R0, R1, etc. that areD > global to all routines.  On Alpha and Itanium, there are sufficentJ > hardware registers for all of that to work although the Calling StandardJ > on OpenVMS I64 makes it a little interesting.  On 8086, I really have noH > clue how to efficiently generate code for Macro32 routines that pass 8C > 'registers' between each other (or between Macro32 and BLISS or C I > routines) for instance.  Macro32's use of GEM is unique compared to the J > other GEM-based compilers.  We have our own code generation and just useG > GEM for object file writing, peepholing, instruction scheduling, etc. J > The fact that there exists a current 8086 GEM doesn't help Macro32 much.A >   For instance, for the port to Itanium, we had to gut the code C > generation modules from the OpenVMS Alpha compiler and completely A > reimplement them for Itanium at a cost of several person-years.   8 Well, there are registers, and then there are registers.  @ Going back to the 6502, at first glance, it seemed real light on? registers.  But after a while programming it, you realized that / basically page zero was a set of 256 registers.   B There's no reason that the calling standards for the IA32 couldn'tA define R0-R15 (or R31, or whatever) as a memory resident block of 7 "registers" pointed to by one of the segment registers.   F I agree that the compilers would be the big issue in porting to IA32. E Without the compilers, it can't be done.  With the compilers, I think  that it would be very doable.   F I still think that there is no technical reason that it can't be done, and done well.  G ----------------------------------------------------------------------- $ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com     Fax: 817-237-3074    ------------------------------  # Date: Sat, 10 Jan 2004 05:22:13 GMT > From: Michael Austin <maustin@no-more-spam.firstdbasource.com> Subject: Re: CSWS V2 and ODS5 < Message-ID: <9_LLb.5802$Lx7.3486@newssvr22.news.prodigy.com>   Rudolf Wingert wrote:    > Hello, > N > did Im read right, that you need ODS5 for CSWS V2? We dont like to use ODS5R > disks. We love ODS2 with its lovely filename restriction. Esspacialy that all isE > uppercase. If I do not have ODS5, am I not able to install CSWS V2?  >  > Best regards R. Wingert  >   F BTW, If you are planning on using Oracle 9iRAC, you will also need an F ODS5 disk.  -- the reason for the "restriction" is that both use JAVA H exensively and the JAR libraries all have explicit multi-dot/mixed-case I file names. (9iRAC also REQUIRES a windowing server -- DECWindows/Motif)  $ no more command-line installations.)  I So, use a very small (9 or 18GB) ODS5 disk for installing these binaries  / and ODS2 disks for all of the data directories.   + This approach seems to work well -- so far.    Michael Austin.    ------------------------------  % Date: Fri, 09 Jan 2004 13:28:34 -0800 , From: Ken Fairfield <My.Full.Name@intel.com> Subject: Re: ds10l blowout+ Message-ID: <btn6e3$8ub$1@news01.intel.com>    Ken Farmer wrote: @ > "Keith Parris" <keithparris_NOSPAM@yahoo.com> wrote in message8 > news:cf15391e.0401081626.b50fd16@posting.google.com... > 3 >>jordan@ccs4vms.com (Rich Jordan) wrote in message  > ; > news:<cc5619f2.0401081152.7f6bef7e@posting.google.com>...  > C >>>If only DS10Ls had two PCI slots... one just won't cut it.  Need I >>>decent graphics _and_ either KZPCA or KZPEA SCSI for it to replace any ( >>>existing EV5x box we have in service. >>E >>There's a Storage/Network/Video Combo card in the Supported Options ' >>List for the DS10L.  Would this work: H >>3X-DEPVZ-AA  PCI TO 10/100MBPS EHTERNET, GRAPHICS,LVD SCSI COMBO ADPTR [...] 9 > I have that card in 6 DS10L's.  They work like a charm.   > The rub is the price: US$999 list.  I suppose David _might_ be9 able to find some at below list, but in any case, I doubt 7 seriously you'll be able to find one for much less than ; David's price for the system!  (E-Bay anyone?  Doubt enough ' were sold for them to now be excess...)   : Ken, do you recall what you paid for yours?  I presume you> bought the DS10L's new, so the incremental cost of the adapter; wouldn't have been such a barrier (as it is for hobbiests)?    	-Ken  --  6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfield ! D1C Automation VMS System Support " who:   kenneth dot h dot fairfield where: intel dot com   ------------------------------  % Date: Sat, 10 Jan 2004 00:41:21 -0600 / From: Chris Scheers <chris@applied-synergy.com>  Subject: Re: ds10l blowout3 Message-ID: <3FFF9E91.9BF6BA1C@applied-synergy.com>    Keith Parris wrote:  > m > jordan@ccs4vms.com (Rich Jordan) wrote in message news:<cc5619f2.0401081152.7f6bef7e@posting.google.com>... D > > If only DS10Ls had two PCI slots... one just won't cut it.  NeedJ > > decent graphics _and_ either KZPCA or KZPEA SCSI for it to replace any) > > existing EV5x box we have in service.  > E > There's a Storage/Network/Video Combo card in the Supported Options ' > List for the DS10L.  Would this work: H > 3X-DEPVZ-AA  PCI TO 10/100MBPS EHTERNET, GRAPHICS,LVD SCSI COMBO ADPTR > B > Graphics capability described as "2D Video with 4 MB of graphicsG > memory (based on 3Dlabs Permedia2V video ASIC), similar to the Compaq @ > SN-PBXGK-BB".  Does this qualify as "decent graphics" for your > purposes?   D Is this an Intraserver 6221L-4U or does it require special firmware?  G ----------------------------------------------------------------------- $ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com     Fax: 817-237-3074    ------------------------------  # Date: Sat, 10 Jan 2004 02:57:46 GMT - From: "Gary Cooper" <gwcooper@mindspring.com> 1 Subject: Experienced OpenVMS Admin/Guru available ? Message-ID: <KSJLb.1059$i4.60@newsread1.news.atl.earthlink.net>   C Over 20 yrs experience on OpenVMS, Alpha, VAX, Network, SAN, etc... % Contact me at gwcooper@mindspring.com    --   Gary Cooper    ------------------------------  % Date: Fri, 09 Jan 2004 22:23:19 -0600 5 From: "Martin P.J. Zinser" <zinser@zinser.no-ip.info>  Subject: Re: GNU build tools ?? 9 Message-ID: <btnuno$9ij7s$1@ID-209632.news.uni-berlin.de>    ASDF wrote: L > Can somebody out here provide me the pointer to GNU build tools (automake, > autoconf etc...) on VMS? >  > I As others have mentioned, the latest version of GNU make can be build on  F   VMS without problems (even using GNU make ;-). One caveat, you will E have to setup VMS specific Makefiles since not all the various shell  ! commands are easily translatable.   E As far as I know there are at least no current versions of automake,   autoconf available.   F What I do have are a number of DCL snippets I do reuse while building F build procedures for Unix to VMS ports. I also have a Perl script you E can run against a config.h.in to produce DCL that then can be used to C generate the "right" config.h automatally depending on your system  I setup. Caveat!! This is work in progress. The script recognizes a number  G of the more common HAVE_FOO and NEED_BAR macros, there is stuff missing < I am adding over time. If you want to help, you are welcome.  G There is some in-line commenting, but no full documentation. The stuff  : is available at http://zinser.no-ip.info/wwwar/mgr/btools/   Greetings, Martin   H P.S. And before somebody asks, the script is written in Perl since I do G most of this work on my notebook while riding the train. And that runs  H Linux. Since Perl is portable enough for this task I can run the script @ on VMS and Linux regardless. Also I like Perl for such work, it F certainly could have been written in C, Fortran, Cobol, DCL or BLISS,  but hey it was my choice ;-)   ------------------------------  % Date: Fri, 09 Jan 2004 15:09:51 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>Y Subject: Re: Hewlett-Packard plans to EOL the HP AlphaServer Tru64. Find out how Sun can  ) Message-ID: <3FFF0A8E.EDD7BAF6@istop.com>    Fabio Cardoso wrote: > E > Hewlett-Packard plans to EOL the HP AlphaServer Tru64. Find out how , > Sun can help you migrate to a new platform    I "plans" ???? Hasn't it been EOLed already ? Or is there still some active 4 development being done with new versions/features ?   N I was under the impresison that Tru64 was going to be developped "for a littleL whitle", but after that would go into maintenance mode. Does anyone have the& timeframe this is expected to happen ?   ------------------------------  % Date: Fri, 09 Jan 2004 20:55:05 -0800 % From: Dean Woodward <deanw@rdrop.com> 3 Subject: Re: Hobbyist Layered Product license PAKs? ( Message-ID: <3FFF85A9.8010609@rdrop.com>   Brian Chase wrote:< > Bob Koehler <koehler@eisner.aspm.encompasserve.org> wrote: >>   Worked for me last week.  > It worked for me that time.    Worked for me on 5-JAN-2004...  L > BTW, the following are showing up in between OPENVMS_HOBBYIST and AMS now: > + > $ LICENSE REGISTER UCX-IP-CLIENT        - # > $ LICENSE REGISTER UCX-IP-NFS   - # > $ LICENSE REGISTER UCX-IP-RT    -   I But I don't see those. You're saying you have licenses for those in your   LP set? Hmm.  H > I don't see these listed in the hobbyist layered product pages yet, so7 > perhaps it was something related to them being added.    ------------------------------  + Date: Sat, 10 Jan 2004 06:20:02 +0000 (UTC) - From: bdc@world.std.com (Brian 'Jarai' Chase) 3 Subject: Re: Hobbyist Layered Product license PAKs? ( Message-ID: <bto5ii$ebl$1@pcls4.std.com>  ( In article <3FFF85A9.8010609@rdrop.com>,' Dean Woodward  <deanw@rdrop.com> wrote:  > Brian Chase wrote:  - > > $ LICENSE REGISTER UCX-IP-CLIENT        - % > > $ LICENSE REGISTER UCX-IP-NFS   - % > > $ LICENSE REGISTER UCX-IP-RT    -  > K > But I don't see those. You're saying you have licenses for those in your   > LP set? Hmm.  D Yes, those came in the license PAKs generated for me on January 7th.   -brian.  --  F --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ ----->       IBM 1130 Club Motto:  "Computer People Are Happy People"   ------------------------------  # Date: Fri, 09 Jan 2004 22:01:21 GMT 9 From: Hein van den Heuvel <hein_netscape@eps.zko.dec.com> 0 Subject: Re: how to flush a file opened by DCL ?/ Message-ID: <3FFF24AC.130355EE@eps.zko.dec.com>   / Phillip Helbig---remove CLOTHES to reply wrote:   3 > > any idea of how to flush a file opened by DCL ?  >  > From another process:  > 3 >    $ OPEN/READ/SHARE=WRITE logical-name file-name  >    $ CLOSE logical-name    I don't think that is enough... - CONVERT/SHARE does not 'see' the data either. H It maybe time for an extra /share switch for type and a modifier for the existing /share for convert   H A program (or DCL) opening with RAC=GET,SHR=UPD will see the new records see log below.   Hein.     --- first session (VMS V7.1) ---  
 $ cre tmp.tmp  aap  noot mies  Exit ! $ open/appen/share=read x tmp.tmp  $ write x "teun"   --- Second session ----    $ type tmp.tmp aap  noot mies! $ open/read/share=write x tmp.tmp 	 $ close x  $ type tmp.tmp aap  noot mies $ ! $ open/read/share=write x tmp.tmp  $ read x record  $ read x record  $ read x record  $ read x record   %RMS-E-EOF, end of file detected $ show symb record   RECORD = "teun" 	 $ close x  $  $ type tmp.tmp aap  noot mies
 $ cre tmp.bas F 1       open "tmp.tmp" for input as file #1, access read, allow modify         for i = 1 to 4           input #1, record$            print record$          next i         end 	 $ bas tmp 	 $ lin tmp 	 $ run tmp  aap  noot mies teun $ type tmp.tmp aap  noot mies $  ----- first session ---  $close x $  ---- second session ---  $ type tmp.tmp aap  noot mies teun $    ------------------------------  + Date: Sat, 10 Jan 2004 00:40:21 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)0 Subject: Re: how to flush a file opened by DCL ?$ Message-ID: <btnhll$7tm$4@online.de>  C In article <3FFF24AC.130355EE@eps.zko.dec.com>, Hein van den Heuvel ( <hein_netscape@eps.zko.dec.com> writes:   5 > > > any idea of how to flush a file opened by DCL ?  > >  > > From another process:  > > 5 > >    $ OPEN/READ/SHARE=WRITE logical-name file-name  > >    $ CLOSE logical-name  > ! > I don't think that is enough... / > CONVERT/SHARE does not 'see' the data either. J > It maybe time for an extra /share switch for type and a modifier for the > existing /share for convert   H I've used this a lot when DIR/SIZ=ALL shows 0 used but many allocated.  F The above command lets one see the data.  Maybe it depends on what is 0 writing to the file and not flushing the buffer.   ------------------------------   Date: 9 Jan 2004 12:55:00 -0600 4 From: kuhrt@nospammy.encompasserve.org (Marty Kuhrt)) Subject: Re: HP TCPIP SMTP Queues (again) 3 Message-ID: <T9Bhtu1aUzNo@eisner.encompasserve.org>   X In article <3ffdccea$1@obsidian.gov.bc.ca>, "Alder" <MUNDDGNTDYTV@spammotel.com> writes:M > ...but the execution queue still hangs (an hour so far) after the mail item M > is successfully delivered to the local VMS Mail account.  Not sure where to I > go from here, so any more suggestions will be welcome.  Here's what the  > config looks like now:  B This is a known bug with one of the SYS patches.  When the system B tries to broadcast the newmail message it will hang.  You need to % get and install the latest SYS patch.    ------------------------------  $ Date: Fri, 9 Jan 2004 12:21:04 -0800* From: "Alder" <MUNDDGNTDYTV@spammotel.com>) Subject: Re: HP TCPIP SMTP Queues (again) + Message-ID: <3fff0d30$1@obsidian.gov.bc.ca>   H I thought I was covered with the VMSUPDATE V3 ECO, but you're right, the? docs SYS ECO (version 7) was not included in the consolidation.   3 Thanks, I'll give it a spin and see what falls out.    Cheers,  Alder   A "Marty Kuhrt" <kuhrt@nospammy.encompasserve.org> wrote in message - news:T9Bhtu1aUzNo@eisner.encompasserve.org... 5 > In article <3ffdccea$1@obsidian.gov.bc.ca>, "Alder" $ <MUNDDGNTDYTV@spammotel.com> writes:J > > ...but the execution queue still hangs (an hour so far) after the mail itemL > > is successfully delivered to the local VMS Mail account.  Not sure where toK > > go from here, so any more suggestions will be welcome.  Here's what the  > > config looks like now: > C > This is a known bug with one of the SYS patches.  When the system C > tries to broadcast the newmail message it will hang.  You need to ' > get and install the latest SYS patch.    ------------------------------  % Date: Fri, 09 Jan 2004 15:38:15 -0500 * From: JF Mezei <jfmezei.spamnot@istop.com>) Subject: Re: HP TCPIP SMTP Queues (again) ( Message-ID: <3FFF1134.B9D7F16@istop.com>   Alder wrote:- >   "TCPIP$SMTP_RECV_SNAPSHOT_BLOCKS" = "200" - >   "TCPIP$SMTP_SYMB_SNAPSHOT_BLOCKS" = "200"   L Do not define those logicals. The snapshop only causes writing to disk afterH some errors the symbiont thinks it has detected. It is possible that the" server hangs before it gets there.  5 This is probably why your log files show no activity.    ------------------------------  # Date: Sat, 10 Jan 2004 03:01:19 GMT ( From: Alder <PGDEHMKOKIMD@spammotel.com>) Subject: Re: HP TCPIP SMTP Queues (again) * Message-ID: <3WJLb.2638$n44.1537@clgrps13>   Marty Kuhrt wrote::   Z > In article <3ffdccea$1@obsidian.gov.bc.ca>, "Alder" <MUNDDGNTDYTV@spammotel.com> writes: > M >>...but the execution queue still hangs (an hour so far) after the mail item M >>is successfully delivered to the local VMS Mail account.  Not sure where to I >>go from here, so any more suggestions will be welcome.  Here's what the  >>config looks like now: >  > D > This is a known bug with one of the SYS patches.  When the system D > tries to broadcast the newmail message it will hang.  You need to ' > get and install the latest SYS patch.   I That's exactly what the problem was.  I read Chris Barratt's second post  F a little too quickly; he'd also suggested updating the VMS_SYS patch, H but I was hasty and only saw his reference to the VMS_UPDATE patch.  So  thanks Marty and Chris!   
