1 INFO-VAX	Fri, 07 Apr 2000	Volume 2000 : Issue 194       Contents: Re: "Galaxy" on Intel?; Availability Manager V1.3 and DECamds V7.2-1A now available  Re: BACKUP/IMAGE, /ALIAS etc! Critical/important bug on VMS ... % Re: Critical/important bug on VMS ... % Re: Critical/important bug on VMS ...  Re: DS20 ADDON CPU FOR SALE  Feedback request - VMS Training 2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory2 Re: Galaxy (or perhaps Wildfire) and shared memory Re: Help: SCSI Clustering 8 Hobbyist VMS V7.2 TK50 DECnet-Plus installation failure?8 How to determine via batch if another node is reachable?< RE: How to determine via batch if another node is reachable?< Re: How to determine via batch if another node is reachable?< Re: How to determine via batch if another node is reachable?< Re: How to determine via batch if another node is reachable?& Re: Identifying unfound target of GOTO Re: Inform & VMS Re: Inform & VMS InfoServer 1000  InfoServer 1000  Re: InfoServer 1000 = Re: OpenVMS and IP Licensing (was Re: OVMS Executive Council)  Re: OpenVMS Times  Re: OpenVMS Times D OT: why i'd be surprised that MSFT ever makes it in Enterprise space out through the in door 7 Re: OVMS : Should I retrain back to the latest version? 7 Re: OVMS : Should I retrain back to the latest version? ) Re: Passing Status Code between Processes  Re: patch on hold?!  Personal Workstation A to AU Re: PMAD-AA  Re: PowerChute for OpenVMS Re: PowerChute for OpenVMS Re: Pro*C on VMS Re: Pro*C on VMS' Probably a stupid OPERATOR question ... * Re: Problem with upgrade from 7.1 -> 7.2-1P Re: should BACKUP automatically dismount the first volume of a two-volume backupP Re: should BACKUP automatically dismount the first volume of a two-volume backup Re: Sign of the times. Re: Sign of the times. Re: So who will buy VMS ? D solved! Re: How to determine via batch if another node is	reachable?H Re: solved! Re: How to determine via batch if another node is	reachable? Re: SSH for VMS  Re: Suggestion for authorize Re: Suggestion for authorize Re: Suggestion for authorize& Re: Sun considers VMS a "mainframe" OS& Re: Sun considers VMS a "mainframe" OS
 Tandem on EV7  Re: UCX's NTP vs. DST Changes 0 Re: VMS today: % users x-widnows vs VT terminals VRC16 Monitor Problem  VRC16 Monitor Problem  Re: VRC16 Monitor Problem  Re: VRC16 Monitor Problem  Re: VRC16 Monitor Problem ( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff!( Re: Warning: circuitsurgeon is a ripoff! Where are TXAn: terminal ports?  Where are TXAn: terminal ports? 2 Why are there no French people in this newsgroup ?6 Re: Why are there no French people in this newsgroup ?6 Re: Why are there no French people in this newsgroup ?  F ----------------------------------------------------------------------  # Date: Fri, 07 Apr 2000 01:48:57 GMT 7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>  Subject: Re: "Galaxy" on Intel? - Message-ID: <38ED3F38.6A922AA0@earthlink.net>    David A Froble wrote:  >  > "David J. Dachtera" wrote: > > K > > A fellow I now work with told me aout a coming product which will allow J > > the simultaneous running of W/9x, Linux, etc. on Intel-based machines.> > > He didn't know the product name/vendor, he just remembered > > hearing/reading about it.  > > K > > Sounded to me like "Galaxy" on Intel. Not sure what the value would be.  > > ) > > ...and of course, VMS won't be there.  > > 
 > > *SIGH* > >  > > -- > > David J. Dachtera  > > dba DJE Systems & > > http://home.earthlink.net/~djesys/ > > > > > Unofficial Affordable OpenVMS Home Page and Message Board:/ > > http://home.earthlink.net/~djesys/vms/soho/  > K > With sub-$1000 Intel based systems available, and the current practice of R > running multiple systems to support one applications, (as an example, a web siteQ > with multiple web serving systems and a seperate database system), I cannot see H > any practical benefit of such a product.  Technically interesting, but > definitely not mainstream.  B How 'bout using the much-touted 8-way Proliant as eight individual virtual machines?    --   David J. Dachtera  dba DJE Systems " http://home.earthlink.net/~djesys/  : Unofficial Affordable OpenVMS Home Page and Message Board:+ http://home.earthlink.net/~djesys/vms/soho/    ------------------------------  % Date: Thu, 30 Mar 2000 11:19:25 -0500 & From: "AvailMan" <AvailMan@compaq.com>D Subject: Availability Manager V1.3 and DECamds V7.2-1A now available6 Message-ID: <8bvv5c$qkt$1@mailint03.im.hou.compaq.com>  9  Compaq Availability Manager Version 1.3 is now available A with kits for OpenVMS Alpha and Windows NT (Intel and Alpha). The ? data analyzer client runs on these platforms and monitors large A numbers of OpenVMS nodes simultaneously (VAX and Alpha) that have C the data collector installed. The details are in the documentation, 8 and everything you need to get started is in these kits.:  If you've been looking for a *FREE* tool (only an OpenVMS< license is required) to help manage your overall data centerD availability, Compaq's Availability Manager for OpenVMS provides theE kind of real-time monitoring and diagnostic capabilities you may find  very useful.>  Availability Manager evolved from the DECamds product that isG Motif-based and VMS-only. The new, easy to use Java GUI runs on both NT H and OpenVMS (Alpha only). Although functionally equivalent to the latest@ release of DECamds, future enhancements are planned only for theH Availability Manager product. VAX users of the data analyzer client will continue to need DECamds. B  Availability Manager Version 1.3 is being shipped coincident withH a release of DECamds, version 7.2-1A. Both fix a number of problems withH earlier releases. There are also significant performance improvements inC Availability Manager Version 1.3. The kits are available on our web > site. To download the software or for more information, visit::  http://www.openvms.digital.com/openvms/products/availman/C Check it out and let us know what you think. Contact information is  on the web page.   ------------------------------    Date: 30 Mar 2000 08:10:41 -05004 From: "Robert Deininger" <rdeininger@mindspring.com>% Subject: Re: BACKUP/IMAGE, /ALIAS etc + Message-ID: <B508BC83-1E8CE@165.247.42.221>   I On Wed, Mar 29, 2000 4:43 PM, Phillip Helbig <helbig@astro.rug.nl> wrote: E >I'm a bit confused about what HELP says for BACKUP/IMAGE and /ALIAS.  > ) >First, although it's not my problem now,  > A >     Specifies that the previous behavior of multiple processing 7 >     of alias and primary file entries be maintained.   > E >will confuse anyone who doesn't know what the previous behaviour is.  >  >I just did  > A >   $  BACK/NOINIT/LOG/VER/IMAG/IGN=(INTERL,NOBA) DKA100: DKA200:  > H >The DKA100 disk had some files and some directories which were created J >with SET FILE/ENTER.  The new disk seems to be identical in this respect   >to the old one.  But HELP says: > F >     Specifying the /IMAGE qualifier without also specifying /NOALIASC >     can result in incomplete disk or file restoration operations. E >     Therefore, Compaq strongly recommends that you specify /NOALIAS ? >     with /IMAGE when performing image mode backup operations.  > H >Naively, one expects /IMAGE to make an exact copy.  That's what it did I >above, with /ALIAS implicitly specified (i.e. the default).  What's the   >story?   G I agree, the documentation for /ALIAS is very confusing.  I *think* you E may need to use /NOALIAS when you restore a saveset made with the old  (V6.x?) backup.   H At V7.1, BACKUP/IMAGE (with the default /ALIAS) seems to save everythingI correctly.  But I'm not sure what happens during a restore.  From reading F the manual, I can't convince myself that a restore will do the aliases right.   ---------------------------  Robert Deininger rdeininger@mindspring.com    ------------------------------  % Date: Thu, 06 Apr 2000 20:09:34 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>* Subject: Critical/important bug on VMS .../ Message-ID: <38ED2735.1783ECE3@vl.videotron.ca>   L I made a few modifications to the "Concorde" in Flight 3.1 and gave it a few million pounds of thrust.   N Carefull use has allowed me to attain escape velocity and left earth. (I think$ I am travelling at about Mach 65 :-)  G Even though I am well outside the earth, the cockpit still indicates an I airspeed. Since airspeed is measured from an air pressure measuring pitot N outside the plane, and since there is no air around me, shouldn't the airspeed! indicator indicate a speed of 0 ?   G While I am at it, has anyone devellopped a space shuttle "aircraft"  ?    F Would the program allow for an orbiting "airport" (international space
