1 INFO-VAX	Wed, 31 Jul 2002	Volume 2002 : Issue 418       Contents:, Bad shadowing and SCSI bus error interaction Re: CGI/Perl on VMS 3 Encompass Board of Directors -- Call for Candidates 7 RE: Encompass Board of Directors -- Call for Candidates  FAQ?M Re: Fortran compiler was(RE: Dev tool docs on www.openvms.compaq.com:8000/ ?) P Re: Fortran compiler was(RE: Dev tool docs on www.openvms.compaq.com:8000/ ?) ?)+ Re: Itanic2 - the cHumPaq spin continues...  license prices - Advance Server  Re: Locked Tape$c device! No Andrew, slowaris can't be VMS! + Re: Only 20% drop in VMS systems (was: wow) + Re: Only 20% drop in VMS systems (was: wow) + RE: Only 20% drop in VMS systems (was: wow) + Re: Only 20% drop in VMS systems (was: wow) & PCMCIA network card support in OpenVMS# Re: PL/I Hobbyist kit now available # RE: PL/I Hobbyist kit now available  RDB Question.. reseller info requested  Re: reseller info requested % Re: Running an .EXE from a .COM file. % Re: Running an .EXE from a .COM file.  Re: Sort file protection  F ----------------------------------------------------------------------  % Date: Tue, 30 Jul 2002 18:51:28 -0400   From: John Santos <JOHN@egh.com>5 Subject: Bad shadowing and SCSI bus error interaction 5 Message-ID: <1020730175437.6055A-100000@Ives.egh.com>   > This situation has occurred about 3 times in the last 6 months= on one customer system, most recently at 1:00 this afternoon.   A They have 4 EZ832 solid-state disks, arranged as two shadow sets. = Each shadow set consists of one drive on each of 2 KZPCM SCSI  interfaces.   ; DSA84: consists of DKH100: and DKI100:, and DSA85: consists  of DKH400: and DKI400:.   4 CPU is a 4-processor ES40 (unsure of clock speed...)  @ They are getting "SCSI bus phase errors" (according to DECevent)@ and processes are simultaneously getting RMS-F-WER write errors.F Also at the same time, the disk is logging a "Power On, Reset, or Bus = Device Reset Occurred."  (I don't know what a "SCSI bus phase  error" actually means.)   @ All the errors seem to be on PKI0: and DKI100: (no errors logged= on the PKH/DKH path.)  I don't know if this is just chance or A if it is pointing specifically to DKI100: or its shelf.  (DKI100: # and DKI400: are on the same shelf.)     D The problem is two-fold.  1st, there is clearly something wrong with; the hardware.  My guess is it is a cable or connector issue = (maybe just a loose cable or terminator?), or flakey power in < the shelf (but why doesn't DKI400: have any problems?)  I've@ told them to call in field service and have them look at things.? (I also told them to do this in May, the last time it happened, 1 but I don't know if the did, or what they found.)   9 The 2nd problem is why doesn't volume shadowing hide this > error from the user apps?  Shouldn't it either declare DKI100:? dead, and drop it from the shadow set, or retry (which seems to . succeed, since the drive is running fine now)?    > This situation has occurred twice since the system was booted,C 163 days ago.  The PKI0: controller now shows 6 errors, and DKI100: : shows 4 errors.  It is possible there have been other SCSI> "events" when the system was idle, that didn't cause processes	 to abort.     8 VMS V7.2-1, possibly relevant ECO's are FIBRE_SCSI V5.0,7 SHADOWING V6.0, UPDATE V3.0, DRIVER V4.0, MOUNT96 V3.0, 6 and SYS V11.0.  There are later versions of several of6 these ECO's, but none of the descriptions seem to show: they would help.  (BTW, both FIBRE_SCSI V5.0 and SYS V11.08 caused problems that were fixed in subsequent ECO's, but6 Compaq provided workarounds at the time that have been9 installed.  FIBRE_SCSI left a file in the wrong directory 4 and we needed new versions of the SECURITY images to> prevent crashes when mounting NFS disks... Fixed in SYS V12.0)     --   John Santos  Evans Griffiths & Hart, Inc. 781-861-0670 ext 539   ------------------------------   Date: 31 Jul 2002 05:10:30 GMT1 From: JONESD@er6.eng.ohio-state.edu (David Jones)  Subject: Re: CGI/Perl on VMS: Message-ID: <ai7rg6$piu$1@charm.magnus.acs.ohio-state.edu>  0 In message <3D46C310.E32631A1@ceris.purdue.edu>,/   Chuck Aaron <caaron@ceris.purdue.edu> writes: G >It appears that when I get the "document contained no data..." message > >that checking the net$server.log file revealed the following: >  >Unoptimized perl...3 >cpu ticks: 49  Buffered I/O: 2051, Direct I/o: 158 3 >-------------------------------------------------- 4 >Connect request received at 30-jul-2002 10:08:18.77( >   from remote process Major::"0=HTTPD" >   for object "WWWEXEC"  J The first two lines are output by WWWEXEC.COM when handling a perl script.D The "Unoptimized perl..." message indicates it is using the standardC perl interpreter image rather the customized webperl interpreter to H run the script.   The second line is just the performance statistics forL the script execution, which includes both the perl execution and cgi_symbols
 execution.  H The last 4 lines are output by sys$system:netserver.exe when it receivesK a new connection request.  This request is a separate request (probably not E a perl request) from the request being reported in the first 2 lines.   F Your perl script is doing a couple of thousand writes.  If the browserJ reports 'no data', the server probably didn't like the CGI response headerH from the script and closed the connection.  The perl interpreter doesn'tC check for write failures and continued to output all data.  Set the E trace level on the server to 5 to see the CGI headers returned by the  script.       < David L. Jones               |      Phone:    (614) 292-6929- Ohio State University        |      Internet: L 140 W. 19th St. Rm. 231a     |               jonesd@er6s1.eng.ohio-state.edu: Columbus, OH 43210           |               vman+@osu.edu  1 Disclaimer: I'm looking for marbles all day long.    ------------------------------  % Date: Tue, 30 Jul 2002 15:30:53 -0500 / From: "Stuart, Ed" <Ed.Stuart@austinenergy.com> < Subject: Encompass Board of Directors -- Call for CandidatesT Message-ID: <92EFB80E551BD511B39500D0B7B0CDCC0642C6CD@ohms.electric.ci.austin.tx.us>  J Encompass is an independent association, consisting of a dynamic communityD of like-minded IT professionals with a vested interest in enterpriseJ computing and--more specifically--the products, services, and technologiesI of HP. Encompass is your industry advocate; the bridge between HP and its $ vendors and you, the customer/user.   J The Encompass Board of Directors currently has three vacancies and has putL out a call for candidates.  I have been selected to serve as a member of theI Nomination Committee and suggested to the committee that c.o.v would be a J good place to find candidates.  If you are interested in becoming a memberK of the Encompass Board of Directors please read the Call for Candidates web J page at: <http://www.encompassus.org/membership/bodcallforcandidates.html>F and submit a Candidate Application using the links found on that page.   EdE **Please apply a generous amount of all the usual disclaimers here.**    ------------------------------  % Date: Tue, 30 Jul 2002 13:47:23 -0700 # From: "Tom Linden" <tom@kednos.com> @ Subject: RE: Encompass Board of Directors -- Call for Candidates9 Message-ID: <CIEJLCMNHNNDLLOOGNJIOEKCFHAA.tom@kednos.com>   D That web page says that the application must be returned by July 26!   >-----Original Message----- 5 >From: Stuart, Ed [mailto:Ed.Stuart@austinenergy.com] % >Sent: Tuesday, July 30, 2002 1:31 PM  >To: Info-VAX@Mvb.Saic.Com= >Subject: Encompass Board of Directors -- Call for Candidates  >  > K >Encompass is an independent association, consisting of a dynamic community E >of like-minded IT professionals with a vested interest in enterprise K >computing and--more specifically--the products, services, and technologies J >of HP. Encompass is your industry advocate; the bridge between HP and its$ >vendors and you, the customer/user. > K >The Encompass Board of Directors currently has three vacancies and has put ? >out a call for candidates.  I have been selected to serve as a  >member of theJ >Nomination Committee and suggested to the committee that c.o.v would be aK >good place to find candidates.  If you are interested in becoming a member L >of the Encompass Board of Directors please read the Call for Candidates webK >page at: <http://www.encompassus.org/membership/bodcallforcandidates.html> G >and submit a Candidate Application using the links found on that page.  >  >Ed F >**Please apply a generous amount of all the usual disclaimers here.** >  >---' >Incoming mail is certified Virus Free. ; >Checked by AVG anti-virus system (http://www.grisoft.com). A >Version: 6.0.377 / Virus Database: 211 - Release Date: 7/15/2002  >  --- & Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.377 / Virus Database: 211 - Release Date: 7/15/2002   ------------------------------   Date: 31 Jul 2002 04:19:20 GMT7 From: International Jetsetter <freedomnow3344@clara.de> 
 Subject: FAQ?  Message-ID: <925c.4ad7.a8@YHBT>   C Hi! nice newsgroup. any rules I should know before I start posting?  is there an faq?   Thanks!    ------------------------------  % Date: Tue, 30 Jul 2002 15:01:56 -0400 + From: Steve Lionel <Steve.Lionel@intel.com> V Subject: Re: Fortran compiler was(RE: Dev tool docs on www.openvms.compaq.com:8000/ ?)8 Message-ID: <kaodku0k6hvkui7igesg0u6hcqh1agpifm@4ax.com>  C On Tue, 30 Jul 2002 19:21:23 +0010, paddy.o'brien@zzz.tg.nsw.gov.au  wrote:  J >Compaq before the seizure gave their Fortran compiler writers to Intel.  L >Fortran on VMS (possibly Tru64) has been put into maintenance mode.  To me 
 >this stinks.   F I don't speak for HP/Compaq any more (if in fact I ever did), but your? statement is overly broad.  HP compiler engineering owns Compaq C Fortran for VMS.  There is certainly active development for Fortran ? for Itanium VMS - how much of that will be applied to the Alpha E compiler, I don't know.  HP has engineers working on Fortran for VMS.   O >Starting from DEC, the Fortran compiler was considered one of the best ever.   K >Many of their proprietary enhancements became de-facto standards and were  J >mimicked in other compilers.  Steve Lionel was a very good spokesman for N >Fortran and VMS both here and in c.l.f.  He now works for Intel and produces E >CVF (obviously a money spinner), but soon to be subsumed as IVF (?).   E CVF is still a Compaq/HP product and is sold by them.  Yes, most (but C not all) of the engineers associated with that product now work for @ Intel, but this is a close partnership with HP and I suggest not) getting too bogged down in who owns what.         D Please send Visual Fortran support requests to vf-support@compaq.com   Steve Lionel Software Products Division Intel Corporation 