 What a group!    Cheers,  Alder    ------------------------------  # Date: Sat, 10 Jan 2004 03:06:29 GMT ( From: Alder <PGDEHMKOKIMD@spammotel.com>) Subject: Re: HP TCPIP SMTP Queues (again) * Message-ID: <V_JLb.2640$n44.2160@clgrps13>   JF Mezei wrote::   > Alder wrote: > - >>  "TCPIP$SMTP_RECV_SNAPSHOT_BLOCKS" = "200" - >>  "TCPIP$SMTP_SYMB_SNAPSHOT_BLOCKS" = "200"  >  > N > Do not define those logicals. The snapshop only causes writing to disk afterJ > some errors the symbiont thinks it has detected. It is possible that the$ > server hangs before it gets there. > 7 > This is probably why your log files show no activity.    JF.   G It turned out that I was missing a VMS_SYS patch kit (version 7).  The  ( queues seem to be working just fine now.  E I want to send you a big thank you, though, for all the patience you  H showed getting me through this.  My grasp of HP's SMTP product is a lot ? better now that I've had the benefit of your posts.  Now I can  % concentrate on blocking spam! oh joy.    Merci beaucoup!  Alder    ------------------------------  % Date: Fri, 09 Jan 2004 12:56:25 -0600 2 From: "-Andy-" <see2go4me@spamdelicious.yahoo.com>. Subject: Re: I just scored an AlphaServer 800!6 Message-ID: <Xns946B8DCEDD4EFsee2go4me@216.196.97.132>  = Steve Kulpa <stevekulpaFUDGE@yahoo.com> enlightened us with:    3 > On Thu, 8 Jan 2004 13:36:53 -0500, "Peter Weaver" 0 > <WeaverConsultingServices@sympatico.ca> wrote:  3 >> and please, nobody tell him that if he posts his 6 >> location then there would probably be someone close' >> to him who will loan him the CD set    0 Shoot, I'm sure I can find a recent distribution3 somewhere... haven't quite finished unpacking since  moving last summer.   8 > You mean if I were to post on c.o.v. and say something6 > like "I live in Knoxville, TN and I was wondering if9 > someone nearby had a hobbyist CD I could borrow", there . > would be a chance that someone would reply?   6 Possibly. Depends on when you want it. This weekend - < you'll have to find Pleasant Ridge Rd. (It's off of Western < Ave.)... during the week (Excepting today)... East Hill Ave.   If that helps.   -Andy-  < f$edit("a n d r e w . s t o f f e l @ j e n z a b a r . n e  t","COLLAPSE")     ------------------------------  # Date: Fri, 09 Jan 2004 21:49:55 GMT , From: Wayne Morrison <Wayne.Morrison@hp.com>; Subject: Re: Kerberos installation fails with 7.3-2 upgrade & Message-ID: <3FFF2132.C0549F96@hp.com>   Harri Klemetti wrote:  >  > Hi,  > G > I'm trying to upgrade a DS15 from 7.3-1 to 7.3-2, but installation of A > Kerberos fails during upgrade and thus the whole upgrade fails.  > A > The installation procedure tells me to perform a manual overlay D > installation of Kerberos 2.0-6, but even after having done this asH > instructed the upgrade procedure still fails at Kerberos installation. >  > Any ideas? >  > BR,  >  > Harri   H The only other time we've seen this happen, it was due to a corrupt PCSIK database.  Please contact me off-line, and we'll determine what's happening  with your upgrade.   	Wayne Morrison  	Kerberos & CDSA Project Leader  	OpenVMS Engineering   ------------------------------  # Date: Fri, 09 Jan 2004 22:06:00 GMT , From: Wayne Morrison <Wayne.Morrison@hp.com>; Subject: Re: Kerberos installation fails with 7.3-2 upgrade & Message-ID: <3FFF24F7.9B932CD0@hp.com>   Harri Klemetti wrote:  >  > Hi,  > G > I'm trying to upgrade a DS15 from 7.3-1 to 7.3-2, but installation of A > Kerberos fails during upgrade and thus the whole upgrade fails.  > A > The installation procedure tells me to perform a manual overlay D > installation of Kerberos 2.0-6, but even after having done this asH > instructed the upgrade procedure still fails at Kerberos installation. >  > Any ideas? >  > BR,  >  > Harri   H The only other time we've seen this happen, it was due to a corrupt PCSIK database.  Please contact me off-line, and we'll determine what's happening  with your upgrade.   	Wayne Morrison  	Kerberos & CDSA Project Leader  	OpenVMS Engineering   ------------------------------  # Date: Fri, 09 Jan 2004 22:16:42 GMT , From: Wayne Morrison <Wayne.Morrison@hp.com>; Subject: Re: Kerberos installation fails with 7.3-2 upgrade & Message-ID: <3FFF2779.6700B396@hp.com>   Harri Klemetti wrote:  >  > Hi,  > G > I'm trying to upgrade a DS15 from 7.3-1 to 7.3-2, but installation of A > Kerberos fails during upgrade and thus the whole upgrade fails.  > A > The installation procedure tells me to perform a manual overlay D > installation of Kerberos 2.0-6, but even after having done this asH > instructed the upgrade procedure still fails at Kerberos installation. >  > Any ideas? >  > BR,  >  > Harri   H The only other time we've seen this happen, it was due to a corrupt PCSIK database.  Please contact me off-line, and we'll determine what's happening  with your upgrade.   	Wayne Morrison  	Kerberos & CDSA Project Leader  	OpenVMS Engineering   ------------------------------  # Date: Fri, 09 Jan 2004 23:16:53 GMT , From: Wayne Morrison <Wayne.Morrison@hp.com>; Subject: Re: Kerberos installation fails with 7.3-2 upgrade & Message-ID: <3FFF3594.82158467@hp.com>   Harri Klemetti wrote:  >  > Hi,  > G > I'm trying to upgrade a DS15 from 7.3-1 to 7.3-2, but installation of A > Kerberos fails during upgrade and thus the whole upgrade fails.  > A > The installation procedure tells me to perform a manual overlay D > installation of Kerberos 2.0-6, but even after having done this asH > instructed the upgrade procedure still fails at Kerberos installation. >  > Any ideas? >  > BR,  >  > Harri   H The only other time we've seen this happen, it was due to a corrupt PCSIK database.  Please contact me off-line, and we'll determine what's happening  with your upgrade.   	Wayne Morrison  	Kerberos & CDSA Project Leader  	OpenVMS Engineering   ------------------------------  # Date: Fri, 09 Jan 2004 22:11:09 GMT # From: Jim Mehlhop <jim@mehlhop.org> ( Subject: Re: Multnet Documentation V4.0?= Message-ID: <1GFLb.10084$uF6.2999117@news1.news.adelphia.net>   G MN4.4 is the current version and is VERY similar with the exception of  
 SSH and XNTP.   I Most of the basic stuff in the 4.0 manuals would still work or vice versa        John Brandon wrote:X > Anyone using Multinet V4.0?  > + > And where could I find the documentation?  > 1 > Is the current version (V4.4 or V4.3) similiar?n >  > TIA! >  >  > J*o*h*n B*r*a*n*d*o*nn > VMS Systems Administrator-, > firstname.lastname.spam.me.not@dalsemi.com   ------------------------------   Date: 9 Jan 2004 11:00:07 -0800 ) From: brian@wadsworth.org (Brian Cuttler)e, Subject: MX - create outbound mail with FQDN= Message-ID: <e6215e11.0401091100.78b36da7@posting.google.com>o   VMS 5.5r MX 4.1  A I'm having trouble directing email to the internet and have found G that if I properly define "MX_VMSMAIL_LOCALHOST" and create (correctly) H all of the proper DNS records on my external DNS server I can send mail.  7 Mail internally to my site has always worked correctly.i  C What I need to do is fix the MX config so that MX_VMSMAIL_LOCALHOST8: is FQDN rather than short/cname following the next reboot.  C The switch to do this isn't obvious to me and the data is stored int an indexed .dat file.g  F Yes - we are planning to decommission, 18 month horizon when I startedF here in '98. I'm not complaining though, I love VMS and you can't kill either the OS nor the hardware.Q   Thanks for your help,M   Briane ---->    Brian R Cuttler                 brian.cuttler@wadsworth.org3    Computer Systems Support        (v) 518 486-1697f3    Wadsworth Center                (f) 518 473-6384 9    NYS Department of Health        Help Desk 518 473-07732   ------------------------------  % Date: Fri, 09 Jan 2004 14:26:08 -0800o" From: Brad Hughes <brad@tgsmc.com>% Subject: Re: MySQL 4.0.17 for OpenVMSe( Message-ID: <3FFF2A80.2010801@tgsmc.com>   jf.pieronne@laposte.net wrote: [...]oL > I have a nearly completely automatic porting process for MySQL, Python,...M > So it is, generally, very easy to port a new version of one of these tools.   E Have you tried throwing PHP at that automatedporting process...   ;-)    brad   ------------------------------   Date: 9 Jan 2004 11:30:36 -0800 7 From: jones.computer.srv@worldnet.att.net (Daryl Jones)i< Subject: Re: Non-WS Process Quotas not in Performance Manual= Message-ID: <8a646952.0401091130.6776e342@posting.google.com>:  U Jan-Erik Sderholm <aaa@aaa.com> wrote in message news:<3FFE64C1.D1CA07AD@aaa.com>... w > > spamsink2001@yahoo.com (Alan E. Feldman) wrote in message news:<b096a4ee.0401081323.2bbf489b@posting.google.com>... C > > > Why aren't process quotas like DIOlm, BIOlm, ASTlm, etc., notu* > > > discussed in the performance manual? > > >  > = > Perhaps becuse that are not *performance* specific quotas ?t; > They are rather process *limits*, either there are enougho= > of them, or there isn't. Other quotas (WS and others) has aa5 > more "soft" limit and are more performance related.e >  > Jan-Erik.o   Dear Jan-Erik Soderholm:  E In the case of Backup performance, these process quotas turned out tooB be performance parameters. During a backup the process quotas willD limit the performance of the backup image and not error out like theC execution of other images. Increasing these parameters will per ther> article will decrease the runtime of the backup. Much like theE WSdefault, WSquota, WSexetent, and other parameters you have mention.-   Regards, Daryl Jonesa   ------------------------------   Date: 9 Jan 2004 19:14:40 -0800r. From: spamsink2001@yahoo.com (Alan E. Feldman)< Subject: Re: Non-WS Process Quotas not in Performance Manual= Message-ID: <b096a4ee.0401091914.17f1d930@posting.google.com>u  | jones.computer.srv@worldnet.att.net (Daryl Jones) wrote in message news:<8a646952.0401091130.6776e342@posting.google.com>...W > Jan-Erik Sderholm <aaa@aaa.com> wrote in message news:<3FFE64C1.D1CA07AD@aaa.com>... y > > > spamsink2001@yahoo.com (Alan E. Feldman) wrote in message news:<b096a4ee.0401081323.2bbf489b@posting.google.com>...2E > > > > Why aren't process quotas like DIOlm, BIOlm, ASTlm, etc., noti, > > > > discussed in the performance manual?  @ Oops. I should have used either "aren't" or "not", but not both!  ? > > Perhaps becuse that are not *performance* specific quotas ?s= > > They are rather process *limits*, either there are enoughe? > > of them, or there isn't. Other quotas (WS and others) has at7 > > more "soft" limit and are more performance related.  [...]tG > In the case of Backup performance, these process quotas turned out todD > be performance parameters. During a backup the process quotas willF > limit the performance of the backup image and not error out like theE > execution of other images. Increasing these parameters will per the @ > article will decrease the runtime of the backup. Much like theG > WSdefault, WSquota, WSexetent, and other parameters you have mention.  [...]i  = Then why do these quotas even exist if they have no effect onoE performance (save for BACKUP)? Why not just set them all to 2^32-1 or  whatever they max out at?a  E At a previous job we had CAM doing remote backups from VAX to offsite-E silo and when the CAM processes ran, anything else on the machine ran E very poorly. The CAM processes had the large BACKUP quotas and ran at B priority 4. I assume it was the large process quotas combined with> priority 4 that mode CAM such a resource hog that slowed other@ processes. (I don't recall the details, but it seemed that otherE processes definitely ran slower than one would expect they would whensC competing with a very few other processes at the same priority, buto that was years ago.)  A And if these quotas don't affect performance, why does the BACKUPSF quotas section in the manual say, "If your disks are highly fragmentedC or if your backups will be performed during periods of heavy system F use, you should reduce the values shown for WSQUOTA and FILLM."? Also,B what about PGFLQUOTA? Couldn't that have performance effects if it were too large or too small?  A And if these non-WS quotas don't affect performance, what do they, affect? What is their purpose?   Thanks.i   Alan E. Feldmans   ------------------------------   Date: 9 Jan 2004 21:25:54 -0600F- From: Kilgallen@SpamCop.net (Larry Kilgallen)t< Subject: Re: Non-WS Process Quotas not in Performance Manual3 Message-ID: <uuMGLBEDDCm8@eisner.encompasserve.org>h  n In article <b096a4ee.0401091914.17f1d930@posting.google.com>, spamsink2001@yahoo.com (Alan E. Feldman) writes:~ > jones.computer.srv@worldnet.att.net (Daryl Jones) wrote in message news:<8a646952.0401091130.6776e342@posting.google.com>...  H >> In the case of Backup performance, these process quotas turned out toE >> be performance parameters. During a backup the process quotas willsG >> limit the performance of the backup image and not error out like the,F >> execution of other images. Increasing these parameters will per theA >> article will decrease the runtime of the backup. Much like theUH >> WSdefault, WSquota, WSexetent, and other parameters you have mention. > [...]g > ? > Then why do these quotas even exist if they have no effect ontG > performance (save for BACKUP)? Why not just set them all to 2^32-1 ore > whatever they max out at?i  ? Process quotas control access to system resources by individuali? processes.  If everybody had infinite quotas, one piggy programr( could bring the machine to a standstill.   ------------------------------  % Date: Fri, 09 Jan 2004 13:16:10 -0600X2 From: "-Andy-" <see2go4me@spamdelicious.yahoo.com># Subject: Re: OpenVMS Certificationsa6 Message-ID: <Xns946B912829F9Esee2go4me@216.196.97.132>  7 "Peter Weaver" <WeaverConsultingServices@sympatico.ca>  # enlightened us on 09 Jan 2004 with:d  9 > That's the second person in this thread to say they gett' > CONDIST through their certifications.d  < Add a couple more to that total :-). A fellow emp. & I have 3 been getting them for a while now. We both did the M; "certification thing" in Fall 1999 because it was a Compaq  : requirement because we sell/install VMS systems to run our