 station) ?   ------------------------------  $ Date: Thu, 6 Apr 2000 20:49:23 -0400' From: "Bill Todd" <billtodd@foo.mv.com> . Subject: Re: Critical/important bug on VMS ...( Message-ID: <8cjb7j$s04$1@pyrite.mv.net>  ; JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote in message ) news:38ED2735.1783ECE3@vl.videotron.ca... J > I made a few modifications to the "Concorde" in Flight 3.1 and gave it a few  > million pounds of thrust.  > J > Carefull use has allowed me to attain escape velocity and left earth. (I think & > I am travelling at about Mach 65 :-) > I > Even though I am well outside the earth, the cockpit still indicates an K > airspeed. Since airspeed is measured from an air pressure measuring pitot G > outside the plane, and since there is no air around me, shouldn't the  airspeed# > indicator indicate a speed of 0 ?   K One might assume that the mechanism adjusts for air density.  Perhaps it is E sufficiently sensitive to be able to compute 'air speed' based on the ' occasional hydrogen atom you encounter.    - bill   > H > While I am at it, has anyone devellopped a space shuttle "aircraft"  ? > H > Would the program allow for an orbiting "airport" (international space > station) ?   ------------------------------  # Date: Fri, 07 Apr 2000 03:23:55 GMT * From: "Jack Peacock" <peacock@simconv.com>. Subject: Re: Critical/important bug on VMS ...B Message-ID: <fxcH4.3974$y4.75787@newsread1.prod.itd.earthlink.net>  ; JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote in message ) news:38ED2735.1783ECE3@vl.videotron.ca... I > Even though I am well outside the earth, the cockpit still indicates an K > airspeed. Since airspeed is measured from an air pressure measuring pitot G > outside the plane, and since there is no air around me, shouldn't the  airspeed# > indicator indicate a speed of 0 ?  > L Very little in the way of gases up there, I thnk you would need some type ofI inertia-based instrument to measure velocity, or something to compare the H time shift based on speed relative to a stationary atomic clock on earthK (remember? time slows down the faster you go).  Surely electronics are fast L enough to measure the difference these days in the age of Ghz devices.  What did the SR-71 use? > H > While I am at it, has anyone devellopped a space shuttle "aircraft"  ? > J I believe that project is well along out at Edwards AFB as we speak.  NASAL is still working on "X" planes, I heard something about it on the radio lastC week but forgot the project name, X-35 maybe?  It's a direct to LEO  aerospace plane.    Jack Peacock    ------------------------------   Date: 6 Apr 2000 15:48:12 PDT T From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)$ Subject: Re: DS20 ADDON CPU FOR SALE3 Message-ID: <RByioQj2SedZ@mccdev.slac.stanford.edu>   . In article <sel3g39ueop6@corp.supernews.com>, 2     	"David Turner" <d_b_turner@yahoo.com> writes:N > I thought this would be the best place to post this as comp.sys.dec seems to > concentrate on older stuff > > > We pulled this out of an almost new ds20 (same as ds20e cpu)  !         Licensed for which O/S???   < > it was an smp upgrade but we can't find the license for it  H         In my experience, all SMP  upgrades  for Alpha _include_ the SMPH     license  (for that CPU).  On VMS (and presumably, for Tru64), you doH     get a PAK that must be loaded in order to use that  additional  CPU.1     Without the PAK, the CPU is pretty useless...            -Ken --  M  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edu :  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  $ Date: Thu, 6 Apr 2000 16:07:46 -05005 From: "Matute, Jaime" <Jaime.Matute@austinenergy.com> ( Subject: Feedback request - VMS TrainingT Message-ID: <CB874B506A79D1118FBC006097306B8905387A0B@ohms.electric.ci.austin.tx.us>  H Does anyone have any feedback on OpenVMS training by Compaq's AuthorizedF Independent Training Partners?  Please speak to Instructor competence,: facilities, ease of registration, course location.  Thanks   ------------------------------  % Date: Thu, 06 Apr 2000 15:17:05 -0400 * From: Joshua Cope <Joshua.Cope@Compaq.com>; Subject: Re: Galaxy (or perhaps Wildfire) and shared memory * Message-ID: <38ECE2B1.3B34EF86@Compaq.com>   Bill Todd wrote: > L > That may be true in currently-shipping systems - as long as the failure of5 > any VMS instance causes the entire box to reboot.     O The set of failures which will cause the entire system to reboot is relatively  N small. In most cases, a Galaxy partition failure doesn't bring down any other ' partitions, even on current platforms.    N > The problem is that VMS itself is not coded to handle such sharing problems,M > because, unlike applications, until the advent of Galaxy there's never been L > more than one instance of VMS running, and if that instance crashes (e.g.,J > due to a hardware problem), the contents of memory becomes irrelevant.    J I wouldn't say it becomes "irrelevant". Shared memory is not owned by any L single partition, and persists as long as at least one member of the Galaxy O is still available. If an instance crashes, then application shared memory may  N be left in an inconsistent state, but it's no worse than if a process crashes N on a traditional SMP system -- any memory shared with other processes is also M in an inconsistent state. If an app can deal with process failures on an SMP  H system, it's not that much of a stretch to deal with partition failures  on a Galaxy system.   ; I don't think SMP cache coherency is really an issue here.    < ------------------------------------------------------------7 The above opinions and information are not necessarily  % those of Compaq Computer Corporation. < ------------------------------------------------------------   ------------------------------  $ Date: Thu, 6 Apr 2000 16:00:29 -0400' From: "Bill Todd" <billtodd@foo.mv.com> ; Subject: Re: Galaxy (or perhaps Wildfire) and shared memory ( Message-ID: <8ciqaa$bor$1@pyrite.mv.net>  5 Joshua Cope <Joshua.Cope@Compaq.com> wrote in message $ news:38ECE2B1.3B34EF86@Compaq.com... > Bill Todd wrote: > > K > > That may be true in currently-shipping systems - as long as the failure  of5 > > any VMS instance causes the entire box to reboot.  > E > The set of failures which will cause the entire system to reboot is 
 relativelyI > small. In most cases, a Galaxy partition failure doesn't bring down any  other ( > partitions, even on current platforms. > F > > The problem is that VMS itself is not coded to handle such sharing	 problems, J > > because, unlike applications, until the advent of Galaxy there's never beenG > > more than one instance of VMS running, and if that instance crashes  (e.g.,J > > due to a hardware problem), the contents of memory becomes irrelevant. > ) > I wouldn't say it becomes "irrelevant".   G Wouldn't you in a box running only a single (now crashed) VMS instance? D That was my statement above, which I then contrasted with the Galaxy environment.  "  Shared memory is not owned by anyF > single partition, and persists as long as at least one member of the GalaxyL > is still available. If an instance crashes, then application shared memory mayaG > be left in an inconsistent state, but it's no worse than if a processe crashes  > on a traditional SMP systemd  K If you'll re-visit what you snipped in this response, you'll see an example(J of a case in which it's quite different:  the possibility that some systemL service (say, some kind of nominally atomic conversion process not requiringJ any I/O) operating on the shared memory may be interrupted by the instanceI crashing in the middle of its processing, whereas in a traditional SMP iftA that happened the whole box crashed and no one could ever see the  partially-updated data.?  H In other words, sharing memory between partitions introduces new failureF modes that applications never had to deal with before.  