 Nashua, NH  : Intel Fortran for Windows and Compaq Visual Fortran forum:6   http://intel.forums.liveworld.com/forum.jsp?forum=76 Intel Fortran for Linux forum:7   http://intel.forums.liveworld.com/forum.jsp?forum=121    ------------------------------  % Date: Wed, 31 Jul 2002 09:47:20 +0010 % From: paddy.o'brien@zzz.tg.nsw.gov.au Y Subject: Re: Fortran compiler was(RE: Dev tool docs on www.openvms.compaq.com:8000/ ?) ?) 5 Message-ID: <01KKQNWLN9EA000SJ5@tgmail.tg.nsw.gov.au>    John Reagan wrote: >>  . >> Is CXML still in the hands of the new HP?   > E >CXML is very much still with us at HP.  You use it on OpenVMS?  Our  A >impressions are that CXML was used more in the Tru64 space, etc.  >   >Which CXML routines do you use?  K DGEEV, DGEHRD, DORGHR are the main ones, but DGEEV itself is effectively a  . package of nearly 50 BLAS and LAPACK routines.  E We had previously used equivalent NAG routines, but when CXML became  K available with the compiler, I swapped to using that directly and used the  N NAG alternative library which excludes their versions of BLAS and LAPACK.  We C still use many NAG routines which themselves call some of the BLAS.   K I have been looking at using some of the other "packages" within CXML, but  9 here the problem of future portability might be an issue.   I The reason I mentioned this in my post was that the last time I had some  K correspondence with Jeff Arnold, the LAPACK routines are a level down from  L what is current. I was interested in the feedback feature regarding optimal J workspace size.  Hopefully workspace will become internal to the routines < anyway eventually, as I have now done with our applications.   Regards, Paddy   ------------------------------  # Date: Tue, 30 Jul 2002 19:42:29 GMT 5 From: "Fred Kleinsorge" <kleinsorge@star.zko.dec.com> 4 Subject: Re: Itanic2 - the cHumPaq spin continues...2 Message-ID: <F_B19.23$qo3.750693@news.cpqcorp.net>  K To be honest, I "try" to not even see Bill's posts, but with bottom posting L including other replies, it becomes very hard to do.  I will resist a little harder.     $ Larry Kilgallen wrote in message ...< >In article <3D42CEC5.326CD70A@fsi.net>, "David J. Dachtera" <djesys.nospam@fsi.net> writes:  > F >> I know we've had words here, also. So, I must be honest: you'd haveE >> garnered more credibility (does that word hold value for you?) for K >> yourself had you addressed Bill's comments about trust and trust broken.  > D >Whereas _my_ reaction is that Fred reduces his usefulness everytimeC >he clutters my newsreader responding to Bill Todd, whose remarks I $ >would otherwise not have to suffer.   ------------------------------  % Date: Tue, 30 Jul 2002 14:24:34 -0400 0 From: "Hank Vander Waal" <hvanderw@novagate.com>( Subject: license prices - Advance Server; Message-ID: <000d01c237f6$5bb0ef80$cd96a8c6@manufact5l8vs8>   F Finally getting our Pathworks server upgraded to v 7/3 Advanced ServerK I have found the part # for the license upgrade that I need QM-5SUAA-CC but I can not find any place to order it on the web.  I found where I can order C the new license but can not find pricing or the ability to order it 2 anywhere.   Is the only place to order via phone??  ; I was hoping to avoid calling anyone if I did not have to .    ------------------------------  # Date: Wed, 31 Jul 2002 00:37:36 GMT ! From: rob.buxton@wcc.spam.govt.nz ! Subject: Re: Locked Tape$c device $ Message-ID: <3d472ff4.88548816@news>  / On Tue, 30 Jul 2002 12:04:47 -0500, Chuck Aaron   <caaron@ceris.purdue.edu> wrote:   >Bob,  > 7 >Thank you for your reply.  The tape is in a state that 6 >can't be aborted by the driver. Looks like a shutdown >and boot to me. >  >Thanks to all,  >Chuck   Chuck,  C I've seen the problem before and have used the following to free up  the device. C Used Availability Manager / AMDS to kill the process. Not sure what * this did differently but the process dies.< The Tape Drive remains allocated to the now deleted process.D There's an image available from HP that can remove this, the details  from the image header are below.(         Image Identification Information  E                 image name: "CSCPAT_0245_USE_WITH_CAUTION_MTDISMOUNT" <                 image file identification: "CSC/CS MTDMT V2"7                 link date/time: 27-OCT-1993 07:06:28.58 .                 linker identification: "05-13"  *   This now clears the owner process field.5 This may not help this time, but might in the future,  Rob.   >  >Bob Koehler wrote:  >>  a >> In article <3D469F8F.138C9F2C@ceris.purdue.edu>, Chuck Aaron <caaron@ceris.purdue.edu> writes:  >> > Group,  >> >8 >> > My batch job terminated with the following message:A >> > %SYSTEM-F-FORCEDEXIT, forced exit of image or process by job  >> > controller  >> >K >> > It shows the job still running with the tape in the slot and wonder if G >> > there is a way to reset the tape$c drive. I have tried to dismount  >> > tape$c and it tells me: >> >A >> > %SYSTEM-W-DEVALLOC, device already allocated to another user  >> >D >> > I checked the system and there is a process running that I have >> > stopped >> >< >> > Copy    _TNA6:        00000149  TNA6:    (disconnected) >> > >>  G >>    It's probably OK by now.  Lot's of times the process can't finish J >>    running down until the I/O completes or aborts.  In the case of someJ >>    tape controllers the timeout is justifiably long, but eventually theI >>    I/O will either complete or timeout.  Then the process can run down & >>    and the tape drive becomes free. >> iE >>    Often a prvileged user can convince the system to abort the I/OfD >>    via dismount/abort.  Since that didn't work for you either youE >>    don't have sufficient privilege, or the tape is in a state thatp% >>    can't be aborted by the driver.w   ------------------------------    Date: 30 Jul 2002 19:24:28 -0700( From: bob@instantwhip.com (Bob Ceculski)* Subject: No Andrew, slowaris can't be VMS!= Message-ID: <d7791aa1.0207301824.20ed36a3@posting.google.com>   2 sorry Andrew, no other os can cluster like VMS ...     HP takes aim at Sun clustering    ( Claims Sun fails to deliver technologies  + By Mike Magee: Tuesday 30 July 2002, 17:08 r  C A DOCUMENT SEEN the INQUIRER shows that HP is taking aim at Sun andbA accusing it of not delivering on promises it made about clusters. C The document, called a "competitive briefing", is aimed at partners C and obviously is intended to show HP in a good light and put Sun in-