 software on. e  9 To get the distributions you had to actually sign up for r8 them, it wasn't automatic like the annoying PC coasters.   -Andy-   ------------------------------   Date: 9 Jan 2004 16:07:02 -0800l7 From: jones.computer.srv@worldnet.att.net (Daryl Jones)iY Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per minute us8= Message-ID: <8a646952.0401091519.25c9eb92@posting.google.com>-   Dear Kieth Parris:  4 Oracle RDB runs on VAX/VMS, OpenVMS, and Tru64 UNIX.   Regards, Daryl Jonesd  s fabiopenvms@yahoo.com.br (Fabio Cardoso) wrote in message news:<f30679fb.0401090506.5766cac2@posting.google.com>....x > keithparris_NOSPAM@yahoo.com (Keith Parris) wrote in message news:<cf15391e.0401081508.2b02c5ac@posting.google.com>...I > > HP OpenVMS operating system and Oracle Rdb software break performance, > > barrierc > > by Marc Courchesne > > H > > Recent performance tests of Oracle Rdb version 7.1.2 and HP OpenVMSA > > version 7.3-2 running on an HP AlphaServer GS1280 system haveoB > > demonstrated breakthrough transaction performance for a singleH > > 32-processor symmetric multiprocessing (SMP) machine configured withG > > 256 gigabytes of main memory. The AlphaServer system, together withhC > > Oracle Rdb software, achieved sustained throughput of 1,010,160MF > > database transactions per minute. In this test, five tables with aH > > total of one billion (109) rows were created and their contents wereJ > > mapped in main memory using Oracle Rdb's Row Cache feature. This test,D > > designed to explore the limits of the Row Cache feature, was notJ > > highly optimized for performance, but additional performance tuning is? > > likely to result in even higher transaction rates. This newfF > > implementation of Row Cache, first available in Oracle Rdb versionD > > 7.1.2, allows customers to fully utilize all the physical memoryI > > present in the largest AlphaServer systems for caching databases thateI > > retain complete transaction integrity and full recovery from process, G > > disk, or system failure. The OpenVMS version 7.3-2 operating systemeJ > > provides a number of key features for improved scheduling and resourceE > > utilization that enhance scaling on systems with many processors.a4 > > These features were fully utilized in this test. > > - > > For more information, visit HP's site at h9 > > http://h71000.www7.hp.com/solutions/oracle/index.htmlm3 > > and Oracle's site at http://www.oracle.com/rdb/  > > ---eD > > Note: This test does not reflect tpm-C or tpm-H results. --Keith >  > B > May be its time to Re-bundle Oracle RDB under OVMS ! Even as anB > Oracle-HP agreement.... may be with some kind of activation key.@ > As RDB is exclusive for OVMS, there is no reason to mantain itA > unbundled...Would be good for Sybase / Ingres customers just toh@ > port their old database/applications to ORDB. And ORacle could > justify RDB investments... > 	 > Regardso >  > FC   ------------------------------  + Date: Sat, 10 Jan 2004 00:35:19 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)Y Subject: Re: Oracle Rdb on GS1280 with 7.3-2 exceeds 1 million transactions per minute us $ Message-ID: <btnhc7$7tm$3@online.de>  = In article <f30679fb.0401090506.5766cac2@posting.google.com>,z1 fabiopenvms@yahoo.com.br (Fabio Cardoso) writes: s  B > May be its time to Re-bundle Oracle RDB under OVMS ! Even as anB > Oracle-HP agreement.... may be with some kind of activation key.@ > As RDB is exclusive for OVMS, there is no reason to mantain itA > unbundled...Would be good for Sybase / Ingres customers just toA@ > port their old database/applications to ORDB. And ORacle could > justify RDB investments...  E Why would ORACLE agree to this?  Either you deprive them of a lot of aG income (not that Larry needs it---I mean, how many Migs can you fly at -G once), or make VMS unnecessarily expensive for folks who don't use Rdb.o   ------------------------------  * Date: Fri, 9 Jan 2004 22:17:39 +0000 (UTC)% From: bdc@world.std.com (Brian Chase)-) Subject: Re: Rally VAX to Alpha Migration ( Message-ID: <btn9a3$f92$1@pcls4.std.com>  R In article <FD827B33AB0D9C4E92EACEEFEE2BA2FB1E10E6@tayexc19.americas.cpqcorp.net>,& Main, Kerry <kerry.main@hp.com> wrote:  J > - While it was not Rally specific, I completed a major VAX Cobol / Rdb /I > CI mission critical, high visibility cluster to 3 node Alpha ES45 / SANeJ > / Rdb cluster earlier this year (went live in mid April). Rdb was latestF > version available. This project was extremely successful in that theG > Cust went fom 6 large VAX 7610's that were always peaked out in prime D > time to an environment whereby their new environment servers (were. > running about 20% utilization at peak time.   J Yeow.  If I were that customer, I might now be feeling a bit like I boughtF /too much/ replacement hardware.  80% is a hell of a lot of head room.   > From a pure costB > perspective, they now have a huge amount of datacenter space andG > cooling/power available for future expansion. In addition, while theyfJ > are still in the evaluation phase, this has allowed the Customer to lookJ > at all sorts of new functionality that they could not add before becauseE > of loading and version issues e.g. Java based development, wrapping D > existing Cobol logic and serving up via Web Services technologies,H > Apache web serving, Adv Server (reduce file serving virus issues while4 > consolidating Intel servers at the same time) etc.  ' Well, there is always that bright side.e   -brian.o -- oF --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ ----->       IBM 1130 Club Motto:  "Computer People Are Happy People"   ------------------------------  + Date: Sat, 10 Jan 2004 00:42:05 +0000 (UTC)tP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)) Subject: Re: Rally VAX to Alpha Migration1$ Message-ID: <btnhos$7tm$5@online.de>  H In article <btn9a3$f92$1@pcls4.std.com>, bdc@world.std.com (Brian Chase) writes:   L > > - While it was not Rally specific, I completed a major VAX Cobol / Rdb /K > > CI mission critical, high visibility cluster to 3 node Alpha ES45 / SANrL > > / Rdb cluster earlier this year (went live in mid April). Rdb was latestH > > version available. This project was extremely successful in that theI > > Cust went fom 6 large VAX 7610's that were always peaked out in primecF > > time to an environment whereby their new environment servers (were0 > > running about 20% utilization at peak time.  > L > Yeow.  If I were that customer, I might now be feeling a bit like I boughtH > /too much/ replacement hardware.  80% is a hell of a lot of head room. >  > > From a pure costD > > perspective, they now have a huge amount of datacenter space andI > > cooling/power available for future expansion. In addition, while theyCL > > are still in the evaluation phase, this has allowed the Customer to lookL > > at all sorts of new functionality that they could not add before becauseG > > of loading and version issues e.g. Java based development, wrappingeF > > existing Cobol logic and serving up via Web Services technologies,J > > Apache web serving, Adv Server (reduce file serving virus issues while6 > > consolidating Intel servers at the same time) etc.  , Yes, that will eat up the 80% headroom.  :-)   ------------------------------  % Date: Fri, 09 Jan 2004 15:29:32 -0500i* From: JF Mezei <jfmezei.spamnot@istop.com>A Subject: Re: Secure Web Server Version 2.0 available for downloadi) Message-ID: <3FFF0F2A.37592519@istop.com>    Rick Barry wrote:nM > It's the record formats with metadata, such as variable-length format, thattM > cause problems determining actual data size of a file. Since Content-Length-4 > is recommended by RFC 2616 for all content types,   F It is in fact "required" if you want to re-use connections. The headerL stipulates the exact content length, the client reads those bytes, and afterI that, instead of the connection closing, the client can send another HTTPjJ request. When you consider all the poorly designed pages that download 100K different files to display just one page that could be simple, this greatlyp. saves on http session establishement overhead.  N > Thanks to all for taking the time to provide feedback. We'll consider adding: > support for variable-length format files in the future,   I What about just simply reading the file into memory, then subtracting the-N start address of the buffer from the address of the last byte in the buffer toJ obtain length ?  After that, you can transmit the file. (Does Apache cacheK contents in memory like OSU does ?, if so, then content length shouldn't be ) such a big problem for any type of file).e       >if we discover thatL > the conversion utility is does not solve the problem in many cases. But we" > need more time to evaluate this.  H It seems rather pointless to me if you have a group dedicated to porting; Apache to VMS, but the result cannot read native VMS files.n   ------------------------------  % Date: Fri, 09 Jan 2004 13:25:53 -0800 0 From: Mark Berryman <Mark.Berryman@Mvb.Saic.Com>A Subject: Re: Secure Web Server Version 2.0 available for download ' Message-ID: <3ffeabe2$1@cpns1.saic.com>    Rick Barry wrote:tM > It's the record formats with metadata, such as variable-length format, that M > cause problems determining actual data size of a file. Since Content-LengthmL > is recommended by RFC 2616 for all content types, not just binary, we madeJ > the decision to limit support in this version to stream and fixed record
 > formats. > @ > We understood that this would affect customers who already hadM > variable-length files, which is why we provided the convert utility to help  > in some of those situations. > N > Thanks to all for taking the time to provide feedback. We'll consider addingM > support for variable-length format files in the future, if we discover thatiL > the conversion utility is does not solve the problem in many cases. But we" > need more time to evaluate this.  I It may be that we have a different reading of RFC2616 but, as I read the e' following from section 4.4 of that RFC:e  I      2.If a Transfer-Encoding header field (section 14.41) is present andsF       has any value other than "identity", then the transfer-length isD       defined by use of the "chunked" transfer-coding (section 3.6),A       unless the message is terminated by closing the connection.   F     3.If a Content-Length header field (section 14.13) is present, itsG       decimal value in OCTETs represents both the entity-length and the6G       transfer-length. The Content-Length header field MUST NOT be sent F       if these two lengths are different (i.e., if a Transfer-EncodingD       header field is present). If a message is received with both aG       Transfer-Encoding header field and a Content-Length header field,:!       the latter MUST be ignored.A   ande  C     All HTTP/1.1 applications that receive entities MUST accept thehI     "chunked" transfer-coding (section 3.6), thus allowing this mechanism H     to be used for messages when the message length cannot be determined     in advance.o  A I read this as meaning you have the choice of either providing a  A Content-Length header or using a Transfer-Encoding of "chunked". ID Chunked transfer-encoding is expressly required to be supported and C using both headers is expressly forbidden.  These requirements are .G unlikely to change since dynamically generated pages frequently cannot   provide a length in advance.  C With this in mind, I would make the following suggestion for how a e! VMS-based web-server should work:U  A 1) If the file type permits the amount the data to be calculated bH directly (e.g., stream or fix-length records) then calculate the length " and send a Content-Length: header.  G 2) If step 1 does not apply but file length hints are present then use :F them to calculate the length and send a Content-Length: header.  (For I anyone who doesn't know, file length hints are obtained the same way you 'E get the number of blocks used by the file as well as how much of the sD last block is being used, both of which are needed to determine the G amount of data in a stream format file.  File length hints contain the /F number of records in the file and the total number of bytes in all of E the records.  In all the files I have ever checked them on, they are a+ either not present or completely accurate).e  F 3) If both steps 1 and 2 fail, Transfer encode the file in chunks and ( send it without a content length header.  D Using these 3 steps will allow any file type to be served from VMS. B Please consider this or some other solution for implementation as H quickly as possible that will allow the HP supported web-server to work I with non-stream or non-fixed format files.  I cannot use the V2.0 server HH with its current limitations but I need a supported server.  Converting G the files is not practical since many are dynamically generated (using rH languages other than C).  If it will help, I will file this as a formal  bug report.   I Another issue with this version is that it appears to treat all files as sH binary and does not do any record interpretation.  In V1.3, whether the F file format was variable-length or stream-lf, each record of the file > was transmitted with a CRLF terminator.  In V2.0, the file is F transmitted as is (i.e. stream-lf files have only LF as a terminator, I stream-CR files have only CR as a terminator, stream files are sent with  A a CRLF terminator and variable-length files sent including their pG metadata but no terminator at all (since there isn't one in the file). iG While the HTTP 1.1 standard permits the use of CR, or LF, or CRLF as a  H terminator there are still many clients out there that expect only CRLF B terminated records.  I ran into this with some pages that use the E <EMBED> tag to activate various plugins.  Until I converted the file sH from stream-lf to stream, the application simply wouldn't work.  Again, 4 something that worked in V1.3 but is broken in V2.0.  