Not to mentionI failure modes in memory shared at the system level between instances thatOE the system never had to contemplate at all when the only inter-system0J communication was message-based (save for Memory Channel, which made clearI the new problems involved and did not even try to masquerade as "just thes( same old shared memory you're used to").  2  -- any memory shared with other processes is alsoJ > in an inconsistent state. If an app can deal with process failures on an SMPtI > system, it's not that much of a stretch to deal with partition failures  > on a Galaxy system.A > < > I don't think SMP cache coherency is really an issue here.  K Inter-partition cache coherence is *exactly* the issue I've tried to raise,tE though you're correct that it is not particularly related to what you J discuss above (which is still an interesting discussion in its own right).  L Sorry to keep pounding on this point, but no one yet seems to understand it.F When they do, the answer may well be that the hardware will handle theJ problem - but that answer has not yet been offered in a context that makes( it clear that it addresses the question.   - bill   >l> > ------------------------------------------------------------8 > The above opinions and information are not necessarily' > those of Compaq Computer Corporation.s> > ------------------------------------------------------------   ------------------------------  $ Date: Thu, 6 Apr 2000 16:08:59 -0400- From: "Peter Weaver" <peter.weaver@stelco.ca>o; Subject: Re: Galaxy (or perhaps Wildfire) and shared memory / Message-ID: <seprr0derfg141@corp.supernews.com>!  : Bill Todd wrote in message <8cihcr$2m8$1@pyrite.mv.net>... >u9 >norm lastovica <nlastovi@us.oracle.com> wrote in message!( >news:38EC90FF.7EF5DD79@us.oracle.com...8 >> remember that galaxy is a software solution to create9 >> partitioning.  All existing rules that make a computerr >> an "Alpha" still apply. >s@ >That may be true in currently-shipping systems - as long as the
 failure ofD >any VMS instance causes the entire box to reboot.  When you move toF >Wildfire/Marvel, it has been said that there will be hardware support for0@ >partitioning which will allow a single instance to fail without	 affectingcA >other instances (and it's possible that this is also true in thek/ >currently-shipping Alphas too - I don't know).  >h    B Our ES40 Galaxy can have one instance crash and the other instanceE usually stays up. We have had crashes during the migration of one CPUeE to another instance that causes the whole machine to go down, but allr- other crashes have only been in one instance.    ------------------------------  " Date: Thu, 6 Apr 2000 20:30:25 GMT* From: young_r@eisner.decus.org (Rob Young); Subject: Re: Galaxy (or perhaps Wildfire) and shared memorya& Message-ID: <2000Apr6.163025.1@eisner>  R In article <8ciqaa$bor$1@pyrite.mv.net>, "Bill Todd" <billtodd@foo.mv.com> writes: > 7 > Joshua Cope <Joshua.Cope@Compaq.com> wrote in message/& > news:38ECE2B1.3B34EF86@Compaq.com... >> Bill Todd wrote:e >> >L >> > That may be true in currently-shipping systems - as long as the failure > of6 >> > any VMS instance causes the entire box to reboot. >>F >> The set of failures which will cause the entire system to reboot is > relativelyJ >> small. In most cases, a Galaxy partition failure doesn't bring down any > other ) >> partitions, even on current platforms.i >>G >> > The problem is that VMS itself is not coded to handle such sharingp > problems,lK >> > because, unlike applications, until the advent of Galaxy there's never: > beenH >> > more than one instance of VMS running, and if that instance crashes > (e.g.,K >> > due to a hardware problem), the contents of memory becomes irrelevant.c >>* >> I wouldn't say it becomes "irrelevant". > I > Wouldn't you in a box running only a single (now crashed) VMS instance?aF > That was my statement above, which I then contrasted with the Galaxy > environment. > $ >  Shared memory is not owned by anyG >> single partition, and persists as long as at least one member of theV > GalaxyM >> is still available. If an instance crashes, then application shared memorya > may H >> be left in an inconsistent state, but it's no worse than if a process	 > crashes  >> on a traditional SMP system > M > If you'll re-visit what you snipped in this response, you'll see an examplerL > of a case in which it's quite different:  the possibility that some systemN > service (say, some kind of nominally atomic conversion process not requiringL > any I/O) operating on the shared memory may be interrupted by the instanceK > crashing in the middle of its processing, whereas in a traditional SMP ifaC > that happened the whole box crashed and no one could ever see theC > partially-updated data.r > J > In other words, sharing memory between partitions introduces new failureH > modes that applications never had to deal with before.  Not to mentionK > failure modes in memory shared at the system level between instances that G > the system never had to contemplate at all when the only inter-systemiL > communication was message-based (save for Memory Channel, which made clearK > the new problems involved and did not even try to masquerade as "just thea* > same old shared memory you're used to"). >   9 	I think I know what you are suggesting... please correcto: 	me if I get it wrong... Yes, what you say can happen.  So: 	wouldn't one at an OS level need to be VERY careful aboutC 	what one puts/does in shared memory?  A nice comparison I stumbledB9 	upon recently was the Cellular Disco paper whereby Disco49 	acts as a hypervisor.  Their argument:  "trust us, it is<8 	64000 lines of code, it is tight."  Okay, I'll buy thatC 	argument.  Maybe also if it runs flawlessly for many months/years.   1 http://www-flash.stanford.edu/~kinshuk/sosp99.pdfe  < 	So maybe whatever they do with shared memory at an OS level= 	they have a thousand eyes peering at the code and it is veryd; 	tight and elegant.  I'll buy that.  Afer all, aren't there 9 	similar risks with the nasty Memory Channel code?  Which1< 	by the way, Greg Pfister points out it is very tricky stuff= 	so much so a special guru wrote the Memory Channel stuff fors 	Oracle (or some such).n  F 	By the way, again .. if I am off the mark or misinterpreting (again),# 	feel free to show me the way. . . y   				RobS   ------------------------------  % Date: Thu, 06 Apr 2000 17:58:16 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>; Subject: Re: Galaxy (or perhaps Wildfire) and shared memorye/ Message-ID: <38ED0877.37F1A823@vl.videotron.ca>I   Bill Todd wrote:M > Inter-partition cache coherence is *exactly* the issue I've tried to raise,eG > though you're correct that it is not particularly related to what youAL > discuss above (which is still an interesting discussion in its own right).  N OK, node 1 and node 2 share an area of memory. Both nodes know that this blockJ of memory is shared and is in fact used to manager the cluster (instead of ethernet etc). Right ?  L Is the shared memory used solely as a cluster communication medium, or wouldE processes have direct access to virtual memory mapped to that block ?d  F If access to the shared memory is done solely though the use of systemK services, the can't these services not be written to sustain/handle corrupt I structures due to another node mis-behaving ? Also, once a node exits thevL cluster (crash etc), won't the remaining nodes know that blocks belonging to+ that nodes are corrupt and not touch them ?    ------------------------------  $ Date: Thu, 6 Apr 2000 21:16:43 -0400' From: "Bill Todd" <billtodd@foo.mv.com>5; Subject: Re: Galaxy (or perhaps Wildfire) and shared memorys( Message-ID: <8cjcqs$10c$1@pyrite.mv.net>  ; JF Mezei <jfmezei.spamnot@vl.videotron.ca> wrote in messageh) news:38ED0877.37F1A823@vl.videotron.ca...e > Bill Todd wrote:H > > Inter-partition cache coherence is *exactly* the issue I've tried to raise,I > > though you're correct that it is not particularly related to what younF > > discuss above (which is still an interesting discussion in its own right).  >eJ > OK, node 1 and node 2 share an area of memory. Both nodes know that this block:L > of memory is shared and is in fact used to manager the cluster (instead of > ethernet etc). Right ? >oH > Is the shared memory used solely as a cluster communication medium, or would G > processes have direct access to virtual memory mapped to that block ?a  I Using memory shared between partitions solely as the cluster interconnecttI can be done pretty much the same way it's done using Memory Channel (someMK simplifications are likely possible), and presumably with no adverse effectiD upon fault isolation (it's just a fast message-passing mechanism, if suitably implemented).  I But as I understand it the intent is to use inter-partition shared memoryeF for both system- and application-accessed 'galactic' (global) sharableI sections, and that's where thinking that it works just about the same way D that SMP global sections work can get one into considerable trouble.   >eH > If access to the shared memory is done solely though the use of systemE > services, the can't these services not be written to sustain/handlen corruptt/ > structures due to another node mis-behaving ?-  H Yes, they could - and should.  But my understanding is that applicationsK will also be able to map such shared memory directly, hence they need to be J aware of its extended failure semantics compared with SMP global sections.    Also, once a node exits theK > cluster (crash etc), won't the remaining nodes know that blocks belongingo to- > that nodes are corrupt and not touch them ?i  L Well, that's one option - and not a bad one for certain cases.  For example,I the mastery - vs. local - structures maintained by the DLM:  if the localdB lock information is maintained in private memory, then the masteryK information can be maintained in shared memory and directly accessed by theiK local lock owning systems under suitable shared interlocks, and if any suchiJ owner system fails - which one assumes is *very* rarely - then the masteryI information - presumed corrupted if the failing node held it locked - cantL simply be re-built from the local information held by the survivors, just asH happens today if the lock mastering node fails.  Note that the 'obvious'L approach of simply stuffing *all* the DLM data structures into shared memoryB would require a great deal more care, however:  for example, everyI lock/unlock operation would have to be made 'atomic' (likely using one ofIL the common approaches:  shadow-'paging', logging, ...), which would drive upJ overhead significantly over the more elegant (but less obvious) separation described above.  J The point is that *some* such approach must be agreed upon by everyone whoK is sharing any particular chunk of memory between partitions.  The possibleoJ approaches do resemble the mechanisms required to recover a global sectionK shared among multiple processes in a 'normal' SMP when one of their members @ fails, but there are additional failure modes (some of which areJ system-service-related, but I'm not sure there aren't others) that need to be covered.m  H None of which (again) is related to the original cache-coherence issue IJ raised, but (as I said before) it's an interesting topic in its own right.   - bill   ------------------------------  $ Date: Thu, 6 Apr 2000 21:26:27 -0400' From: "Bill Todd" <billtodd@foo.mv.com>i; Subject: Re: Galaxy (or perhaps Wildfire) and shared memory ( Message-ID: <8cjdd4$18m$1@pyrite.mv.net>  K Apologies for replying to myself (again), but I had an interesting off-line J conversation that may help shed light on the kind of problem I'm trying to0 describe (name omitted to protect the innocent):    J Actually, my real worry was that the hardware interconnections required toL join caches across partitions constituted a potential cross-partition singleL failure point in and of themselves (of a kind that your option '2' would notH necessarily protect against, though hardware isolation on these paths inK excess of that normally found on SMP systems could) - a failure point whichfE conventionally-separated cluster nodes do not have since they have nog' inter-node shared memory to begin with.m  L That may be kind of a nit (or a failure mode of sufficiently low probabilityH that no one is really going to care much), but if true it does mean thatJ fault-isolation truly equivalent to that of conventional clustering cannotJ be achieved if cachable shared memory is used between partitions - whereasJ uncached shared memory could maintain the same level of fault-isolation in
 this area.  L Good point about rebooting needing to flush rather than destroy cached data,% and it brought to mind another issue:   L If write-back caching is used with such shared memory, modified data not yetI requested by anyone else could be sitting in a processor cache long afterrH the interlocks protecting shared access to it had been released - and ifI that partition then died in any manner such that the dirty cache contents L were lost, the shared memory would be left in an undetectibly-corrupt state.H On the other hand, if write-back caching is not used when modifying suchH shared memory, then I'm not sure that preserving cache contents across aL reboot is necessary:  whatever interlocks are guarding the modifications canJ remain in place until whatever recovery mechanisms executing subsequent toJ the failure clean up the shared state (though I'm not quite sure that suchL sequencing fits well with the existing VMS distributed lock manager recovery  H model - something I've observed w.r.t. distributed file-system issues asL well), and they won't really care just how much of that state did or did notD get modified before the failure.  On the *third* hand, however, thisI highlights the necessity if write-back caching *is* used of ensuring that-J the rebooting is somehow synchronized with the recovery (being executed byF someone else) such that dirty but now stale data won't be written back$ *after* the recovery has executed...  J All this may mean that one simply does not write-back-cache data in memoryK shared across partitions, though if you ensure that a reboot never destroys G cache contents and believe that any hardware problem that would destroypH cache contents or cache connectivity is low enough to ignore I guess youK could write-back cache.  On the other hand (not sure which one we've got togK by now, but I feel like an octopus), just prohibiting write-back caching toeE such memory might be enough that no special additional steps would be D necessary to protect surviving sharing members against virtually anyG hardware failure - and preserving the ability to read-cache such sharedrJ memory likely is far more important to its performance than the ability to write-back-cache it.   - bill  0 Bill Todd <billtodd@foo.mv.com> wrote in message" news:8cihdv$2mb$1@pyrite.mv.net... > @ > Fred Kleinsorge <kleinsorge@star.zko.dec.com> wrote in message' > news:8ci8ih$nfn$1@lead.zk3.dec.com...tJ > > As of this writing, memory-is-memory, and this is a software construct toK > > create something called "shared" memory versus "local" memory.  The guy  > whowK > > is doing most of the lock management work is the one to comment on this  > typeI > > thing.  But I believe that the plans are to provide locks both in thetK > > conventional way, where the shared memory is simply a transport (like as CIK > > or a memory channel), and it will handle failures in the same way, witht > theoI > > same robustness.  Then there are plans to allow locks to exist in thel > sharedJ > > memory itself, and under those conditions I believe that you lose someJ > > robustness - however, the guy doing the work is a pretty clever guy... soI > > you never know what he may have come up with (I can't say I have beeng6 > > following it, I'm kind of busy with other things). >rL > Whoops - I think I covered this in the response I just wrote elsewhere.  IK > believe I understand the software issues involved in sharing memory amongaH > partitions, its the hardware-cachability of such shared memory and theG > implications for inter-partition fault isolation (due to the required J > cross-partition cache-consistency hardware coupling) I was asking about. >s > - bill >y > >n > >h > >  > >m> > > Bill Todd wrote in message <8cgo3b$3tf$1@pyrite.mv.net>...J > > >Though Galaxy and Wildfire/Marvel seem pretty neat, some of the hopes > thatL > > >have been expressed for them may overlook certain complexities inherent > inD > > >sharing memory across independent system instances executing in separateL > > >partitions.  Interesting applications of shared memory that may requireK > > >significant re-thinking to achieve optimal results include things likemL > > >shared-memory DLM coordination (rather than just using shared memory to > > passC > > >DLM messages of the current variety) and shared-memory caches.y > > >oJ > > >In thinking about the latter, I started wondering whether the (small) > added K > > >latency involved in fetching cache lines from non-local memory in suchs aiF > > >shared-memory cache might be largely offset by the fact that once fetchedtK > > >the lines would likely remain in local processor cache, such that eachoI > > >required line would only be fetched remotely once (given the size of L > > >outermost-level processor caches these days).  That sounded pretty goodG > > >until I asked myself just how multiple independent OS instances inaJ > > >partitions that might be hardware-isolated for robustness (to support > trueH > > >cluster-in-a-box availability with no inter-instance shared failure > modes)C > > >might allow write activity to shared memory by one instance toe