 the shade.  @ The document claims HP has competitive intelligence that Sun had@ promised its customers they would have a 32-64 CPU shared memory cluster machine.  E This, said HP, would be based on a high speed interconnect giving the0C look and feel of a single Solaris copy across multiple nodes, usinge! multiple 12 way Sun Fire servers.t  B HP claims Sun had told its customers they would have such machines@ installed by now, but has been unable to deliver. Instead it has@ substituted Sun Fire 1500s to satisfy the contracts, HP alleges.  E And HP also claims that Sun said it was talking about a shared memory0' cluster which it also couldn't deliver.   F The document asks partners to find out as much about what it dubs as aC "Sun fiasco" as possible. Obviously, for competitive purposes only.e   ------------------------------  % Date: Tue, 30 Jul 2002 18:22:34 +0100 U From: Andrew Harrison SUNUK Consultancy <andrew_nospam.harrison_remove_this@sun#.com>e4 Subject: Re: Only 20% drop in VMS systems (was: wow)0 Message-ID: <ai6i11$jrc$1@new-usenet.uk.sun.com>   jlsue wrote:  G > On Mon, 29 Jul 2002 18:04:26 +0100, Andrew Harrison SUNUK Consultancy 6 > <andrew_nospam.harrison_remove_this@sun#.com> wrote: >  >  >> >>jlsue wrote: >> >>H >>>BTW, re-read the thread I responded to.  You fired the first salvo of >>>abuse.  Grow up.m >>>j >>>  >>0 >>Where did I abuse you in the previous thread ? >>8 >>If you can find it then you are a much more sensitive  >> >>soul than you appear to be.h >> > % > Ah, yes.  Re-read your responses ind( > <ah6cf6$6vi$1@new-usenet.uk.sun.com>.  > " >  In the original response I made@ > <27jmiuo83hh1j33ri0egsna0e6vomg1i50@4ax.com>, I was counteringE > specific points you made.  There was no request in there concerning:@ > specific performance/scalability requirements.  I address yourH > specific points with as much information as they call for, and in yourD > response (the first one mentioned above) you can't seem to containB > yourself enough to stay on topic.  Instead, you feel the need toF > attack me personally.  Then what?  Now you're standing around saying
 > "Who?  Me?"e >  eE > I work in the real world.  I know what customers are demanding fromIF > their applications and environments, and it's not always the perfect > solution.r >  >     5 So in your world disagreeing with the points that yous3 have made counts as a personal insult but resortingF. to phrases such as grow up and worse does not.  0 When you can find a personal attack in there get1 back to me at the moment all you can accuse me ofe. is rubishing your claims, not pleasant for you$ I know but hardly a personal insult.  / If you advance and argument that is contentiousF1 like for example you "support" claims then expecto/ it to get attacked and don't get upset about itt/ that would be childish now wouldn't it and I am10 sure that isn't the impression you want to give.    D >>You still seem terminally incapable of answering the question, yesC >>clustering gives you N+1 resiliance, that we know. What you don't3D >>seem to want to answer is what scalability you got. Was it 2x when9 >>moving from 1 to 2 nodes, was it less, how much was it.d >> > E > In some cases it was very close to 2x.  In others, not so much.  It:G > really is dependent on the environment.  But performance alone is noto > the only scalability issue.- >  > ? >>The whole discussion was until you lept in flailing about the:8 >>work involved in acheiving scalability in a cluster or< >>using a large SMP machine. Nothing that you have posted so: >>far illuminates this discussion with anything that could >>be described as a fact.s >> > H > No, you're wrong.  Re-look at the specific response I made.  There wasE > nothing in there specifically about measuring scalability to some %i& > accuracy.  In fact, the topics were: >  > 1.	Training requirements > 2.	Tuning  > 3.	Financial Penalties > 4.	Storage requirementse) > 5.	Apps support of cluster environmentsn >     1 You are mistaking your response, which was as youl7 say the above with the topic actually under discussion.n  . You skirted arround the issue. But don't be to- despondant, your responses to cluster supporte. issues were highly revealing though not as you) belatedly realised for the right reasons.n    H > Not a one was scalability vs SMP.  I do not claim that VMScluster willB > provide the same performance scalability as a larger SMP system. >  >     6 Quite and the first usefull response you have provided4 because it differs markedly from your collegue Kerry$ who appears not to share your views.  2 If you remember Kerry was the one claiming that it) was much of a muchness clustering vs SMP.o     >  > F >>>Well, this isn't a sales pitch.  I don't have to be exact.  This isG >>>merely a topical, informal discussion.  If we get into anything that,) >>>requires more exact data, let me know., >>>< >>>u >>4 >>Well if you arn't going to be exact or even mildly2 >>inexact then what did you think you contribution2 >>was going to add to the discussion. All you have2 >>managed to do so far is maintain what is for you >>the status quo.n >> > H > I was addressing your specific points in one particular posting.  Your< > status quo seems to be, if you can't counter with anything% > intelligent, throw insults instead.o >     , Since when is a statement of fact an insult.  - Fact the discussion was about scalability andp1 the relative difficulty of getting it in Clusterss. or SMP systems and thus their suitability as a" platform for server consolidation.  / Fact you hadn't at that point provided anythingo1 quantative appart from statements like wonderfullc etc.  3 Go back and re-read your previous responses in thise1 thread, they hardly add up to anything usefull dof0 they. Your comments on support, for example very% helpfull in an entirely negative way.h     >  > C >>>No.  If I said that, please explain where.  Servers are servers,dG >>>workstations are (generally) not.  