 Mark Berryman   E P.S. Some of you have suggested I simply switch to WASD or OSU.  The  G reason I cannot is the phrase "HP supported web server".  Neither WASD n  nor OSU fall into that category.   ------------------------------  * Date: Fri, 9 Jan 2004 23:54:29 +0000 (UTC), From: "Robert" <robert.heyes@btinternet.com>0 Subject: Silly Users with Password as "Password"2 Message-ID: <btnevk$dm0$1@hercules.btinternet.com>  I Ive written a program today to check the SYSUAF.DAT for all my users in a K group and find any naughty users who havent changed their password from theyG default I would set (which for example would be PASSWORD), eg. Usernames8 JBLOGGS Password PASSWORD by doing $dir nodename"JBLOGGSJ PASSWORD"::Login*.com and checking the return code. If the dir command wasJ successful then output that user name to a text file as the password hadnt
 been changed.   H What I want to do now doesnt appear to be possible... (or I cant get VMS 7.2-2 to do it!)  H I want to modify each user's account in my text file so that the user isK forced to change the password the next time they log in. I have a couple ofs3 ideas on how to do it but they may or may not work.>  H A. Set password lifetime to 0 so that the user is forced to change theirH password the next time they log in. However, how would I then change theJ password lifetime back to say 30 days once it had been changed? VMS doesnt# appear to do this off its own back.k  L AA. Set password lifetime to 7 days and wait for it to kick in automaticallyK but then the password would be forced to be changed every 7 days then and It dont want that.S  G B. If I set the account to pwdexpired then they cant log in. Full stop.4  K C. I was trying the flag where they could login once and then the next timeQL they had to change the password, but that didnt appear to be working for me.J Will keep trying. I suppose this one in theory would solve the problem but why doesnt it work!?   Any Suggestions?   ------------------------------   Date: 9 Jan 2004 18:29:10 -0600e- From: Kilgallen@SpamCop.net (Larry Kilgallen)n4 Subject: Re: Silly Users with Password as "Password"3 Message-ID: <Vn18eZ3mVQI2@eisner.encompasserve.org>1  a In article <btnevk$dm0$1@hercules.btinternet.com>, "Robert" <robert.heyes@btinternet.com> writes:s  I > B. If I set the account to pwdexpired then they cant log in. Full stop.e    That sounds like you were using:  4 	MCR AUTHORIZE MODIFY transgressor/FLAGS=PWD_EXPIRED  # whereas you should have been using:   - 	MCR AUTHORIZE MODIFY transgressor/PWDEXPIRED    ------------------------------   Date: 10 Jan 2004 00:38:26 GMT1 From: JONESD@er6.eng.ohio-state.edu (David Jones) 4 Subject: Re: Silly Users with Password as "Password": Message-ID: <btnhi2$68m$1@charm.magnus.acs.ohio-state.edu>  2 In message <btnevk$dm0$1@hercules.btinternet.com>,0   "Robert" <robert.heyes@btinternet.com> writes:J >Ive written a program today to check the SYSUAF.DAT for all my users in aL >group and find any naughty users who havent changed their password from theH >default I would set (which for example would be PASSWORD), eg. Username9 >JBLOGGS Password PASSWORD by doing $dir nodename"JBLOGGS K >PASSWORD"::Login*.com and checking the return code. If the dir command was K >successful then output that user name to a text file as the password hadnta >been changed.  O You're likely to trigger breakin evasion doing it that way, so you'll get a lotaJ of false negatives.  A more reliable method is to retrieve the appropriateP SYSUAF fields for the account via SYS$GETUAI() and test via SYS$HASH_PASSWORD().L For the most recent VMS versions, you should do the test with SYS$ACM, IIRC.    I >I want to modify each user's account in my text file so that the user isoL >forced to change the password the next time they log in. I have a couple of4 >ideas on how to do it but they may or may not work.  L Use SYS$SETUAI() to set the UAI$_PWD_DATE octaword to -1 to flag the accountJ as having a pre-expired password, which forces a password change.  This is: different than the pwdexpired bit in the UAI$_FLAGS field.     >Any Suggestions?s  3 Get a copy of the System Services Reference Manual.e      < David L. Jones               |      Phone:    (614) 292-6929- Ohio State University        |      Internet:rL 140 W. 19th St. Rm. 231a     |               jonesd@er6s1.eng.ohio-state.edu: Columbus, OH 43210           |               vman+@osu.edu  1 Disclaimer: I'm looking for marbles all day long.    ------------------------------   Date: 9 Jan 2004 18:25:03 -0600F- From: Kilgallen@SpamCop.net (Larry Kilgallen)P4 Subject: Re: Silly Users with Password as "Password"3 Message-ID: <Ljr5XbaBDKbz@eisner.encompasserve.org>n  a In article <btnevk$dm0$1@hercules.btinternet.com>, "Robert" <robert.heyes@btinternet.com> writes:tK > Ive written a program today to check the SYSUAF.DAT for all my users in a M > group and find any naughty users who havent changed their password from theeI > default I would set (which for example would be PASSWORD), eg. Usernamee: > JBLOGGS Password PASSWORD by doing $dir nodename"JBLOGGSL > PASSWORD"::Login*.com and checking the return code. If the dir command wasL > successful then output that user name to a text file as the password hadnt > been changed.N  D The password change date from SYSUAF will tell you when the password* has been changed without invasive testing.  J > I want to modify each user's account in my text file so that the user is: > forced to change the password the next time they log in.  7 Set the Password Change date to -1 (in both longwords).   C Try it on a test account before you take production action based onT$ advice you got over the Internet :-)   ------------------------------  # Date: Sat, 10 Jan 2004 00:35:00 GMTs4 From: brad@.gateway.2wire.net (Bradford J. Hamilton)4 Subject: Re: Silly Users with Password as "Password", Message-ID: <TMHLb.9033$5V2.13583@attbi_s53>  a In article <btnevk$dm0$1@hercules.btinternet.com>, "Robert" <robert.heyes@btinternet.com> writes:s !snip!I !What I want to do now doesnt appear to be possible... (or I cant get VMSi !7.2-2 to do it!)  ! I !I want to modify each user's account in my text file so that the user isrL !forced to change the password the next time they log in. I have a couple of4 !ideas on how to do it but they may or may not work. !aI !A. Set password lifetime to 0 so that the user is forced to change theiraI !password the next time they log in. However, how would I then change the K !password lifetime back to say 30 days once it had been changed? VMS doesnt-$ !appear to do this off its own back. !9M !AA. Set password lifetime to 7 days and wait for it to kick in automaticallynL !but then the password would be forced to be changed every 7 days then and I !dont want that. !eH !B. If I set the account to pwdexpired then they cant log in. Full stop. !nL !C. I was trying the flag where they could login once and then the next timeM !they had to change the password, but that didnt appear to be working for me. K !Will keep trying. I suppose this one in theory would solve the problem butt !why doesnt it work!?e !i    This seems to work (on VAX 7.2):  > mcr authorize mod XXXXXXXXXXXX/pwdexp/pass=password/pwdl="30-"  G This forces the user to login with the password "password" (your statedtO default), and then to change the password immediately upon login.  It will alsoyO fulfill the (implied) requirement for a 30-day password lifetime, if a password- lifetime did not exist before.   !Any Suggestions?0  J __________________________________________________________________________A Bradford J. Hamilton                    "All opinions are my own"0K bMradAhamiPltSon-at-coMmcAast.nPeSt     "Lose the MAPS, and replace '-at-'  0                                          with @"   ------------------------------  + Date: Sat, 10 Jan 2004 00:46:11 +0000 (UTC)lP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)4 Subject: Re: Silly Users with Password as "Password"$ Message-ID: <btni0j$7tm$6@online.de>  ; In article <btnevk$dm0$1@hercules.btinternet.com>, "Robert"a& <robert.heyes@btinternet.com> writes:   K > Ive written a program today to check the SYSUAF.DAT for all my users in a,M > group and find any naughty users who havent changed their password from the I > default I would set (which for example would be PASSWORD), eg. Username-: > JBLOGGS Password PASSWORD by doing $dir nodename"JBLOGGSL > PASSWORD"::Login*.com and checking the return code. If the dir command wasL > successful then output that user name to a text file as the password hadnt > been changed.  > J > What I want to do now doesnt appear to be possible... (or I cant get VMS > 7.2-2 to do it!) > J > I want to modify each user's account in my text file so that the user isM > forced to change the password the next time they log in. I have a couple ofu5 > ideas on how to do it but they may or may not work.a  F Use pre-expired passwords.  Users can log in the first time with your G password and are forced to change it.  This has been possible for ages.s  eM > C. I was trying the flag where they could login once and then the next timewN > they had to change the password, but that didnt appear to be working for me.L > Will keep trying. I suppose this one in theory would solve the problem but > why doesnt it work!? >  > Any Suggestions?   $ mc authorize help modif/pwdex:   MODIFY  
   /PWDEXPIREDf           /PWDEXPIRED (default)-         /NOPWDEXPIREDh  D      Specifies the password is valid for only one login. A user mustF      change a password immediately after login or be locked out of theF      system. The system warns users of password expiration. A user canF      either specify a new password, with the DCL command SET PASSWORD,D      or wait until expiration and be forced to change. By default, aE      user must change a password when first logging in to an account. D      The default is applied to the account only when the password is      being modified.   ------------------------------  % Date: Fri, 09 Jan 2004 19:22:55 -0500p* From: JF Mezei <jfmezei.spamnot@istop.com>4 Subject: Re: Silly Users with Password as "Password") Message-ID: <3FFF45CD.8BF1A68A@istop.com>a  