 invalidate6 > > >data in the processor caches of another instance. > > >eJ > > >So my question is:  is inter-instance shared memory likely to have to beI > > >uncached memory to avoid such (potentially infectious) interactions?D Or > isL > > >it envisioned that there will be 'rules' dictating the handling of suchJ > > >shared memory (perhaps involving suitable lock acquisition) that willF > > >eliminate the need for such cross-instance invalidation (e.g., byL > > >guaranteeing that any stale data in an instances caches should never beF > > >accessed)?  Or am I creating a problem where none in fact exists? > > >o > > >Thanks for any insight, > > >s > > >- bille > > >l > > >t > > >t > >  > >r >s >c   ------------------------------  % Date: Fri, 31 Mar 2000 13:46:26 +0930l$ From: "sseng" <sseng@ozemail.com.au>" Subject: Re: Help: SCSI Clustering3 Message-ID: <ZEVE4.59643$3b6.234536@ozemail.com.au>v  K Of course you'll also need to ensure that the letter designator of the SCSI G controllers connected to the same SCSI bus match on both systems.  ThisnI is so that the disk names are identical from both systems.  Otherwise the3A same disk could be seen as DKA100 from one system and DKB100 fromGK the other.  Also if you have MSCP serving turned on you may see the "other"1B disk via MSCp from the other system (may explain your "ghosting").  F e.g, System1 sees DKA100 through the direct SCSI connection and DKB100L via MSCP, and System2 sees them the other way round.  This will lead to dataK corruption as VMS thinks they are two separate disks and will write to themi accordingly.    D Malcolm Dunnett wrote in message <13jHrsOKJo0k@malvm2.mala.bc.ca>...* >In article <8bv1d4$3f3$1@nnrp1.deja.com>,+ >    rob.mason@bdgh-tr.trent.nhs.uk writes:a >>G >> My organisation has recently purchased a Compaq RAID 8000 box.  ThisnB >> unit will be clustered to two Alpha nodes via SCSI controllers,H >> accessed as shared storage.  We have hit a problem where the physicalJ >> devices are 'ghosted' on each node (i.e. each node sees the disks twice- >> rather than once when we do a 'sh dev d').  >>I >> The SCSI clustering documentation is barely helpful - does anyone havem >> any ideas ? >> >c= >   Most likely an ALLOCLASS problem. What are the values fortB >ALLOCLASS on each node ( and the port allocation classes for each& >SCSI bus if you've used this feature) >oL >=========================================================================== ==; >Malcolm Dunnett      Malaspina University-College   Email:e dunnett@mala.bc.caI >Information Systems  Nanaimo, B.C. CANADA V9R 5S5     Tel: (250)755-8738m >r   ------------------------------  + Date: Thu, 06 Apr 2000 23:11:53 -0500 (CDT)e From: sms@antinode.orgA Subject: Hobbyist VMS V7.2 TK50 DECnet-Plus installation failure?n) Message-ID: <00040623115367@antinode.org>f  H    I've recently been playing with putting the Hobbyist VMS V7.2 CD-ROM F installation kit onto TK50 tape.  After extracting the data concerningG which save sets go onto which tapes, I thought that all was well.  Then A I (foolishly, it seems) tried an installation which failed at thet DECnet-Plus stage:   [...]p  8 * Enter name of drive holding the DECnet-Plus kit: mua0:( %SYSTEM-F-IVLOGNAM, invalid logical name  P ********************************************************************************  0     Installation of OpenVMS VAX V7.2 incomplete.  :     Error - exiting from Mandatory Update phase procedure./     Restart this phase by rebooting the system.t   [...]t  F    When I attempted to "[r]estart this phase by rebooting the system", the error recurred.h  F    The cause turned out to be a typo in [VMS$COMMON.SYSEXE]STARTUP.MANE (in VMS072.B), proving conclusively that "net_def" is not the same asI? "net_dev".  It affects only installations from tape, naturally.I  B     Has anyone else run into this one, or am I the only one looney enough to try this?   H ------------------------------------------------------------------------  C    Steven M. Schweda               (+1) 651-699-9818  (voice, home)4C    382 South Warwick Street        (+1) 763-781-0308  (voice, work)lG    Saint Paul  MN  55105-2547      (+1) 763-781-0309  (facsimile, work)i9    sms@antinode.org                sms@provis.com  (work)f   ------------------------------  % Date: Thu, 06 Apr 2000 14:17:07 -0400h/ From: "Joseph Pomeroy" <jpomeroy@holycross.edu> A Subject: How to determine via batch if another node is reachable? / Message-ID: <s8ec9c7f.026@dudley.holycross.edu>s  
 Greetings:  B Using VMS-ALPHA 7.1 and MULTINET 4.1--  We would like to have a=20  I method of determining whether another TCPIP node is  reachable or not.=20t  I We would like to use a batch job so it can be repetitively run. We HAD=20a  H thought of using MU PING and checking for a return status  other than=20  G NORMAL. Unfortunately,we discovered that PING never returns when the=200  F remote node is unreachable. PROCESS asserts that this is  conscious=20   design.p  3 Does anybody have an idea for a different approach?7   Thanks,i   Joe Pomeroy, S.J.T Ops Mgr-ITS  Holy Cross College Worcester, Mass.   Jpomeroy@holycross.edu=20o   ------------------------------  $ Date: Thu, 6 Apr 2000 15:42:18 -03001 From: "Boyle, Darren" <boyledj@bankofbermuda.com>sE Subject: RE: How to determine via batch if another node is reachable?eK Message-ID: <F150836441C5D311A11700508B6FF01A59527D@bdant024.bda.bobda.com>a   what's wrong with FTPu - Darren   > ----------4 > From: 	Joseph Pomeroy[SMTP:jpomeroy@holycross.edu]) > Sent: 	Thursday, April 06, 2000 3:17 PMr > To: 	Info-VAX@Mvb.Saic.ComD > Subject: 	How to determine via batch if another node is reachable? >  > Greetings: > B > Using VMS-ALPHA 7.1 and MULTINET 4.1--  We would like to have a  > I > method of determining whether another TCPIP node is  reachable or not. l > I > We would like to use a batch job so it can be repetitively run. We HAD   > H > thought of using MU PING and checking for a return status  other than  > G > NORMAL. Unfortunately,we discovered that PING never returns when the l > F > remote node is unreachable. PROCESS asserts that this is  conscious  > 	 > design.s > 5 > Does anybody have an idea for a different approach?e > 	 > Thanks,v >  > Joe Pomeroy, S.J.s