We (the company I worked for, not H >>>DEC/CPQ/HP)  were using workstations as servers in this one instance. >>>m >>>m >>< >>So Digital charged ludicrous license fees for "servers" so: >>its was sensible to use similar capacity but licenced as/ >>workstations systems which were much cheaper.t >> > F > Oh yeah, in the previous comment I forgot one other "status quo" forE > you:  If you can't offer anything of value, pretend to restate youryC > opponents' view in YOUR words (i.e., *not* mine).  This helps youp9 > appear to make a point, even if it *is* make-up-a-fact.c > D > First off, I didn't say they were "similar capacity".  I said theyE > were similar CPU speed.  Certainly you could get many more users on B > one VAX 6000-610 than on a VAXstation Model 90, though they wereE > similar CPU speed.  However, cluster a few of the M90s together ande/ > you had a decent cluster at a bit lower cost.. >     ? Why clearly, it isn't clear at all. They had the same CPU speed C what else differed between them that makes it clear that a 6000-610a< would be have much more capacity than a VAXstation Model 90.      H > And, I'm not saying that Digital's software licensing wasn't ludicrousF > either... ;-)  However, If a Sun workstation has a similar CPU speed@ > as a server, are you telling me that software licensing on theG > workstation is sane enough compared to the server that it would nevertA > make sense to forego high-end server costs by using clusters ofe > smaller workstations?       @ Not for licensing reasons. As with most vendors the per CPU cost? is lowest in the smallest systems so if you have a horizontally.= scalable application like say a web service it makes sense tos= buy say 8 V120's rather than a 4 CPU V480 (the V120 CPU's areF7 slower). But you don't need to do this for license costt reasons.       > F > I don't know how clustering works in the Sun world, but, truly, in aE > VMScluser world, the managment and workload balancing can sometimese( > make it a very attractive alternative. >     > As it can in the Sun world but only for a limitted set of apps8 where for example you don't get penalised for being in a< cluster in the first place, or where the app is horizontally5 scalable like a web service or some kinds of HPC app..     Regardsc Andrew Harrison:   ------------------------------  % Date: Tue, 30 Jul 2002 22:08:03 +0200 9 From: Jan-Erik =?iso-8859-1?Q?S=F6derholm?= <aaa@aaa.com> 4 Subject: Re: Only 20% drop in VMS systems (was: wow)' Message-ID: <3D46F222.26863FFD@aaa.com>t  ( Andrew Harrison SUNUK Consultancy wrote: > A > Why clearly, it isn't clear at all. They had the same CPU speedtE > what else differed between them that makes it clear that a 6000-610 > > would be have much more capacity than a VAXstation Model 90.   Have you seen them ?  A I could more or less just get half my arm into a MV3100/90, whilemB the 6000 box (the "refrigerator" typ of box) easily could take oneC or perhaps two of the smaller type of users if you took out all theF stuff inside...   = Seariously, it's the same reason the Grayhound Bus takes moret@ passagers then some popular sportscar even if they have the same4 amount of h.p. It's the storage space and the I/O...       Jan-Erik Sderholm.e   ------------------------------  % Date: Tue, 30 Jul 2002 17:34:22 -0400r' From: "Main, Kerry" <Kerry.Main@hp.com>l4 Subject: RE: Only 20% drop in VMS systems (was: wow)T Message-ID: <BE56C50EA024184DAF48F0B9A47F5CF4026608AC@kaoexc01.americas.cpqcorp.net>   Andrew,i  D <<< Quite and the first useful response you have provided because itG differs markedly from your collegue Kerry who appears not to share youro	 views.>>>   - Ahh Andrew .. Up to your old tricks again eh?e   :-)v  H Take what one person says and then twist it around to use as fud against
 someone else.a  D Clustering vs SMP scalability and availability discussions can go on forever.=20r  E Bottom line is that it depends on the application(s) and both SMP andtB Clustering have scalability issues and considerations one needs to	 consider.-  D You can cite examples where SMP is better for performance reasons (IF might even agree) and I can cite area's where clustering is better.=20  C Give it up - this is not something you are going to make any groundg with.=20   Regards,  
 Kerry Main Senior Consultant  Hewlett-Packard Canada! Consulting & Integration Services  Voice: 613-592-4660  Fax   : 613-591-4477 Email: Kerry.Main@hp.com     -----Original Message-----' From: Andrew Harrison SUNUK Consultancyo7 [mailto:andrew_nospam.harrison_remove_this@sun#.com]=20o Sent: July 30, 2002 1:23 PMc To: Info-VAX@Mvb.Saic.Coml4 Subject: Re: Only 20% drop in VMS systems (was: wow)         jlsue wrote:  J > On Mon, 29 Jul 2002 18:04:26 +0100, Andrew Harrison SUNUK Consultancy=206 > <andrew_nospam.harrison_remove_this@sun#.com> wrote: >=20 >=20 >> >>jlsue wrote: >> >>H >>>BTW, re-read the thread I responded to.  You fired the first salvo of   >>>abuse.  Grow up.  >>>n >>>u >>0 >>Where did I abuse you in the previous thread ? >>7 >>If you can find it then you are a much more sensitive  >> >>soul than you appear to be.o >> >=20( > Ah, yes.  Re-read your responses in=20' > <ah6cf6$6vi$1@new-usenet.uk.sun.com>.a >=20% >  In the original response I made=20sC > <27jmiuo83hh1j33ri0egsna0e6vomg1i50@4ax.com>, I was countering=20eH > specific points you made.  There was no request in there concerning=20C > specific performance/scalability requirements.  I address your=20 H > specific points with as much information as they call for, and in your  G > response (the first one mentioned above) you can't seem to contain=20-E > yourself enough to stay on topic.  Instead, you feel the need to=20pI > attack me personally.  Then what?  Now you're standing around saying=20b