 Robert wrote:0I > B. If I set the account to pwdexpired then they cant log in. Full stop.t  K Nop, /pwdexpired will force them to change password before login completes.n  I What you could do is send an email to all guilty users advising them thatf" their password would expire soon.   K Then, a week later, you rerun the script on the list of guilty users and ifnL the password-change date still dates before the date the email was sent, youK then /PWDEXPIRED the account. Next time they login, they need to change it.I   ------------------------------  # Date: Fri, 09 Jan 2004 21:06:21 GMTs2 From: "Ken Farmer" <KFarmer@NOSPAM.SpyderByte.com>8 Subject: SKHPC: First Batch of 2003 SKHPC PDFs Available? Message-ID: <hJELb.227928$dl.11480729@twister.southeast.rr.com>o   2003 Archives:2 http://www.shannonknowshpc.com/pages.php?page=2003   To My Fellow HP Enthusiasts,  E In a perfect world, everyone with an interest in HP, high performance C computing, and related developments in the IT industry would opt tot; subscribe to SKHPC. The world is far from perfect, and many J people--especially hobbyists--cannot afford or cost-justify a subscriptionC to my newsletter. Since I've been earning my living researching andSH publishing SKHPC for over a decade, and since the newsletter contains noH advertising, I'm not in a position to distribute current issues of SKHPC free of charge.   L That said, I believe that many of you find the back issues of SKHPC archivedJ on this site to be of value--the ~200 issues dating from 1997 through 2002K contain a wealth of information about Digital, Compaq, and HP respectively.rL In addition, the issues provide unique insights into how and why Digital wasL acquired by Compaq--and how and why Compaq was assimilated into HP on May 7,I 2002. Hence I decided to place this information in the public domain. The-E Open Source approach to OS and software development--and the numerousgK contributors to the Open Source movement--have fundamentally changed the ITFF industry and the manner in which it is evolving, and I submit that the3 changes have been positively received by end users.r  E Rendering the SKHPC archives available to all and sundry will neithernE fundamentally change the IT industry or its evolution, but placing myeE material online and rendering it free for the downloading is bound to F benefit those with an interest in HP or its predecessors. What's more,L amassing this material in a single, easily-accessed location should simplifyH the task of researching and reviewing some very significant developmentsD that have transpired during an unprecedented consolidation of the IT	 industry.   H The latest addition to the archives covers the first calendar quarter ofK 2003. SKHPC issues published in the second calendar quarter of 2003 will beaH released at the start of the second calendar quarter of 2004, and so on.L I hope you find this information to be interesting and informative, and thatI it--together with frequent and regularly updated news briefs posted on mytL home page--will entice you and your colleagues to visit my site on a regular basis.J If you have any suggestions or feedback, feel free to email your comments,@ article ideas, observations, and constructive criticism to me at terry@shannonknowshpc.com.   2003 Archives:2 http://www.shannonknowshpc.com/pages.php?page=2003  
 Best Regards,  Terry C. Shannon Publisher, Shannon Knows HPC+ President, T.C. Shannon and Associates, LLCh  J SKHPCV10N1 - January 10, 2003 - Forward Into The Past: SKHPC's 2002 Report Card    J SKHPCV10N2 - January 14, 2003 - HP's Business Critical Server Business: NoK Longer on the Critical List; Attention Anti-Linux Bigots: Beware of the Dog     K SKHPCV10N3 - January 20, 2003 - HP's AlphaServer GS1280 Launch: A MARVELousi Development     C SKHPCV10N4 - January 26, 2003 - SKHPC Reports on the VMS Ambassador C Roundhouse; SKHPC Provides an Update on the VMS-on-IPF Boot Programw    H SKHPCV10N5 - February 1, 2003 - Mary McDowell and the ISS Unsheathe MoreC Blade Technology; HP's Penguin Lowers the Windows Flag to Half-Mast     ? SKHPCV10N6 - February 5, 2003 - HP's OpenVMS: One OS, Now THREE0 Architectures with IPF Boots    F SKHPCV10N7 - February 12, 2003 - The Sun Also Rises... or Does It? Sun8 Microsystems in the Shade; SKHPC Rectifies MARVEL Errata    B SKHPCV10N8 - February 18, 2003 - HP Gives ProLiant 8-Way Servers AB Much-Needed Midlife Kicker; HP Debuts Affordable Pavilion Desktops    H SKHPCV10N9 - February 22, 2003 - Spring 2003 IDF Conference: Intel's San Jose Summit a Stellar SuccessW    J SKHPCV10N10 - February 27, 2003 - HP Brings Home Its 1FQ03 Fiancial Report Card: No Straight-A's Here    I SKHPCV10N11 - March 5, 2003 - The Results Are In: SKHPC's VMS-on-IPF BootrL Survey; Mr. Fink's Opus: The Business and Economics of Linux and Open Source Book Review     J SKHPCV10N12 - March 10, 2003 - HP and Its ISVs Hone Blade Strategy, RivalsI Ready Responses; Chill Out! HP Announces "Smart" Cooling for Data Centerse    I SKHPCV10N13 - March 12, 2003 - HP ProLiants' Progress: Think Small, Think L Dense, Think Fast; Carly Fiorina and BEA Work the Web at BEAworld in Orlando    F SKHPCV10N14 - March 19, 2003 - Deja Vu: Rightsizing, Retrenchment, andK Reorganization at HP?, Einux Opts for Opteron and Linux Clusters to FlusterV Intelf    J SKHPCV10N15 - March 21, 2003 - HP's Linux Legions Give a Tip of the Hat toK Red Hat Software; The Latest Spin on HP Storage: New Products Real Soon Nowi       2003 Archives:2 http://www.shannonknowshpc.com/pages.php?page=2003   -- Kenneth Farmer  <><e OpenVMS.org  |  dcl.OpenVMS.org.  EnterpriseUnix.org  |  Tru64.org   ------------------------------  # Date: Fri, 09 Jan 2004 21:08:02 GMTc# From: "John Smith" <a@nonymous.com>v< Subject: Re: SKHPC: First Batch of 2003 SKHPC PDFs AvailableG Message-ID: <SKELb.46171$AJB.4465@news04.bloor.is.net.cable.rogers.com>B  
 Thanks Terry.c         Ken Farmer wrote:  > 2003 Archives:4 > http://www.shannonknowshpc.com/pages.php?page=2003 >n > To My Fellow HP Enthusiasts, >uG > In a perfect world, everyone with an interest in HP, high performancetE > computing, and related developments in the IT industry would opt toe= > subscribe to SKHPC. The world is far from perfect, and manyr? > people--especially hobbyists--cannot afford or cost-justify a B > subscription to my newsletter. Since I've been earning my livingC > researching and publishing SKHPC for over a decade, and since the > > newsletter contains no advertising, I'm not in a position to4 > distribute current issues of SKHPC free of charge. >aE > That said, I believe that many of you find the back issues of SKHPCyC > archived on this site to be of value--the ~200 issues dating from.B > 1997 through 2002 contain a wealth of information about Digital,E > Compaq, and HP respectively. In addition, the issues provide unique G > insights into how and why Digital was acquired by Compaq--and how anduG > why Compaq was assimilated into HP on May 7, 2002. Hence I decided toaG > place this information in the public domain. The Open Source approacheF > to OS and software development--and the numerous contributors to theF > Open Source movement--have fundamentally changed the IT industry andC > the manner in which it is evolving, and I submit that the changes - > have been positively received by end users.  >vG > Rendering the SKHPC archives available to all and sundry will neitheraG > fundamentally change the IT industry or its evolution, but placing mydG > material online and rendering it free for the downloading is bound tooH > benefit those with an interest in HP or its predecessors. What's more,E > amassing this material in a single, easily-accessed location shouldmF > simplify the task of researching and reviewing some very significant; > developments that have transpired during an unprecedented<# > consolidation of the IT industry.> > G > The latest addition to the archives covers the first calendar quartersC > of 2003. SKHPC issues published in the second calendar quarter of F > 2003 will be released at the start of the second calendar quarter of > 2004, and so on.E > I hope you find this information to be interesting and informative,aG > and that it--together with frequent and regularly updated news briefs F > posted on my home page--will entice you and your colleagues to visit > my site on a regular basis. B > If you have any suggestions or feedback, feel free to email yourF > comments, article ideas, observations, and constructive criticism to" > me at terry@shannonknowshpc.com. >R > 2003 Archives:4 > http://www.shannonknowshpc.com/pages.php?page=2003 >e > Best Regards,  > Terry C. Shannon > Publisher, Shannon Knows HPC- > President, T.C. Shannon and Associates, LLCV > E > SKHPCV10N1 - January 10, 2003 - Forward Into The Past: SKHPC's 2002.
 > Report Carda >s >w? > SKHPCV10N2 - January 14, 2003 - HP's Business Critical Server.@ > Business: No Longer on the Critical List; Attention Anti-Linux > Bigots: Beware of the Dog5 >2 >cC > SKHPCV10N3 - January 20, 2003 - HP's AlphaServer GS1280 Launch: Ae > MARVELous Developmentr >c > E > SKHPCV10N4 - January 26, 2003 - SKHPC Reports on the VMS AmbassadorsE > Roundhouse; SKHPC Provides an Update on the VMS-on-IPF Boot Program  >a >PE > SKHPCV10N5 - February 1, 2003 - Mary McDowell and the ISS Unsheathet@ > More Blade Technology; HP's Penguin Lowers the Windows Flag to > Half-Mastm >i >cA > SKHPCV10N6 - February 5, 2003 - HP's OpenVMS: One OS, Now THREEy > Architectures with IPF Boote >b >fH > SKHPCV10N7 - February 12, 2003 - The Sun Also Rises... or Does It? Sun: > Microsystems in the Shade; SKHPC Rectifies MARVEL Errata >o >bD > SKHPCV10N8 - February 18, 2003 - HP Gives ProLiant 8-Way Servers AD > Much-Needed Midlife Kicker; HP Debuts Affordable Pavilion Desktops >d > F > SKHPCV10N9 - February 22, 2003 - Spring 2003 IDF Conference: Intel's# > San Jose Summit a Stellar Successr >m >eE > SKHPCV10N10 - February 27, 2003 - HP Brings Home Its 1FQ03 Fiancial # > Report Card: No Straight-A's Here, >r >aF > SKHPCV10N11 - March 5, 2003 - The Results Are In: SKHPC's VMS-on-IPFG > Boot Survey; Mr. Fink's Opus: The Business and Economics of Linux ande > Open Source Book Reviewa >  > E > SKHPCV10N12 - March 10, 2003 - HP and Its ISVs Hone Blade Strategy, E > Rivals Ready Responses; Chill Out! HP Announces "Smart" Cooling fore > Data Centers >m > E > SKHPCV10N13 - March 12, 2003 - HP ProLiants' Progress: Think Small,h@ > Think Dense, Think Fast; Carly Fiorina and BEA Work the Web at > BEAworld in Orlandoz >i >hH > SKHPCV10N14 - March 19, 2003 - Deja Vu: Rightsizing, Retrenchment, andE > Reorganization at HP?, Einux Opts for Opteron and Linux Clusters to  > Fluster Intelc >c >rE > SKHPCV10N15 - March 21, 2003 - HP's Linux Legions Give a Tip of theuF > Hat to Red Hat Software; The Latest Spin on HP Storage: New Products > Real Soon Now  >r >e >  > 2003 Archives:4 > http://www.shannonknowshpc.com/pages.php?page=2003   ------------------------------  # Date: Fri, 09 Jan 2004 21:37:23 GMTg# From: "John Smith" <a@nonymous.com>t< Subject: Re: SKHPC: First Batch of 2003 SKHPC PDFs AvailableH Message-ID: <naFLb.46574$AJB.13818@news04.bloor.is.net.cable.rogers.com>  6 http://www.shannonknowshpc.com/archives/SKHPCV10N4.pdf   Terry wrote:F "OpenVMS boasts imminent Itanium support, a great hardware roadmap andL engineering leverage, a comprehensive hardware and software storage roadmap,D and increased emphasis on Web-based standards (e.g. SOAP, XML, Java,G Netbeans, et al) driven by HP and partner offerings. Mr. Gorham and hisoG engineering staff believe this represents a significant growth area forcA OpenVMS, especially as customers become more acutely aware of thet# deficiencies in Windows offerings."R      L It's wonderful that customers are "more acutely aware of the deficiencies inH Windows offerings." The big question is why isn't HP advertising the VMSL solution to these Windows deficiencies to all and sundry? Terry, did you askC Mark Gorham that question, and if so what was his learned response?(      ' Build a better mousetrap and who cares.   , Build crap, advertise it, and people buy it.  < Build a better mousetrap and advertise it and people buy it.  , The lowest common denominator - advertising.   ------------------------------  % Date: Fri, 09 Jan 2004 17:27:42 -0500t* From: JF Mezei <jfmezei.spamnot@istop.com>< Subject: Re: SKHPC: First Batch of 2003 SKHPC PDFs Available) Message-ID: <3FFF2AD4.EBA03D9C@istop.com>a  J > publishing SKHPC for over a decade, and since the newsletter contains noJ > advertising, I'm not in a position to distribute current issues of SKHPC > free of charge.A   By itself, this is acceptable.  G > Rendering the SKHPC archives available to all and sundry will neithera8 > fundamentally change the IT industry or its evolution,   Which is very laudable.6   <SARCASM level="high">H However, my personal impression is that since the HP takeover began, theH author has been scrambling to become the official channel of informationO between an HP who refuses to advertise VMS and the remaining VMS customer base.a  M As such, I view it as an subtle advertising medium, and I find it very ironiceN that HP wouldn't pay to have that material freely available, at least to DECUSL members. (HP woudln't want to advertise VMS outside of the intalled base for fear of growing VMS).a    J It is also ironic that one would have to wait a full year before seing the4 "advertising" HP intended its installed base to see.