 > Ops Mgr-ITS0 > Holy Cross College > Worcester, Mass. >  > Jpomeroy@holycross.edu   >  >  >  >  >  >  >  >  >     F **********************************************************************C This message and any files transmitted with it are confidential andsJ may be privileged and/or subject to the provisions of privacy legislation.M They are intended solely for the use of the individual or entity to whom theyiL are addressed. If the reader of this message is not the intended recipient, B please notify the sender immediately and then delete this message.I You are notified that reliance on, disclosure of, distribution or copyinga of this message is prohibited.   Bank of BermudasF **********************************************************************   ------------------------------  " Date: Thu, 6 Apr 2000 18:29:41 GMT From: briggs@eisner.decus.orgaE Subject: Re: How to determine via batch if another node is reachable?m& Message-ID: <2000Apr6.142941.1@eisner>  a In article <s8ec9c7f.026@dudley.holycross.edu>, "Joseph Pomeroy" <jpomeroy@holycross.edu> writes:i > Greetings: > D > Using VMS-ALPHA 7.1 and MULTINET 4.1--  We would like to have a=20 > K > method of determining whether another TCPIP node is  reachable or not.=20c > K > We would like to use a batch job so it can be repetitively run. We HAD=20h > J > thought of using MU PING and checking for a return status  other than=20 > I > NORMAL. Unfortunately,we discovered that PING never returns when the=20t > H > remote node is unreachable. PROCESS asserts that this is  conscious=20 > 	 > design.s  B Whoever you talked to at PROCESS needs to be taught the differenceB between a certain portion of his or her anatomy and a concavity in the Earth's surface.  ! $ MULTI PING /NUM=3 /QUIET targeti   Adjust the "3" to taste.  & 	John Briggs			briggs@eisner.decus.org   ------------------------------  " Date: Thu, 6 Apr 2000 18:55:10 GMT* From: young_r@eisner.decus.org (Rob Young)E Subject: Re: How to determine via batch if another node is reachable?r& Message-ID: <2000Apr6.145510.1@eisner>  a In article <s8ec9c7f.026@dudley.holycross.edu>, "Joseph Pomeroy" <jpomeroy@holycross.edu> writes:  > Greetings: > D > Using VMS-ALPHA 7.1 and MULTINET 4.1--  We would like to have a=20 > K > method of determining whether another TCPIP node is  reachable or not.=20d > K > We would like to use a batch job so it can be repetitively run. We HAD=20  > J > thought of using MU PING and checking for a return status  other than=20 > I > NORMAL. Unfortunately,we discovered that PING never returns when the=20h > H > remote node is unreachable. PROCESS asserts that this is  conscious=20 > 	 > design.  > 5 > Does anybody have an idea for a different approach?  > 	 > Thanks,y >   5 	You can use ping for a rudimentary WATCHDOG process:i  1 $ define/user sys$output sys$scratch:pingpong.tmpy9 $ Multinet ping /number_of_packets=3 ip_address_goes_herei $ define/user sys$output nl: $ define/user sys$error nl:e; $ search sys$scratch:pingpong.tmp " 100% packet loss"/exact ) $ if ($status .ne. 1) then goto good_pingo  < 	Give it a try, works for me but I didn't write it found it/ 	inherited it.   				Robm   ------------------------------   Date: 6 Apr 2000 20:28:46 GMTl3 From: gartmann@immunbio.mpg.de (Christoph Gartmann)iE Subject: Re: How to determine via batch if another node is reachable?h0 Message-ID: <8cis1u$oal$1@n.ruf.uni-freiburg.de>  a In article <s8ec9c7f.026@dudley.holycross.edu>, "Joseph Pomeroy" <jpomeroy@holycross.edu> writes:g >Greetings:  > C >Using VMS-ALPHA 7.1 and MULTINET 4.1--  We would like to have a=20rJ >method of determining whether another TCPIP node is  reachable or not.=20J >We would like to use a batch job so it can be repetitively run. We HAD=20I >thought of using MU PING and checking for a return status  other than=20tH >NORMAL. Unfortunately,we discovered that PING never returns when the=20G >remote node is unreachable. PROCESS asserts that this is  conscious=20  >design.4 >Does anybody have an idea for a different approach?  @ Here is a piece of code we use to check whether a printer is up:   $ !aF $ SET NOON                             ! because of unsuccessfull PING$ $ dphone == "$mpi_exe:phone_message" $label1:3 $    MULTINET PING special_host /QUIET/DATA=4/NUM=1s $    IF .NOT. $status  $       THEN $       now = F$TIME()7 $       dnow = F$CVTIME( now, "COMPARISON", "DATETIME")a* $       dday = F$CVTIME( now, , "WEEKDAY")? $       dstart = F$CVTIME("09:00:00", "COMPARISON", "DATETIME")o> $       dstop = F$CVTIME("18:00:00", "COMPARISON", "DATETIME")O $       IF dnow .GTS. dstart .AND. dnow .LTS. dstop .AND. dday .NES. "Sunday" - &            .AND. dday .NES. "Saturday" $          THENcF $          dphone diskb:[prg.mpi.data]msg1.vox 6 464 465 93,0123456789 $       ELSEK $          dphone diskb:[prg.mpi.data]msg1.vox 6 93,0123456789 464 93,12345t
 $       ENDIFi $       EXIT
 $    ENDIF $    WAIT 00:00:10.00  $    GOTO label1 $ EXIT     Works perfectly.   Regards,    Christoph Gartmannd  H -----------------------------------------------------------------------+H | Max-Planck-Institut fuer      Phone   : +49-761-5108-464   Fax: -452 |H | Immunbiologie                                                        |H | Postfach 1169                 Internet: gartmann@immunbio.mpg.de     |H | D-79011  Freiburg, FRG                                               |H +------------ http://www.immunbio.mpg.de/english/menue.html -----------+   ------------------------------  % Date: Fri, 31 Mar 2000 12:46:01 +1000o/ From: "Phil Howell" <howellp@snowyhydro.com.au> / Subject: Re: Identifying unfound target of GOTOg3 Message-ID: <LjUE4.59610$3b6.234370@ozemail.com.au>   * but wasn't that the point of the exercise? Phil   ------------------------------   Date: 6 Apr 2000 20:13:32 GMTo0 From: sander@vmsbiz.enet.dec.com (Warren Sander) Subject: Re: Inform & VMS,* Message-ID: <8cir5c$tuo@usenet.pa.dec.com>   : ps files of openvms times.  9 I've created 2 new ps files of the current openvms times.h  G another LPS20, postscript level 1 and an LPS17 version again postscripta level 1.  9 Hopefully one of these will print correctly for you all..u  = [If someone makes a postscript that does work let me know how & you did it and I'll post that version]     -warren    -- sB ------------------------------------------------------------------6 Warren Sander                        OpenVMS MarketingD Compaq Computer Corporation          Work:  warren.sander@compaq.comB 200 Forest Street MR01-3/J1          Personal: sander@ultranet.com3 Marlboro, MA 01752                   (508) 467-4875e6    My opinions are my own and I only speak for myself /           Read http://www.openvms.digital.com/ tB ------------------------------------------------------------------   ------------------------------   Date: 6 Apr 2000 14:57:17 PDTiT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515) Subject: Re: Inform & VMSy3 Message-ID: <J4monsSBiIYv@mccdev.slac.stanford.edu>   + In article <38eb64d3$1@news.kapsch.co.at>,  0     eplan@kapsch.net (Peter LANGSTOEGER) writes:- > In article <8cfmc3$gno@usenet.pa.dec.com>, d7     	sander@vmsbiz.enet.dec.com (Warren Sander) writes:e@ >>OpenVMSTIMES is not produced on Windows NT.. It is produced on >>a MAC using Quark Express. >  > Yeah. Resist M$. > J >>The Quark Express PDF file is then sent over here and I do a print... toH >>an LPS20 style printer to get the PS file. I happen to have done it onM >>a Windows NT system (I have 4 OpenVMS workstations also (and another PC)). - > @ > I did print the PS file to a HP LJ5 and was very disappointed:L > Overlapping text, single sided (though requested double side via DCPS),...N > I so far never had this happen, when using a PDF to PS converter on OpenVMS.  H         I'm a little surprised at  this.   Did you download the PDF fileH     and  then  convert?  I downloaded the zip'd PS file and  it  printedH     very nicely to  our  HP  8000DN.   One-sided,  like  you  said,  butA     otherwise no complaints.  I don't see any overlapping text...o           -Ken -- pM  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.Edue:  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  $ Date: Mon, 3 Apr 2000 16:19:04 -0600+ From: "Chris Morrill" <cmorrill@micron.net>e Subject: InfoServer 10000 Message-ID: <8cb5cq$3kc$1@admin-srv3.micron.com>  L I am having trouble getting a InfoServer to work.  The lights come on, I canL do a show dev and see the CDs.  But when I go to a node and use ESS$LADCP to" check services - nothing shows up.  A They are on the same net - no routers in-between the node and thea InfoServer.r   Any help would be great!   --  ? OpenVMS: The operating system God runs the Earth simulation on.r   ------------------------------  $ Date: Thu, 6 Apr 2000 14:03:37 -0600+ From: "Chris Morrill" <cmorrill@micron.net>  Subject: InfoServer 1000/ Message-ID: <8ciqij$pa$1@admin-srv3.micron.com>L  < >>"The lights come on, I can do a show dev and see the CDs." >>< >  Did you try show services? Does it list them? What is the? >   'Service Class' of these services? They ought to be [ODS_2]i( >   for VMS to 'see' them via ESS$LADCP.  