 > "Who?  Me?"  > =20nH > I work in the real world.  I know what customers are demanding from=20I > their applications and environments, and it's not always the perfect=207 > solution." >=20 >=20    5 So in your world disagreeing with the points that youv3 have made counts as a personal insult but resortinga. to phrases such as grow up and worse does not.  0 When you can find a personal attack in there get1 back to me at the moment all you can accuse me of . is rubishing your claims, not pleasant for you$ I know but hardly a personal insult.  / If you advance and argument that is contentiousa1 like for example you "support" claims then expect-/ it to get attacked and don't get upset about it0/ that would be childish now wouldn't it and I amn0 sure that isn't the impression you want to give.    G >>You still seem terminally incapable of answering the question, yes=20iH >>clustering gives you N+1 resiliance, that we know. What you don't seem  I >>to want to answer is what scalability you got. Was it 2x when moving=20 2 >>from 1 to 2 nodes, was it less, how much was it. >> >=20H > In some cases it was very close to 2x.  In others, not so much.  It=20J > really is dependent on the environment.  But performance alone is not=20 > the only scalability issue.o >=20 >=20G >>The whole discussion was until you lept in flailing about the work=20 H >>involved in acheiving scalability in a cluster or using a large SMP=20B >>machine. Nothing that you have posted so far illuminates this=20= >>discussion with anything that could be described as a fact.  >> >=20H > No, you're wrong.  Re-look at the specific response I made.  There was  H > nothing in there specifically about measuring scalability to some %=20& > accuracy.  In fact, the topics were: >=20 > 1.	Training requirements > 2.	Tuning  > 3.	Financial Penalties > 4.	Storage requirementsi) > 5.	Apps support of cluster environmentsf >=20    1 You are mistaking your response, which was as yout7 say the above with the topic actually under discussion.e  . You skirted arround the issue. But don't be to- despondant, your responses to cluster supporti. issues were highly revealing though not as you) belatedly realised for the right reasons.T    H > Not a one was scalability vs SMP.  I do not claim that VMScluster will  B > provide the same performance scalability as a larger SMP system. >=20 >=20    6 Quite and the first usefull response you have provided4 because it differs markedly from your collegue Kerry$ who appears not to share your views.  2 If you remember Kerry was the one claiming that it) was much of a muchness clustering vs SMP.s     >=20 >=20I >>>Well, this isn't a sales pitch.  I don't have to be exact.  This is=20TJ >>>merely a topical, informal discussion.  If we get into anything that=20) >>>requires more exact data, let me know.i >>>e >>>  >>4 >>Well if you arn't going to be exact or even mildly2 >>inexact then what did you think you contribution2 >>was going to add to the discussion. All you have2 >>managed to do so far is maintain what is for you >>the status quo.s >> >=20H > I was addressing your specific points in one particular posting.  Your  ? > status quo seems to be, if you can't counter with anything=20s% > intelligent, throw insults instead.e >=20    , Since when is a statement of fact an insult.  - Fact the discussion was about scalability and 1 the relative difficulty of getting it in Clustersu. or SMP systems and thus their suitability as a" platform for server consolidation.  / Fact you hadn't at that point provided anything>1 quantative appart from statements like wonderfully etc.  3 Go back and re-read your previous responses in this 1 thread, they hardly add up to anything usefull dof0 they. Your comments on support, for example very% helpfull in an entirely negative way.w     >=20 >=20F >>>No.  If I said that, please explain where.  Servers are servers,=20G >>>workstations are (generally) not.  We (the company I worked for, notsH >>>DEC/CPQ/HP)  were using workstations as servers in this one instance. >>>o >>>o >>G >>So Digital charged ludicrous license fees for "servers" so its was=20nJ >>sensible to use similar capacity but licenced as workstations systems=20 >>which were much cheaper. >> >=20F > Oh yeah, in the previous comment I forgot one other "status quo" forH > you:  If you can't offer anything of value, pretend to restate your=20F > opponents' view in YOUR words (i.e., *not* mine).  This helps you=209 > appear to make a point, even if it *is* make-up-a-fact.m >=20G > First off, I didn't say they were "similar capacity".  I said they=20lH > were similar CPU speed.  Certainly you could get many more users on=20E > one VAX 6000-610 than on a VAXstation Model 90, though they were=20sH > similar CPU speed.  However, cluster a few of the M90s together and=20/ > you had a decent cluster at a bit lower cost.s >=20    D Why clearly, it isn't clear at all. They had the same CPU speed whatG else differed between them that makes it clear that a 6000-610 would be 3 have much more capacity than a VAXstation Model 90.a      H > And, I'm not saying that Digital's software licensing wasn't ludicrous  I > either... ;-)  However, If a Sun workstation has a similar CPU speed=20"C > as a server, are you telling me that software licensing on the=20-J > workstation is sane enough compared to the server that it would never=20D > make sense to forego high-end server costs by using clusters of=20 > smaller workstations?0    C Not for licensing reasons. As with most vendors the per CPU cost is.E lowest in the smallest systems so if you have a horizontally scalablewE application like say a web service it makes sense to buy say 8 V120's#H rather than a 4 CPU V480 (the V120 CPU's are slower). But you don't need$ to do this for license cost reasons.       >=20I > I don't know how clustering works in the Sun world, but, truly, in a=20uH > VMScluser world, the managment and workload balancing can sometimes=20( > make it a very attractive alternative. >=20    H As it can in the Sun world but only for a limitted set of apps where forC example you don't get penalised for being in a cluster in the firstvF place, or where the app is horizontally scalable like a web service or some kinds of HPC app.     Regards  Andrew Harrisonn   ------------------------------    Date: 30 Jul 2002 19:21:50 -0700( From: bob@instantwhip.com (Bob Ceculski)4 Subject: Re: Only 20% drop in VMS systems (was: wow)< Message-ID: <d7791aa1.0207301821.1472742@posting.google.com>   "Main, Kerry" <Kerry.Main@hp.com> wrote in message news:<BE56C50EA024184DAF48F0B9A47F5CF4026608AC@kaoexc01.americas.cpqcorp.net>... 	 > Andrew,r > F > <<< Quite and the first useful response you have provided because itI > differs markedly from your collegue Kerry who appears not to share your  > views.>>>s > / > Ahh Andrew .. Up to your old tricks again eh?a >  > :-)l > J > Take what one person says and then twist it around to use as fud against > someone else.s > F > Clustering vs SMP scalability and availability discussions can go on > forever. g > G > Bottom line is that it depends on the application(s) and both SMP andfD > Clustering have scalability issues and considerations one needs to > consider.m > F > You can cite examples where SMP is better for performance reasons (IF > might even agree) and I can cite area's where clustering is better.  > E > Give it up - this is not something you are going to make any grounde > with.  > 
 > Regards, >  > Kerry Main > Senior Consultanto > Hewlett-Packard Canada  @ of course, what else can he do when sun can't offer a clusteringA solution ... they promised it, but they couldn't deliver, becausen> no other os can cluster the way vms does ... of course sun hasA plenty of smp ... thats the only way they can even try to show upe, on the performance screen againset alpha ...   ------------------------------  % Date: Tue, 30 Jul 2002 17:39:16 -0600k2 From: David Richardson <DRichardson@interbaun.com>/ Subject: PCMCIA network card support in OpenVMSe, Message-ID: <3D46CF44.DF8B0F8@interbaun.com>  2 Is any PCMCIA network card supported in OpenVMS?    H I have a hobbyist Alphastation 200 4/233 (7.2-2 currently) with a PCMCIAE slots in the back and would like to add a second network card.  I wasm? hoping that something like the D-Link 660+ series would work.  e  - Does any card work, without being supported? s   -- SF Microsoft:  The "Would you like fries with that" of the IT profession.   David Richardson DRichardson@interbaun.com    ------------------------------   Date: 30 Jul 2002 23:38:13 GMT2 From: "Zane H. Healy" <healyzh@shell1.aracnet.com>, Subject: Re: PL/I Hobbyist kit now available, Message-ID: <ai781501qki@enews2.newsguy.com>  " Tom Linden <tom@kednos.com> wrote:< > We have made fully functional versions of the compiler for> > bot VAX and Alpha which repsect the OPENVMS-HOBBYIST license= > paks.  You may downlad the kits from freja.kednos.com using # > anonymous ftp or ftp web browser.t  7 I'm getting the following on the AXP files on the site:     J DPLI043.A;2                   <%SYSTEM-F-NOPRIV, insufficient privilege or objec  t protection violation>SJ DPLI043.B;2                   <%SYSTEM-F-NOPRIV, insufficient privilege or objeci t protection violation>,J DPLI043.C;2                   <%SYSTEM-F-NOPRIV, insufficient privilege or objecl t protection violation> J DPLI043.D;2                   <%SYSTEM-F-NOPRIV, insufficient privilege or objeco t protection violation> J DPLI043H.ZIP;1                   <%SYSTEM-F-NOPRIV, insufficient privilege or oba ject protection violation>  , At least I think these are the needed files.   		Zane   ------------------------------  % Date: Tue, 30 Jul 2002 17:48:59 -0700c# From: "Tom Linden" <tom@kednos.com>t, Subject: RE: PL/I Hobbyist kit now available9 Message-ID: <CIEJLCMNHNNDLLOOGNJIEEKGFHAA.tom@kednos.com>b  H My fault, I'll fix the protections shortly.  All you need however is theG .ZIP file.  The other files are only there in case somebody didn't haved the ability to UNZIP.'   Toma   >-----Original Message-----n8 >From: Zane H. Healy [mailto:healyzh@shell1.aracnet.com]% >Sent: Tuesday, July 30, 2002 4:38 PMe >To: Info-VAX@Mvb.Saic.Com- >Subject: Re: PL/I Hobbyist kit now availablec >e > # >Tom Linden <tom@kednos.com> wrote:l= >> We have made fully functional versions of the compiler forh? >> bot VAX and Alpha which repsect the OPENVMS-HOBBYIST license > >> paks.  You may downlad the kits from freja.kednos.com using$ >> anonymous ftp or ftp web browser. >e8 >I'm getting the following on the AXP files on the site: >d >dK >DPLI043.A;2                   <%SYSTEM-F-NOPRIV, insufficient privilege ori >objec >t protection violation>K >DPLI043.B;2                   <%SYSTEM-F-NOPRIV, insufficient privilege ort >objec >t protection violation>K >DPLI043.C;2                   <%SYSTEM-F-NOPRIV, insufficient privilege or  >objec >t protection violation>K >DPLI043.D;2                   <%SYSTEM-F-NOPRIV, insufficient privilege ora >objec >t protection violation>K >DPLI043H.ZIP;1                   <%SYSTEM-F-NOPRIV, insufficient privilegee >or ob >ject protection violation>e >e- >At least I think these are the needed files.  >e >		Zaneu >t >---' >Incoming mail is certified Virus Free.r; >Checked by AVG anti-virus system (http://www.grisoft.com).aA >Version: 6.0.377 / Virus Database: 211 - Release Date: 7/15/2002h >h ---e& Outgoing mail is certified Virus Free.: Checked by AVG anti-virus system (http://www.grisoft.com).