 </SARCASM>   ------------------------------  + Date: Fri,  9 Jan 2004 19:40:07 +0100 (CET)e% From: Nomen Nescio <nobody@dizum.com>i, Subject: Re: To all "patriotic" americans...8 Message-ID: <d85d3690b07ae9b1058bab9791d3b0ec@dizum.com>  O JF, it's selfish of you to only share your wisdom with rec.travel.air alone and P leave out your friends in the other groups you frequent.  Surely they would want8 to read such precious gems as this too, don't you think?  : JF Mezei, in nobody <nobody@nobody.com> disguise, wrote in <3FFE74F9.8DD770F@nobody.com>:  M >All patriotic americans who strongly believe that their government's actions K >were honest and necessary to preserve the United States, you should take ap  >look at the following document: >aD >http://www.ceip.org/files/Publications/IraqReport3.asp?from=pubdate >aJ >This is from the Carnegie Endowment for International Peace, a washington< >think tank which has spent 6 months researching the issue.  >wO >While you may choose to disbelieve your allies outside the USA to protect yourxM >own government, perhaps there is a chance you might believe a serious reporta, >prepared by americans in the United States. >n >tM >The BBC reported the issuance of this report, and as part of the BBC report, 7 >it was mentioned how it had been ignored by USA media.i > M >jesus Christ, you have proof your government lied to the american public andaN >the USA media chose not to show this ?????? What the fuck is happening in theM >USA ? Has everyone been anesthetised and remains totally paralysed, allowinguD >the government to do everything it wants without being challenged ?   ------------------------------  % Date: Fri, 09 Jan 2004 17:24:20 -0600u9 From: Steve +1 608 278 7700 <Stephen.L.Arnold@Arnold.com>l; Subject: Ultra 3 drives in ES40 systems with EV6 processorsn/ Message-ID: <btnd49$bg4$1@grandcanyon.binc.net>e  G Sorry about all the details below, but I'm getting to the point where Ie$ can't afford to gloss over anything.  F A client purchased, on my advice, upgrade options for an ES40 model 1: 	a BA610-6D card cage-  	two more 500 Mhz EV6 processors? 	another power supply (needed for the card cage and processors)o 	36GB U3 disk drivesC 	a Smart Array 5302A 2-channel U3 PCI RAID controller (3X-KZPEA-DB)DE We put in the memory and power supplies right away, but had to get tor' VMS V7.3-1 or later to use the SA5302A.-  H We upgraded to VMS V7.3-2 two weeks ago, and in preparing to install andG configure the Smart Array this weekend, I ran into a restriction that IpG hadn't seen before.  The restrictions are ambiguous or conflicting, andeF HP has not yet been able to find anyone who knows whether our proposed configuration will work.  D We intend to connect a PCI-SCSI adapter (whether RAID or not) to theG BA610-6D card cage that we purchased with the SA5302A, and populate thenB card cage with six 3R-A0919-AA 36GB 10K-rpm Ultra 3 disk drives weH purchased with the SA5302A.  If we use a RAID controller (preferred), weC will configure 5 drives in a RAID-5 configuration and use the sixtho@ drive as a spare.  If we don't use a RAID controller, we plan toG configure the drives as three host-based volume shadowing (HBVS) shadow  sets.b  B The lastest version of the ES40 QuickSpec, 10-Oct-2003, on line atD http://h18000.www1.hp.com/products/quickspecs/10392_na/10392_na.PDF,E shows, for the 3X-KPDC-BE, no restriction for VMS and "EV68 only" for G Tru64 UNIX.  This is improved support from the QuickSpec of 6-Nov-2002, : which shows "EV68 only" for VMS and "no support" for UNIX.  F However, the ES40 Supported Options Details for the 3X-KPDC-BE clearlyD state "Support restricted to EV68 systems (no EV6 or EV67 support)".  H I would like to use this device if I can, to run the disks at U3 speeds.D If I can find out it works (with maybe minor problems, e.g., InsightH Management Agents don't work all the time), I would go ahead and try it.C If not, I'd like to use the non-RAID 3X-KZPEA-DB adapter to get theB benefit of U3 speeds.e  E For the 3X-KZPEA-DB, both versions of the QuickSpec have the notationaE "support restricted to EV68 833-MHz systems only and Tru64 V5.1A as atF minimum".  Does this mean that all processors are supported under VMS,0 that none are, or that only EV68 processors are?  F The ES40 Supported Options Details for the 3X-KZPEA-DB make no mention& of this restriction under VMS or UNIX.  E I am very eager to hear from anyone who has tried to use one of these H configurations, or any configuration of the ES40 with EV6 processors andF U3 disks.  What was your configuration?  Did it work, and if not, what was the mode of failure?  ? I would also be grateful for the insights of anyone involved inn@ engineering, qualification, or product management of the ES40 orH SA5300A.  Can you clarify any of these restrictions?  Is there anyway to9 run U3 disks at U3 speeds with EV6 processors on an ES40?   - Thanks in advance for any help you can offer.t   Regards,F "Steve"   Stephen L. Arnold, Ph.D., President, Arnold Consulting, Inc.E Address   2530 Targhee Street, Madison, Wisconsin  53711-5491  U.S.A.sA Telephone +1 608 278 7700               Facsimile +1 608 278 7701 = Internet  Stephen.L.Arnold@Arnold.com   http://WWW.Arnold.com M Arnold is a registered trademark and service mark of Arnold Consulting, Inc.y   ------------------------------   Date: 9 Jan 2004 18:18:01 -0800e From: cellctl@lycos.com (Ben)t Subject: VAXstation question= Message-ID: <dd8c4466.0401091818.31c5f296@posting.google.com>   F I have a VAXstation 4000 VLC which runs VMS 5.something quite nicely. F I originally purchased it to learn VMS, but since I come from the UNIXE side, am finding it quite difficult to adapt.  Straight (v7) UNIX wascC ported to the VAX quite early on, in the form of 32/v, which can betD obtained from www.tuhs.org.  Anyone know if the VAXstations or other; non-11/7xx VAXen are capable of running 32/v?  Thanks much!g   -Ben   ------------------------------  # Date: Sat, 10 Jan 2004 02:28:39 GMTaL From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr")  Subject: Re: VAXstation question6 Message-ID: <00A2BA66.DD91707C@SSRL.SLAC.STANFORD.EDU>  ] In article <dd8c4466.0401091818.31c5f296@posting.google.com>, cellctl@lycos.com (Ben) writes:   G >I have a VAXstation 4000 VLC which runs VMS 5.something quite nicely. wG >I originally purchased it to learn VMS, but since I come from the UNIXhF >side, am finding it quite difficult to adapt.  Straight (v7) UNIX wasD >ported to the VAX quite early on, in the form of 32/v, which can beE >obtained from www.tuhs.org.  Anyone know if the VAXstations or otherh< >non-11/7xx VAXen are capable of running 32/v?  Thanks much!  N I don't know nuthin about 32/v, but if the 780 (etc) can do it, the VaxstationC ought to be able to.  Those machines could run Ultrix-32, BSD, etc.w   -- Alane   -- tO ===============================================================================i0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056 M  Paper mail to: SSRL -- SLAC BIN 99, 2575 Sand Hill Rd, Menlo Park CA   94025nO ===============================================================================n   ------------------------------  + Date: Sat, 10 Jan 2004 05:50:47 +0000 (UTC)h% From: bdc@world.std.com (Brian Chase)   Subject: Re: VAXstation question( Message-ID: <bto3rn$517$1@pcls4.std.com>  6 In article <00A2BA66.DD91707C@SSRL.SLAC.STANFORD.EDU>,K Alan Winston - SSRL Admin Cmptg Mgr <winston@SSRL.SLAC.STANFORD.EDU> wrote:a? > In article <dd8c4466.0401091818.31c5f296@posting.google.com>,e! > cellctl@lycos.com (Ben) writes:   I > >I have a VAXstation 4000 VLC which runs VMS 5.something quite nicely. gI > >I originally purchased it to learn VMS, but since I come from the UNIXeH > >side, am finding it quite difficult to adapt.  Straight (v7) UNIX wasF > >ported to the VAX quite early on, in the form of 32/v, which can beG > >obtained from www.tuhs.org.  Anyone know if the VAXstations or otherf> > >non-11/7xx VAXen are capable of running 32/v?  Thanks much! > E > I don't know nuthin about 32/v, but if the 780 (etc) can do it, thenF > Vaxstation ought to be able to.  Those machines could run Ultrix-32, > BSD, etc.   G There are a lot of underlying hardware differences between systems likebF the 11/780 and the more recent VAXstation 4000's.  My understanding isC also that DEC didn't even support Ultrix on their later desktop VAXmC systems, nothing more recent than the VAXstation 3100/M38, favoring H instead to get their Unix customers on MIPS RISC DECstation hardware andE then onto Alpha.  The only Unix I know of to run on the VS4000/VLC iss NetBSD/vax.t  J Given than 32V was mostly a proof of concept release, I don't think anyoneJ bothered to maintain it on more recent hardware.  Especially not after the2 BSD and AT&T System III releases became available.   -brian.s -- pF --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ ----->       IBM 1130 Club Motto:  "Computer People Are Happy People"   ------------------------------  % Date: Fri, 09 Jan 2004 22:55:52 +0000u) From: Tony Arnold <tony.arnold@man.ac.uk>r( Subject: Re: Vaxstation some question...9 Message-ID: <btnbhp$99chc$1@ID-207001.news.uni-berlin.de>S   Bill,o   Bill Gunshannon wrote:; > In article <btkl30$8cim7$1@id-207001.news.uni-berlin.de>,d. > 	Tony Arnold <tony.arnold@man.ac.uk> writes: > B > You know, considering how little people here actually know aboutD > Unix you really should stop trying to compare it to VMS.  You justE > come off looking silly.  While the possibility of editing the filesaB > (in this case, /etc/passwd) exists, command line utilities to doD > this in a more reasonable way have pretty much always existed.  AsD > far back as Ultrix-11 (which pre-dates VMS) on the PDP-11 this was$ > not only possible but recommended:  C I've used both for more years than I care to remember! I also said iC 'most' sysadmin, not necessarily 'all' sysadmin on Unix is done by tF editing files. I think also the flavours of Unix that came out of DEC I were much better at providing scripts and utilities to perform sysadmin, d  where as other flavours did not.  A > If you're going to continue to criticize Unix at least take thet/ > time to learn at least a little bit about it.c  H It was not a criticism of Unix, merely a fairly obvious comparison that F I thought would be useful to point out to some one about to play with ' VMS having come from a Unix background.t   Regards, Tony.e --  F Tony Arnold, Deputy to the Head of COS Division, Manchester Computing,: University of Manchester, Oxford Road, Manchester M13 9PL.F T: +44 (0)161 275 6093, F: +44 (0)870 136 1004, M: +44 (0)773 330 0039E E-mail: tony.arnold@man.ac.uk, Home: http://www.man.ac.uk/Tony.Arnolda   ------------------------------  % Date: Fri, 09 Jan 2004 15:13:00 -0500n* From: JF Mezei <jfmezei.spamnot@istop.com>5 Subject: Re: VMS naming conventions for disk volumes?s) Message-ID: <3FFF0B4B.BF0AE31C@istop.com>u   Bob Koehler wrote:I >    Generally the OS is on one disk (aka the system disk) and the user's J >    directories are at the top level of other disks.  It's not unusual toI >    have a rooted directory to put the user's directories in, instead of 
 >    the top.   ' Another consideration to make: backups.g  M If you want to backup only user files, it then makes it much easier to have a6J rooted directory or directories under which are all your user directories.A This way, you can backup the rooted directory with a "..." in the.D specification which will grab any and all user directories below it.   ------------------------------  * Date: Fri, 9 Jan 2004 21:25:43 +0000 (UTC)% From: bdc@world.std.com (Brian Chase)a5 Subject: Re: VMS naming conventions for disk volumes? ( Message-ID: <btn68n$t11$1@pcls4.std.com>  N In article <btlmfs07mr@news1.randori.com>, L. Blunt <someone@out.there> wrote:  ? > eg:  Put something like the following into systartup_vms.com:*< >         $ define/system/translate=(conc,term)    user_root  > $disk1:[user.],$disk2:[users.] > < > Then all SYSUAF accounts are defined using the location asM > USER_ROOT:[username], and this is also what the user will see as their homef > (SYS$LOGIN) directory. > M > Similarly you create other rooted logical names for specific uses, such as:$M > PROJ_ROOT for development, UTILS_ROOT, etc... Using logical names like thisoN > is one of the reason why VMS systems are so easy to manage and maintain, andL > therefore require far fewer administrators/operators than just about everyN > other OS. Being able to shift data/directories/applications/operating systemM > locations, without having to edit/alter vast numbers or scripts, databases,l0 > and uaf entries. Just alter the logical names.  J I think the main requirement in this situation is that the systems's adminG be clueful about their OS. :-)  In Windows OSes, you've got the (albeitmC weak) option of mapping logical drive letters to provide a layer of C indirection. Under Unix there's the fundamental ability of mountingnE filesystems anywhere into the directory tree, and also the ability too9 setup symbolic filesystem links as a form of indirection.a  D One thing that I do like about what you've shown above though is theG ability of a single logical to be mapped to multiple directories acrossgH volumes.  You can get a similar end result under Unix using a filesystemF automount facility or by using a directory of symlinks, but neither ofG those options are as elegant as your VMS logical example.  I'm a littleeI curious about how the precedence is handled when there's a name collisionAF between files in too or more of the referenced directories, but that's9 either an RTFM or a "try it and see" answerable question.   H I've gotten a lot of good feedback on this thread both here and offline,. so I want to thank everyone for their answers.  J At the moment, I'm primarily playing with a VS4000/VLC with a 1GB internalG system drive.  I'm going to add a second larger external disk to hold anE combination of freeware and user accounts, with top level [USERS] andhG [SOFT] directories mapped to sensible logical names.  When I get to theiF point of pulling some of my bigger systems into a cluster, I'll spread things out a bit more.   -brian.h -- gF --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ ----->       IBM 1130 Club Motto:  "Computer People Are Happy People"   ------------------------------  * Date: Fri, 9 Jan 2004 21:45:08 +0000 (UTC)% From: bdc@world.std.com (Brian Chase) 5 Subject: Re: VMS naming conventions for disk volumes?:( Message-ID: <btn7d4$bfs$1@pcls4.std.com>  $ In article <btlspi$218$1@online.de>,J Phillip Helbig---remove CLOTHES to reply <helbig@astro.multiCLOTHESvax.de> wrote:  / > At the moment, on my hobbyist cluster I have:  >  > DSA122:     OVMSVAXSYS_2 > DSA133:     ALPHASYS_3 > DSA144:     OVMSVAXSYS_4 > DSA510:     USER         > DSA520:     SOFT         > $22$DKA400: OVMSDOC071   > $33$DKA0:   DATA         > $44$DKA200: SCRATCH      > F > In a cluster, each disk must have a unique label; this explains the ' > suffixes for the system disks.  [...]   J I can certainly understand this being desirable for devices shared betweenE nodes on some sort of common storage bus (e.g. DSSI) but I'm a littletH suprised that this would be necessary for locally attached SCSI storage.  I So I can't have nodes named RED and BLUE in my cluster, both with locallynJ attached SCSI system drives labelled as "OVMSVAXSYS"?  Likewise, it soundsF like I couldn't mount two OVMSDOC071 CDs on two different nodes of the6 same cluster since there'd be a volume label conflict.   -brian.y -- dF --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ ----->       IBM 1130 Club Motto:  "Computer People Are Happy People"   ------------------------------  % Date: Fri, 09 Jan 2004 14:12:25 -0800r, From: Ken Fairfield <My.Full.Name@intel.com>5 Subject: Re: VMS naming conventions for disk volumes?w+ Message-ID: <btn90a$aam$1@news01.intel.com>h   Brian Chase wrote:   [snip]F > One thing that I do like about what you've shown above though is theI > ability of a single logical to be mapped to multiple directories acrossiJ > volumes.  You can get a similar end result under Unix using a filesystemH > automount facility or by using a directory of symlinks, but neither ofI > those options are as elegant as your VMS logical example.  I'm a littleMK > curious about how the precedence is handled when there's a name collision H > between files in too or more of the referenced directories, but that's; > either an RTFM or a "try it and see" answerable question.   B Once you've got your VLC booted, take a look at the logical names,@ SYS$SYSROOT, SYS$SPECIFIC and SYS$COMMON, and their relationship7 to SYS$MANAGER, SYS$SYSTEM, SYS$LIBRARY, SYS$HELP, etc.1  : Then create a file in SYS$SPECIFIC:[SYSMGR] and copy it toC SYS$COMMON:[SYSMGR].  Then do a $ DIRECTORY SYS$MANAGER:<the file>.A= As a second exercise, $ SET DEFAULT SYS$MANAGER: and create ao< new file.  Do a DIRECTORY to see where it was created.  This should be enlightening. :-)   ; One thing you must be aware of in using search-list logical$: names: _new_ files will be (must be) created _only_ in the; first translation (member) of the search list.  (However, cp< reating a new version of an existing file, e.g., by editing,; will write the new version back to the same location as theo: previous version.) Search lists are great for read access;< they can be confusing to difficult if you need write access.   	Good luck!  Ken  ; P.S. In case you're worried about typing so many characterso:       for DCL commands, just don't do it. :-)  As a habit,<       I always provide the full DCL command "verb" name when>       describing an action (as well as in command procedures),;       but not when actually _using_ the commands.  DCL wille:       accept the any truncation that is not ambiguous with>       another verb, e.g., "DIR" for "DIRECTORY", but not "DI",>       which is ambiguous between "DIRECTORY" and "DIFFERENCE". -- l6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfieldh! D1C Automation VMS System SupportD" who:   kenneth dot h dot fairfield where: intel dot com   ------------------------------  % Date: Fri, 09 Jan 2004 14:18:48 -0800 , From: Ken Fairfield <My.Full.Name@intel.com>5 Subject: Re: VMS naming conventions for disk volumes? + Message-ID: <btn9c9$ac9$1@news01.intel.com>n   Brian Chase wrote:   [snip]K > So I can't have nodes named RED and BLUE in my cluster, both with locallyeL > attached SCSI system drives labelled as "OVMSVAXSYS"?  Likewise, it soundsH > like I couldn't mount two OVMSDOC071 CDs on two different nodes of the8 > same cluster since there'd be a volume label conflict.  = Correct.  Volume labels, as well as storage device names (notr= sure about other devices, e.g., LAT, terminal, BG, etc.), aree= visible cluster-wide and must be unique across all members ofi the cluster.   	-Keno --  6 I don't speak for Intel, Intel doesn't speak for me...  
 Ken Fairfieldt! D1C Automation VMS System Supportm" who:   kenneth dot h dot fairfield where: intel dot com   ------------------------------  % Date: Sat, 10 Jan 2004 00:03:48 -0000s$ From: "L. Blunt" <someone@out.there>5 Subject: Re: VMS naming conventions for disk volumes?e+ Message-ID: <btnfdt010l7@news1.randori.com>t  2 "Brian Chase" <bdc@world.std.com> wrote in message" news:btn68n$t11$1@pcls4.std.com...I > In article <btlmfs07mr@news1.randori.com>, L. Blunt <someone@out.there>r wrote: >0A > > eg:  Put something like the following into systartup_vms.com:>> > >         $ define/system/translate=(conc,term)    user_root" > > $disk1:[user.],$disk2:[users.]  F I made a little error here, which I am surprised was not picked up on.  J If you use the "/trans=terminal" qualifier (tells VMS to stop logical nameC translation at this point) you cannot use other logical names (liketJ "$disk1") in the definition. You must provide real device names or not useJ the "terminal" qualifier, the "/trans=conceil" is needed for root logicals to work correctly.   > >r> > > Then all SYSUAF accounts are defined using the location asJ > > USER_ROOT:[username], and this is also what the user will see as their home > > (SYS$LOGIN) directory. > >oK > > Similarly you create other rooted logical names for specific uses, suchI as:eJ > > PROJ_ROOT for development, UTILS_ROOT, etc... Using logical names like thisL > > is one of the reason why VMS systems are so easy to manage and maintain, andoH > > therefore require far fewer administrators/operators than just about everynI > > other OS. Being able to shift data/directories/applications/operating  systemD > > locations, without having to edit/alter vast numbers or scripts,
 databases,2 > > and uaf entries. Just alter the logical names. > L > I think the main requirement in this situation is that the systems's adminI > be clueful about their OS. :-)  In Windows OSes, you've got the (albeitwE > weak) option of mapping logical drive letters to provide a layer ofyE > indirection. Under Unix there's the fundamental ability of mountinglG > filesystems anywhere into the directory tree, and also the ability tov; > setup symbolic filesystem links as a form of indirection.e  L Agreed, there are some similar things available, but I freely admit to being a VMS Bigot :^)t  K This main difference with logical names is that they are translated by VMS,iJ which makes them very flexible. Not just for directories, or single files,F but also queues (print and batch), devices (serial lines), and general4 software seting/parameters (think Windows registry).  F > One thing that I do like about what you've shown above though is theI > ability of a single logical to be mapped to multiple directories acrossnJ > volumes.  You can get a similar end result under Unix using a filesystemH > automount facility or by using a directory of symlinks, but neither ofI > those options are as elegant as your VMS logical example.  I'm a littleyK > curious about how the precedence is handled when there's a name collisioniH > between files in too or more of the referenced directories, but that's; > either an RTFM or a "try it and see" answerable question.s  L Basically the list is searched in order until a translation that can satisfyJ the command if found. When using this ensure that the user accounts do notF have the ability to create directories in the root directoy. This willH ensure that directory/file creation fails until the correct logical name translation is reached.r  J > I've gotten a lot of good feedback on this thread both here and offline,0 > so I want to thank everyone for their answers. >2L > At the moment, I'm primarily playing with a VS4000/VLC with a 1GB internalI > system drive.  I'm going to add a second larger external disk to hold aeG > combination of freeware and user accounts, with top level [USERS] and5I > [SOFT] directories mapped to sensible logical names.  When I get to theoH > point of pulling some of my bigger systems into a cluster, I'll spread > things out a bit more. >V	 > -brian.  > --H > --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ -----@ >       IBM 1130 Club Motto:  "Computer People Are Happy People"   ------------------------------  + Date: Sat, 10 Jan 2004 00:28:40 +0000 (UTC)aP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)5 Subject: Re: VMS naming conventions for disk volumes?I$ Message-ID: <btngvo$7tm$1@online.de>  H In article <btn68n$t11$1@pcls4.std.com>, bdc@world.std.com (Brian Chase) writes:   L > At the moment, I'm primarily playing with a VS4000/VLC with a 1GB internalI > system drive.  I'm going to add a second larger external disk to hold asG > combination of freeware and user accounts, with top level [USERS] and1I > [SOFT] directories mapped to sensible logical names.  When I get to thepH > point of pulling some of my bigger systems into a cluster, I'll spread > things out a bit more.  G Note that by default DISK$<volume-label> logicals will be created when i you MOUNT the disk.e  H When you have more systems in the cluster (or even before that), I wouldE really, really, really recommend host-based volume shadowing.  Except ? for the system disks, have direct connections of the members to1B different nodes, MSCP serve all disks and mount all volumes on allG nodes.  Have 3 voting nodes in the cluster with their own system-disk, 3I and you have a very, very reliable system---which can be set up with old t? hardware and hobbyist licenses, and run the newest version VMS.o   ------------------------------  + Date: Sat, 10 Jan 2004 00:32:33 +0000 (UTC)8P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)5 Subject: Re: VMS naming conventions for disk volumes?B$ Message-ID: <btnh71$7tm$2@online.de>  H In article <btn7d4$bfs$1@pcls4.std.com>, bdc@world.std.com (Brian Chase) writes:   H > > In a cluster, each disk must have a unique label; this explains the ) > > suffixes for the system disks.  [...]i > L > I can certainly understand this being desirable for devices shared betweenG > nodes on some sort of common storage bus (e.g. DSSI) but I'm a littlecJ > suprised that this would be necessary for locally attached SCSI storage. > K > So I can't have nodes named RED and BLUE in my cluster, both with locallyhL > attached SCSI system drives labelled as "OVMSVAXSYS"?  Likewise, it soundsH > like I couldn't mount two OVMSDOC071 CDs on two different nodes of the8 > same cluster since there'd be a volume label conflict.  G It's obvious if the disks are mounted on more than one node (no shared fI bus required, just MSCP serving of the disks).  I don't know if a) it is oH supported and b) it will work to have the same volume label if the disk C is directly connected to only one machine and only mounted on that cD machine.  By using the third argument to MOUNT you can probably get H around this without changing the volume label (I have never tried this, I since I always have labels appropriate to the DISK$volume-label logicals h  which will be created by MOUNT).   ------------------------------  # Date: Sat, 10 Jan 2004 05:44:47 GMTh> From: Michael Austin <maustin@no-more-spam.firstdbasource.com>5 Subject: Re: VMS naming conventions for disk volumes? < Message-ID: <jjMLb.5804$tF7.4577@newssvr22.news.prodigy.com>  / Phillip Helbig---remove CLOTHES to reply wrote:   J > In article <btn68n$t11$1@pcls4.std.com>, bdc@world.std.com (Brian Chase)
 > writes:  >  > L >>At the moment, I'm primarily playing with a VS4000/VLC with a 1GB internalI >>system drive.  I'm going to add a second larger external disk to hold a G >>combination of freeware and user accounts, with top level [USERS] and I >>[SOFT] directories mapped to sensible logical names.  When I get to the H >>point of pulling some of my bigger systems into a cluster, I'll spread >>things out a bit more. >  > I > Note that by default DISK$<volume-label> logicals will be created when e > you MOUNT the disk.o > J > When you have more systems in the cluster (or even before that), I wouldG > really, really, really recommend host-based volume shadowing.  Excepth  F Unless you have a storage sub-system that would make it immune to the H failures and frustrations you will end up with by using HBVS in a large C cluster environment -- like using an EVA with Business Copy - cool  G stuff!!!  Essentially the same thing, only the mirror is maintained at  A the controller level - offloading the system CPU to do real work.w  G I also use HSG80 Mirror sets (single HSG80 controller-pair) for System SI and application files and raid-5 for database/data drives --and thus far iI have not had system downtime due to a controller or spindle failure.  -- HE well except for the time when one of my predecessors slammed a drive P7 into the backplane a bit too hard frying the backplane.e  H And I also do not use MSCP -- we have NO local drives in any server. If C you do, then you might as  well have a Sun or other Unix "cluster"  : because if that system dies, so does your disk connection.  F Although if all you have is an LAVC, then there is basically no other  choice.e   Can you spell SAN? :)r   Michael Austin.K   ------------------------------  % Date: Fri, 09 Jan 2004 15:06:05 -0500B* From: JF Mezei <jfmezei.spamnot@istop.com>- Subject: Re: VMS to VMS copying between Nodes6) Message-ID: <3FFF09AC.2C46D656@istop.com>    Simon wrote:G > But....Isn't it one heck of a load of trouble installing Decnet V (ore- > IV) just to do a proper copy between nodes.e  K DECNET 4 is very easy to install and configure. There a a command procedure K sys$manager:netconfig.com which prompts you for everything. All you need toPL worry about is deciding on your area number as well as node number and name.  H > sub-directories at all well. Would it be possible to 'share' a disk at? > a level and then 'mount' the share on another OpenVMS node...1  M That is what clustering is for. once two nodes are clustered, either node canc* MOUNT and then access each other's drives.  M And if you MOUNT/SYSTEM/CLUSTER a drive, it automatically gets mounted on all  other nodes.   ------------------------------  % Date: Sat, 10 Jan 2004 05:31:54 +0100 2 From: martin@radiogaga.harz.de (Martin Vorlaender)- Subject: Re: VMS to VMS copying between Nodesn; Message-ID: <3fff803a.524144494f47414741@radiogaga.harz.de>P  + JF Mezei (jfmezei.spamnot@istop.com) wrote:s > Simon wrote:I > > But....Isn't it one heck of a load of trouble installing Decnet V (ora/ > > IV) just to do a proper copy between nodes., >dM > DECNET 4 is very easy to install and configure. There a a command proceduresM > sys$manager:netconfig.com which prompts you for everything. All you need toTN > worry about is deciding on your area number as well as node number and name.  E But then, you decided on an SCSSYSTEMID when installing VMS, which isn areanumber * 1024 + nodenumber.    cu,-   Martin -- -D                     | Martin Vorlaender    |    VMS & WNT programmer-   Smiert Spamionem  | work: mv@pdv-systeme.de D                     |       http://www.pdv-systeme.de/users/martinv/4                     | home: martin@radiogaga.harz.de   ------------------------------   Date: 9 Jan 2004 11:02:00 -0800 0 From: keith.cayemberg@conti.de (Keith Cayemberg)M Subject: Re: Well Andrew, "3" count them "3" security patches for VMS in fivea= Message-ID: <3a65a5c8.0401091049.199aa813@posting.google.com>M   > > ? > > listen Andrew, VMS security mup kits are rarely issued, and < > > don't confuse ucx flaws with VMS os and kernel flaws ... >  > F > Aah, that old chestnut.  Whenever you discuss security with VMS guysE > they always trot it out.  Does that mean we can take your figure ofaC > 1000+ solaris holes and cross off anything not in the kernel ? :)f  C Yes, but first redesign and rewrite your unix to cleanly catagorizen and separateF Kernel Mode from Supervisor Mode and from User Mode. Three modes are a minimum,D for a correct ring protection system. The use of three or more rings
 happens toC be a fully patented methodology by OpenVMS Engineering. OpenVMS hase four.uE OpenVMS also has 40 groups of higher mode functionality classified asa	 requiringt special named privileges.    And, then...  6  - allow access to higher mode services only through a DESCRIPTOR-basedD    calling standard which rules out "by design" the primary cause of securityE    holes - buffer-overflows. The secure Calling Standard is a centrale design    theme in OpenVMS.  F  - rewrite and install your TCP/IP stack so that it doesn't live in orB    directly access kernel mode services except through the calling	 standard. @    If the previous condition was met, your tcp/ip stack probablyD    won't work in Supervisor mode or User Mode without these changes.B    This is the reason why most security holes for which OpenVMS isA    affected does not in fact lead to a security vulnerability. IniF    this sense I agree with Andrew. Security vulnerability listings are9    innaccurate for OpenVMS. Because they do not correctly 
 differentiateeE    whether only a user-mode process can be affected or a higher mode,HD    and whether a higher privilege can be attained. A correct listing>    must rate the severity of the security hole. In OpenVMS the severityE    is usually lower (or meaningless) in comparison to other operatingv    systems.u  F  - design privilege assignments to be attached to a mode. If a programC    installed in a higher mode breaks out to a user-mode prompt. AllaC    privileges assigned during the program run must be automatically  lost.:C    This prevents program privilege tailgating. OpenVMS Hackers (yes8 they>    do exist, an admirably persistent if unsuccessful lot) have recently9    discovered this functionality in OpenVMS, inwhich they 