@ Version: 6.0.377 / Virus Database: 211 - Release Date: 7/15/2002   ------------------------------    Date: 30 Jul 2002 15:39:21 -0700- From: contracer11@uol.com.br (Shiva MahaDeva)s Subject: RDB Question..s= Message-ID: <ddf392ea.0207301439.29f218d1@posting.google.com>k  * In my job, many times I issue the command:  $ RMU/SH STATUS dbrf0301    and I press D, T, A keys to see locked processl  5 My question: Can I put this in a procedure ? Example:s    $ RMU/SH STATUS DBRF030  D  T  A  : I make this procedure, but it doesnt work. Any sugestion? Thanks in advance...   ------------------------------  # Date: Tue, 30 Jul 2002 23:01:09 GMTs. From: "Tom Simpson" <simpsont@attbi.com.fubar>  Subject: reseller info requested9 Message-ID: <VUE19.449$Ru5.266@rwcrnsc52.ops.asp.att.net>h  E Has anyone had any recent experiences with purchases from CPU Options,= (a reseller/broker out of MN.) that they would care to share?    Thanks,a Tom    ------------------------------    Date: 30 Jul 2002 19:28:00 -0700( From: bob@instantwhip.com (Bob Ceculski)$ Subject: Re: reseller info requested= Message-ID: <d7791aa1.0207301827.39a11099@posting.google.com>e  o "Tom Simpson" <simpsont@attbi.com.fubar> wrote in message news:<VUE19.449$Ru5.266@rwcrnsc52.ops.asp.att.net>... G > Has anyone had any recent experiences with purchases from CPU Options6? > (a reseller/broker out of MN.) that they would care to share?m > 	 > Thanks,e > TomD  ? we have bought alphas from them in the past ... good company to 
 deal with ...h   ------------------------------   Date: 31 Jul 2002 05:03:07 GMT- From: djweath@attglobal.net (Dave Weatherall)a. Subject: Re: Running an .EXE from a .COM file.5 Message-ID: <DTiotGxQ0bj6-pn2-OpwoXXePQDb8@localhost>m  E On Mon, 29 Jul 2002 09:33:05 UTC, robert_kersey@bat.com (Rob Kersey)   wrote:  G > I am trying to run an .exe from within a .com file. The .exe requireso@ > input back from the user before it exits on the users request. > H > The problem that I am having is that the .com file does not pause whenE > it gets to the running of the .exe unless I use SPAWN/WAIT. However.H > the .com file is being run from within a captive account and therefore > no spawning is allowed.- > H > Is there another way to run an .exe and get the com file to pause/wait > before it continues.   Hi Rob9                 you've received lots of good suggestions.4  E Most of my programming is in Fortran so when I embed a Fortran based u" image in a .COM file I usually use  ! $ DEFINE /USER FOR005 SYS$COMMANDa  F The advantage is that I can split READ (5,*) statements and ACCEPTs ifB I need to. ACCEPT reads directly from SYS$COMMAND (IIRC). There's 7 another advantage too but it escapes me for the moment.d   The /USER does its usual job.e  F I have one big .COM which has many .EXE's so there I DEFINE at the top and DEASSIGN on exit.a   -- p Cheers - Dave.   ------------------------------  % Date: Wed, 31 Jul 2002 15:45:33 +0010e% From: paddy.o'brien@zzz.tg.nsw.gov.au . Subject: Re: Running an .EXE from a .COM file.5 Message-ID: <01KKR0FP8CXU000SVO@tgmail.tg.nsw.gov.au>'    Dave Weatherall wrote (in part):  F >Most of my programming is in Fortran so when I embed a Fortran based # >image in a .COM file I usually user >n" >$ DEFINE /USER FOR005 SYS$COMMAND >lG >The advantage is that I can split READ (5,*) statements and ACCEPTs if C >I need to. ACCEPT reads directly from SYS$COMMAND (IIRC). There's w8 >another advantage too but it escapes me for the moment. >r >The /USER does its usual job. >yG >I have one big .COM which has many .EXE's so there I DEFINE at the topp >and DEASSIGN on exit.  H I would still be inclined to use a DEFINE/USER (or ASSIGN, depending on = personal preference) before each .EXE (some may not need it).i  M My reason is that if one of the programs "crashes" and aborts the whole job, hK in your case, the logical will still be defined, whereas in mine, not.  It >M depends what other safeguards you have for a graceful exit from your command nK file to clean these things up.  And, if this is the only thing you need to eJ clean up, /user would give simpler code than status checking to go to the  DEASSIGN statement.    Regards, Paddy   ------------------------------    Date: 30 Jul 2002 13:28:57 -0700" From: cstranslations@msn.com (Joe)! Subject: Re: Sort file protectione= Message-ID: <d56d1c2d.0207301228.5205834c@posting.google.com>y  g cstranslations@msn.com (Joe) wrote in message news:<d56d1c2d.0207291342.4b6a5c87@posting.google.com>...r  E Solved. Somewhere between when they (the company I work for) bought a2D VMS system and when I got here someone (probably not sure what (s)heB was doing) decided on the above nifty definitions for the SORTWORKF logicals. They get in the way of one another (they're suppose to point to a device only).  A For some reason there isn't any problem on our 7.1-1H2 productionoF system. We just moved out test system to 7.3 and something is "gettingC stuck in the works." What, how, and why I'll save for another time.l   Joe    ------------------------------   End of INFO-VAX 2002.418 ************************onal attack in the                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                 
    
    
    
    
    
    
    
    
    	
    

    
    
    

    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
     
    !
    "
    #
    $
    %
    &
    '
    (
    )
    *
    +
    ,
    -
    .
    /
    0
    1
    2
    3
    4
    5
    6
    7
    8
    9
    :
    ;
    <
    =
    >
    ?
    @
    A
    B
    C
    D
    E
    F
    G
    H
    I
    J
    K
    L
    M
    N
    O
    P
    Q
    R
    S
    T
    U
    V
    W
    X
    Y
    Z
    [
    \
    ]
    ^
    _
    `
    a
    b
    c
    d
    e
    f
    g
    h
    i
    j
    k
    l
    m
    n
    o
    p
    q
    r
    s
    t
    u
    v
    w
    x
    y
    z
    {
    |
    }
    ~
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    