 intentionallyuF    installed an application with privileges and with a buffer overflow leading E    to a DCL prompt. Their experiment failed. This OpenVMS "knockdown"nA    functionality can also be extended to disable the privilege ofr?    receiving a DCL Prompt when breaking out of a program or DCLn
 procedure,E    just by assigning the CAPTIVE and RESTRICT flags to user accounts.d  @  - design your Unix to provide only strictly separated (and from overflow?    controlled) user and system stacks to prevent stack crashingP leadinga&    to access to higher mode functions.  F  - lets also not forget a redesign of the internal logon  mechanism to    be carried out by one  @ These are only a few of the unique, patented design decisions in OpenVMSeC resulting in a world-beating matrix of Functionality, Reliability,  D Availability, Security, Stability, and Scalability(RT, APMP, SMP andF Cluster). It's an OS that was "Designed" first by 4 competing teams of experts,E and then the best results of these competing design teams merged intoa at? final design team. They knew of the older Unix, MVS and Multicsn designs, andD naturally they innovated and improved on them for the Enterprise OS  problem space.  A When you are done making these elementary design changes to Unix -A (many of which were intentionally excluded or ignored by the Unix 	 designers F in 1969 - Multics already had early forms many of them) you will find A most of the commercial products on the Unix Market will no longern functionF correctly on your New-Unix, and will also require a redesign, and then
 a rewrite.  ? But at least you will finally have an OS and TCP/IP stack which @ "begins" to technically compare with OpenVMS within the frame of	 security.rF And you'll have a product which pays royalties to OpenVMS Engineering.  F Each OS has it's strengths and weaknesses in design and implementationE which will have a different evaluation depending on the problem spacep? it will be applied to, and depending on the design goals of thet
 designers.> For the general Enterprise OS problem space, I believe OpenVMS Engineering ; has most consistently made the best decisions in design andi implementeda? them with an admirably consistent high quality and methodology.   E OpenVMS enthusiasts can righteously bemoan that the Computer Science sA Profession (Informatics) have failed to recognize and teach theirp studentsC the sophisticated mechanisms and high principals found in OpenVMS,  D preferring instead to favoritize the minimalistic asthetics of Unix,E or the marketing level sophistication in OS selection. This is a real > loss for enterpise efficiency (money), mission-critical system	 stabilityoF (lives), and the computer science profession (maturity as a science). @ A more balanced and impartial framework of scientific thought is needed.HF Computer Science needs some independence from commercial and marketing> interests to even discover the value of many existing designs, technologiesD and ideas. The last major papers over OS design were written over 10/ years ago, but their work is far from complete.r  A Critics of OpenVMS should first study and compare it's internals  E (Professional OS comparisons and choices should not be reduced to an lA application layer beauty contest) with an open mind concering OS t> design paradigms, system operations principals and reliability methodologies.A After recovering from the shock, they will likely no longer be asm	 critical.m   Cheers!s   Keith Cayemberge) IBM Business Services - Hannover, Germanye   Semi-Nonstandard Disclaimer:3 Any non-official claims concerning my semi-official + opinions are hereby officially disclaimed. ,  i.e. I said it, not my employer.0 (and no I didn't steal this one from Yogi Berra)  ? I welcome rebuttal, however a lack of response on my part only h> indicates a lack discretionary time to indulge in discussions ' peripheral to my employment activities.S   ------------------------------   Date: 9 Jan 2004 16:13:49 -0800 0 From: keith.cayemberg@conti.de (Keith Cayemberg)M Subject: Re: Well Andrew, "3" count them "3" security patches for VMS in fivec< Message-ID: <3a65a5c8.0401091542.1df169b@posting.google.com>  E Excuse me. I just noticed I didn't finish writing the last condition.o It should read...l  F  - lets also not forget a redesign of the internal logon  mechanism toF    be carried out by one program/process first created at user request and =    has complete responsibility for the entire login sequence.   @ By the way, that was not by any means a complete list of OpenVMS, design advantages.  It was only a beginning.  u keith.cayemberg@conti.de (Keith Cayemberg) wrote in message news:<3a65a5c8.0401091049.199aa813@posting.google.com>...5 > > > A > > > listen Andrew, VMS security mup kits are rarely issued, and7> > > > don't confuse ucx flaws with VMS os and kernel flaws ... > >  > > H > > Aah, that old chestnut.  Whenever you discuss security with VMS guysG > > they always trot it out.  Does that mean we can take your figure ofmE > > 1000+ solaris holes and cross off anything not in the kernel ? :)d >   C Yes, but first redesign and rewrite your unix to cleanly catagorize0A and separate Kernel Mode from Supervisor Mode and from User Mode.sC Three modes are a minimum for a correct ring protection system. TheiE use of three or more rings happens to be a fully patented methodology D by OpenVMS Engineering. OpenVMS has four. OpenVMS also has 40 groupsB of higher mode functionality classified as requiring special named privileges.o   And, then...  7  - allow access to higher mode services only through a  5    DESCRIPTOR-based calling standard which rules out L5    "by design" the primary cause of security holes - r6    buffer-overflows. The secure Calling Standard is a #    central design theme in OpenVMS.   4  - rewrite and install your TCP/IP stack so that it 2    doesn't live in or directly access kernel mode 4    services except through the calling standard. If 5    the previous condition was met, your tcp/ip stack t7    probably won't work in Supervisor mode or User Mode d6    without these changes. This is the reason why most 9    security holes for which OpenVMS is affected does not o5    in fact lead to a security vulnerability. In this o5    sense I agree with Andrew. Security vulnerability 36    listings are innaccurate for OpenVMS. Because they ;    do not correctly differentiate whether only a user-mode e9    process can be affected or a higher mode, and whether A8    a higher privilege can be attained. A correct listing;    must rate the severity of the security hole. In OpenVMS O5    the severity is usually lower (or meaningless) in  )    comparison to other operating systems.3  :  - design privilege assignments to be attached to a mode. 9    If a program installed in a higher mode breaks out to =:    a user-mode prompt. All privileges assigned during the 9    program run must be automatically lost. This prevents  :    program privilege tailgating. OpenVMS Hackers (yes they:    do exist, an admirably persistent if unsuccessful lot) ;    have recently discovered this functionality in OpenVMS,  7    inwhich they intentionally installed an application A6    with privileges and with a buffer overflow leading :    to a DCL prompt. Their experiment failed. This OpenVMS 5    "knockdown" functionality can also be extended to 58    disable the privilege of receiving a DCL Prompt when 7    breaking out of a program or DCL procedure, just by  =    assigning the CAPTIVE and RESTRICT flags to user accounts.   7  - design your Unix to provide only strictly separated c9    (and from overflow controlled) user and system stacks e9    to prevent stack crashing leading to access to higher r    mode functions.  :  - lets also not forget a redesign of the internal logon  <    mechanism to be carried out by one program/process first ;    created at user request and has complete responsibility o!    for the entire login sequence.s  @ These are only a few of the unique, patented design decisions in= OpenVMS resulting in a world-beating matrix of Functionality,lC Reliability, Availability, Security, Stability, and Scalability(RT,hA APMP, SMP and Cluster). It's an OS that was "Designed" first by 4r> competing teams of experts, and then the best results of theseD competing design teams merged into a final design team. They knew ofE the older Unix, MVS and Multics designs, and naturally they innovatedo9 and improved on them for the Enterprise OS problem space.n  F When you are done making these elementary design changes to Unix (manyE of which were intentionally excluded or ignored by the Unix designersdE in 1969 - Multics already had early forms many of them) you will find"A most of the commercial products on the Unix Market will no longer5F function correctly on your New-Unix, and will also require a redesign, and then a rewrite..  ? But at least you will finally have an OS and TCP/IP stack whichl@ "begins" to technically compare with OpenVMS within the frame ofC security. And you'll have a product which pays royalties to OpenVMSe Engineering.  F Each OS has it's strengths and weaknesses in design and implementationE which will have a different evaluation depending on the problem spaceo? it will be applied to, and depending on the design goals of thehA designers. For the general Enterprise OS problem space, I believeiD OpenVMS Engineering has most consistently made the best decisions inE design and implemented them with an admirably consistent high qualityy and methodology.  D OpenVMS enthusiasts can righteously bemoan that the Computer ScienceA Profession (Informatics) have failed to recognize and teach theirsB students the sophisticated mechanisms and high principals found inD OpenVMS, preferring instead to favoritize the minimalistic astheticsD of Unix, or the marketing level sophistication in OS selection. ThisA is a real loss for enterpise efficiency (money), mission-criticale= system stability (lives), and the computer science professionoC (maturity as a science). A more balanced and impartial framework ofuF scientific thought is needed. Computer Science needs some independenceE from commercial and marketing interests to even discover the value oft# many existing designs, technologiesoD and ideas. The last major papers over OS design were written over 10/ years ago, but their work is far from complete.:  @ Critics of OpenVMS should first study and compare it's internalsD (Professional OS comparisons and choices should not be reduced to an@ application layer beauty contest) with an open mind concering OS> design paradigms, system operations principals and reliabilityC methodologies. After recovering from the shock, they will likely no  longer be as critical.   Cheers!(   Keith Cayemberg)) IBM Business Services - Hannover, Germanyh   Semi-Nonstandard Disclaimer:3 Any non-official claims concerning my semi-official + opinions are hereby officially disclaimed. t  i.e. I said it, not my employer.0 (and no I didn't steal this one from Yogi Berra)  ? I welcome rebuttal, however a lack of response on my part only y> indicates a lack discretionary time to indulge in discussions ' peripheral to my employment activities.    ------------------------------  * Date: Fri, 9 Jan 2004 20:32:21 +0000 (UTC)% From: bdc@world.std.com (Brian Chase)-@ Subject: Re: Where to get layered product software for hobbyist?( Message-ID: <btn34l$nsi$1@pcls4.std.com>  4 In article <3fff0a83$0$22331$626a54ce@news.free.fr>,' Didier Morandi  <info@vaxus.org> wrote:t > Brian Chase wrote:  K > > The inevitable next step of setting up my VMS system(s), beyond loadinglM > > the base OS install and the layered products provided on the hobbyist CD, K > > involves getting my hands on some distribution media for all the other s
 > > fun toys.i > G > Some of us here are considering to store the whole SPL on servers forlI > free or controlled access from the Internet, but the legal part of thiso$ > operation is not yet seen with HP.  G At the very least, it would be sensible to have those layered products i> for which the hobbyist license covers made available this way.   -brian.  -- eF --- Brian Chase | bdc@world.std.com | http://world.std.com/~bdc/ ----->       IBM 1130 Club Motto:  "Computer People Are Happy People"   ------------------------------  % Date: Fri, 09 Jan 2004 21:07:41 +0100i% From: Didier Morandi <info@vaxus.org>s@ Subject: Re: Where to get layered product software for hobbyist?4 Message-ID: <3fff0a83$0$22331$626a54ce@news.free.fr>   Brian Chase wrote:I > The inevitable next step of setting up my VMS system(s), beyond loadingiK > the base OS install and the layered products provided on the hobbyist CD,rI > involves getting my hands on some distribution media for all the other t > fun toys.r  N Some of us here are considering to store the whole SPL on servers for free or Q controlled access from the Internet, but the legal part of this operation is not   yet seen with HP.    D.
 www.vaxus.orgt   ------------------------------   End of INFO-VAX 2004.018 ************************