H    On the InfoServer - when I do a show services I do get a list and the class is [ODS_2]   InfoServer> show systeme   Digital Equipment CorporationnE InfoServer 1000 V3.1         ( BL22)       System Image:  Nov  8 1993e 17:21:02  D Server Name:            BOOK_SERVER       Uptime:                002 17:57:047 Ethernet Address: 08-00-2b-a3-bf-2a       Server State:  OnA Write Access Policy:    Pre-emptive       Current LASTport Group:" 0 7 Service Extension Limit:      50000       Client Limit:- 50   Remote Device List:    Enabled characteristics:)   MOP downline loading,   Automount disks     Last software crash information:    None available. InfoServer> show service  4  Service Name         Service Class   Disk:Partition<  -------------------- --------------- ----------------------*  TEST                 [ODS_2]         DK3:*  VAXBINSEP975         [ODS_2]         DK3:   InfoServer>        > @ >>"But when I go to a node and use ESS$LADCP to check services - >> nothing shows up."u >>) >   Have you started LAD on the VMS host?i  <    Yup - LAD (sys$startup:ess$lad_startup) has been started.     $ mcr ess$ladcp    LADCP Version V13.0O  LASTPort Disk (LAD) Version V3.0 ESS$DADDRIVER Ident X-18 LADCP> show serviceo  7 Interrogating network for Disk services, please wait...i LADCP> show service/disk  7 Interrogating network for Disk services, please wait...  LADCP> exit    > G >   These are some of the things to look for. If still having troubles,y- >   please post the output of these commands.p >  >   Cheers,c >s >   -- mahendras     -- Thank you - Shane.F.Smith:  ? OpenVMS: The operating system God runs the Earth simulation on.    ------------------------------  % Date: Thu, 06 Apr 2000 16:03:28 -0700v# From: Mickey Stein <yekkim@cts.com>  Subject: Re: InfoServer 1000' Message-ID: <38ED17C0.550B8902@cts.com>-  O Ok: It's been awhile since I've stared at the console or a lat session on an ISpL 1000, but I'll try to remember, and if that fails, I think I can log in fromO home and figure out what you might need to see on the IS and *do on VMS to make0I this work. First thing that strikes me is that $sho services only shows 2gH services and no IS OS type services. Also you'd need to do a show of theH functions that the server is enabled to perform. (like disk serving, VXTP serving, cd serving, etc..). Even at this state, you should be able to do a $setM host/lat "servername" from VMS as long as OUTGOING connections are enabled inh+ Latcp. $mc latcp , set node/conn=both, exitt  J        well -- I can't remember<g> so I'm going to log on and take a look.5             back in a bit if I can find anything out.                    Mick   ---- Ok -- I checked it out. --f  P Forget that stuff above :) maybe not all of it, since you need to be able to setL host/lat to the IS just to prove you're really connected. You may be ok withP just the server defined , the service enabled, the cd serve function enabled andI some vms commands that resemble the following: (you should be able to $mcwN ess$ladcp and find out the specifics for cd-roms since these commands are just3 for infoserver 1000 disks in an attached SW array):M  ) $ if f$getsyi("NODENAME") .eqs. "MT_FUJI"O $   then% $     @sys$startup:ess$startup clienta $!H $!    bind the infoserver disks (check the /disk qualifier for something: cd-specific, unit just means how the dev name will appear) $! $     mc ess$ladcp1         bind/disk/conn/unit=1 name_of_cd_as_on_ISh         exit $! $! $! $!> $!    mount the infoserver disks that are currently available. $!L $     mount/noassist /system dad1: name_of_cd_as_on_IS logical_name_you_want	 $   endif,       Chris Morrill wrote:  N > I am having trouble getting a InfoServer to work.  The lights come on, I canN > do a show dev and see the CDs.  But when I go to a node and use ESS$LADCP to$ > check services - nothing shows up. > C > They are on the same net - no routers in-between the node and thei
 > InfoServer.  >| > Any help would be great! >  > -- > A > OpenVMS: The operating system God runs the Earth simulation on..   ------------------------------  # Date: Thu, 30 Mar 2000 14:53:40 GMT 7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>eF Subject: Re: OpenVMS and IP Licensing (was Re: OVMS Executive Council)- Message-ID: <38E36B14.5F70BAA2@earthlink.net>@   JF Mezei wrote:  >  > Bob Kaplow wrote: K > > That, and the license fees themselves. My previous point is that if the2P > > license fee were lower, volume would increase to make up the difference. Add= > > support to the larger installed base and revenue goes up.l > > L > > A larger installed base means more apps migrated to VMS, and more timely6 > > migrations. It's all an intertwined upward spiral. > >n: > > Or a downward spiral if things continue as they are... > L > If a stupid person like me knows what you said, then surely the experts at > Compaq would too.a  C I dare say that there are those here who would argue that "Compaq :-C experts" is roughly akin to "Military : Intelligence". Not saying I < agree or disagree - the weight of evidence in either case isD substantial. Compaq (thinks it) is protecting its stock holders. TheG market thinks Compaq is flushing a potential gold mine down the toilet. ? Pretty tough call. Even gold mines eventually stop producing, Im
 suppose...  7 > But if their mandate is to control the erosion of VMS-O > installed base to an acceptable rate, then they have no incentive to do great- > things with VMS.  # ...as evidenced by what we've seen.C  aN > What this tells me is that the problem is high up at Compaq. If the top guysO > were to give mandates to make VMS competitive on all business markets againstbN > W2K, then the lower end grunts at Compaq would find ways to make it so. SineP > it hasn't happened, you can only conclude that the top guys are happy with the9 > way VMS is treated right now (controlled demise/death).i  @ All the more reason to buy it away from them. Been weighing your piggy-banks, folks?>   -- r David J. Dachteraa dba DJE Systemss" http://home.earthlink.net/~djesys/  : Unofficial Affordable OpenVMS Home Page and Message Board:+ http://home.earthlink.net/~djesys/vms/soho/i   ------------------------------  % Date: Thu, 06 Apr 2000 14:13:16 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca> Subject: Re: OpenVMS Times/ Message-ID: <38ECD3B2.F3F40765@vl.videotron.ca>a  $ system@niuhep.physics.niu.edu wrote:? > I would encourage the editors to make sure it remains legible B > with xpdf and preferably with Ghostscript as those are the tools > available to most VMSites.  M Not much that can be done about that. With the Acrobat PDF compiler, you justeG feed it a PS file and it spews out a PDF file. There are a few switchesiC though. Perhaps the author could ask for Acrobat version 3 output ?    ------------------------------  " Date: Thu, 6 Apr 2000 21:12:16 GMT2 From: kilgallen@eisner.decus.org (Larry Kilgallen) Subject: Re: OpenVMS Times& Message-ID: <2000Apr6.171216.1@eisner>  b In article <38ECD3B2.F3F40765@vl.videotron.ca>, JF Mezei <jfmezei.spamnot@vl.videotron.ca> writes:& > system@niuhep.physics.niu.edu wrote:@ >> I would encourage the editors to make sure it remains legibleC >> with xpdf and preferably with Ghostscript as those are the toolsy >> available to most VMSites.y > O > Not much that can be done about that. With the Acrobat PDF compiler, you justaI > feed it a PS file and it spews out a PDF file. There are a few switcheseE > though. Perhaps the author could ask for Acrobat version 3 output ?c  F I have had good luck reading large PDF files converted with Acrobat V2C on a Macintosh from Postscript files created on VMS.  For DistillercC runs, I am not convinced there is any need to run a recent Acrobat.D   ------------------------------  % Date: Thu, 06 Apr 2000 11:35:33 -0700 5 From: "Larry D Bohan, Jr" <LBohan@dbc.spam_less..com>hM Subject: OT: why i'd be surprised that MSFT ever makes it in Enterprise spacen2 Message-ID: <kNfsOJlRzKcFLwOhLFh1PzJMSBK7@4ax.com>  9 .8665 - .8447 = 218 builds in (11/14 - 04/27)  ~210 days. 9 .8397 - .8293 = 104 builds in (07/11 - 02/02)  ~159 days.c  + obviously, MSFT puts significant time into f/ their regression testing.  what, 28-36 hrs per?t  7 taken w/ some grains of salt, as I'm sure some of this e, is due to their  practice of daily rebuilds  (probably including weekends)s  0 Can y'all imagine having this many versions of  ' C-runtime sharables  available on VMS ?     @ <...>  has gotten calls from Clients who have upgraded to <X.y> A and started getting page fault errors.  It is not the majority ofnC Clients, but enough to make me concerned.  All the errors have beeniD related to mfc42.dll and msvcrt.dll.  When I checked what we have onD the source server to go into our installer, I noticed that there areB more recent versions on my PC for both of them.  Is there a chanceE that the latest builds need a higher version than we are including iny< the installer, or something else that anyone has an idea on?   On Source Server:o  . MFC42.dll 	Ver. 6.00.8447.0	Mod Date 4/27/1999- MSVCRT.dll	Ver. 6.00.8397.0	Mod Date 2/2/1999e  	 On my PC:l/ MFC42.dll 	Ver. 6.00.8665.0	Mod Date 11/14/1999r. MSVCRT.dll	Ver. 6.10.8293.0	Mod Date 7/11/1999   ------------------------------  $ Date: Thu, 6 Apr 2000 21:01:41 -04002 From: "Brian Kennedy" <kennedy-brian@ameritel.net>  Subject: out through the in door( Message-ID: <38ed2d9b@glorpnob.tqci.net>  L I have a problem that I have been trying to solve for several weeks and I amF out of ideas. I am able to get great throughput when initiating DECnetK transfers from outside my cluster's area to any member within my cluster. IsJ also get great throughput when I transfer data between the cluster membersG within my cluster. However,  transfers from any member of my cluster to1G systems that are outside the cluster's area are ALWAYS terrible. When I L watch the counters I see way too many timeouts. It can take up to 45 minutesK to transfer a 1MB file. Utilization on the network is less than one percent K and COPY/FTP works great, so I know it isn't a network performance problem. L It has to be a DECnet problem, but I just can't figure it out. If anyone hasK seen these symptons before, please let me know what you did do clear it up.b> Below is the cluster/network configuration. Thanks in advance.       Alias: WHATSUP     Alias Address: 1.999  /     Members: (each member is a Routing IV node)0           Node Name: WHATSUPAl         Node Address: 1.1001         Hardware: Alpha 4100(         Operating System: OpenVMS 6.1-H2           Node Name: WHATSUPBs         Node Address: 1.1002         Hardware: Alpha 4100(         Operating System: OpenVMS 6.1-H2       Area Router:  @         RouteAbout Access EW 90 with the latest firmware upgrade         Address: 1.1023c       All wiring is CAT-5.   ------------------------------  % Date: Thu, 30 Mar 2000 15:13:26 +0100t From: bfb <bfb@tesco.net>h@ Subject: Re: OVMS : Should I retrain back to the latest version?8 Message-ID: <l3o6es85h8l7gsa6dnv3oee5s8ko1s4fl7@4ax.com>  2 On Wed, 29 Mar 2000 23:43:36 -0500, David A Froble <davef@tsoft-inc.com> wrote:   >fM >If you can work in a VMS environment, then I don't see the issue of 'current-Q >experience' being a problem.  In many cases companies are having trouble finding ! >VMS help, and you'd be a 'find'.  >     Thats good news to hear, thanks.  P >You say it's been a while, but either I should be kicked out, or I question youM >experience on a 3100/105e system.  No such beast.  Possibly a VAX 4000 model / >105, or a MicroVAX 3100 model 10e.  Or a typo.s  B Ooops, yes I stand corrected, it was a 4000 not a 3100, apologies.   > P >Hey, if you want to work with VMS, look around.  This isn't an employment list,# >but I believe there is at least 1.  >n  > Ok, I shall look at vmsnet.employment and see what I can find.  E I have received a fair few emails, I thank you all for your help. Now*/ on to job hunting (even with IR35 nearly here!)m     Regards  bfb    ------------------------------  # Date: Thu, 30 Mar 2000 14:55:04 GMTb= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)r@ Subject: Re: OVMS : Should I retrain back to the latest version?0 Message-ID: <009E7DB6.7C231C47@SendSpamHere.ORG>  T In article <l3o6es85h8l7gsa6dnv3oee5s8ko1s4fl7@4ax.com>, bfb <bfb@tesco.net> writes:3 >On Wed, 29 Mar 2000 23:43:36 -0500, David A Frobler ><davef@tsoft-inc.com> wrote:  >o >>N >>If you can work in a VMS environment, then I don't see the issue of 'currentR >>experience' being a problem.  In many cases companies are having trouble finding" >>VMS help, and you'd be a 'find'. >> >v! >Thats good news to hear, thanks.S >eQ >>You say it's been a while, but either I should be kicked out, or I question you N >>experience on a 3100/105e system.  No such beast.  Possibly a VAX 4000 model0 >>105, or a MicroVAX 3100 model 10e.  Or a typo. >eC >Ooops, yes I stand corrected, it was a 4000 not a 3100, apologies.h >c >>Q >>Hey, if you want to work with VMS, look around.  This isn't an employment list,-$ >>but I believe there is at least 1. >> >r? >Ok, I shall look at vmsnet.employment and see what I can find.   G vmsnet.employment has been rather sparse.  The only thing that seems tobH get posted to vmsnet.employment these days are COMPLETELY-OFF-TOPIC typeG postings and SPAM, SPAM, SPAM, SPAM, SPAM.  I am having a great time ofsI it though finding the SPAMmers and having their ISPs pull their accounts. F I was just on the phone yesterday with PANIX to have one of their cus-> tomer's yanked for the abusive SPAMs they have been posting.    8 If, however, you are interested in any of the following:  #  -- making money fast (MMF) schemesc0  -- Multi-Level Marketting (MLM/pyramid) schemes  -- hot stock tipsD  -- losing 150 lbs in a week with a miracle fat burner (acetylene ;)@  -- nutritional diet plans to cure (your favorite disease here)   -- where to buy viagra on-linee<  -- new porno web sites (visit after you buy viagra on-line)8  -- golf tips and high-tech golf balls (golf is so lame)A  -- working at home (licking envelopes, assembling widgets, etc.) <  -- Sexy Shelia's Sultry Sundries and Sex Surrogate SuppliesF  -- penis enlargement tools -- posted by little.woody@cantmaker.com ;)   -- jobs for russian translators  -- ads for russian wivesr@  -- how to pick up women (why, when you can buy a russian wife?)A  -- check out this URL!  (God knows what could be lurking there!)u$  -- get the lowdown on your neighbor  -- spy on your neighborJ  -- find that lost friend (If they were friends, you know where they are!)  -- discount phone services   -- discount food services2  -- discount (your favorite expense here) services   F please let me know and I will graciously 'approve' them for posting to vmsnet.employment.  F BTW, anyone reading this, I'd really like to 'approve' some legitimate VMS job postings.    --N VAXman- OpenVMS APE certification number: AAA-0001           VAXman@TMESIS.COM   ------------------------------   Date: 30 Mar 2000 16:26:25 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)2 Subject: Re: Passing Status Code between Processes6 Message-ID: <8bvv7h$qjd$1@mailint03.im.hou.compaq.com>  _ In article <8bt02p$l0n$1@news.laserlink.net>, "Madhuri Garikipati" <madhuri@compaq.net> writes:M :We have a two node cluster.    %   OpenVMS version(s) and platform(s)?b  6   How many queue databases are present in the cluster?  / :I am submitting a job on one of the nodes thatcM :submits another job on the second node.  The first job synchronizes with theoM :second job's completion and continues execution.  The problem I am facing isnM :that the status code returned by the second job is not getting passed to thep :first job.e  D   Exactly what are you trying to do?  Some additional background on    the problem, please? 6  = :This kind of thing is done routinely between jobs running onn :single node.  f  C   I'm not sure what you are doing when you are passing status codeseE   around, as this is not something I think of as particularly routinet7   communications between one batch job and the next....r  H   A *short* *concise* *brief* *less-than-a-page-of-DCL* example of just G   how you think the batch jobs are communicating on the same node mights   also be of interest.  I :Is there something I am missing?  Any help will be very much appreciate.o  J   You could write the data into a file in a shared area.  This is trivial K   from DCL.  Depending on what particular requirement(s) you need to fufill2@   here, there may well be other alternatives available to you...  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 30 Mar 2000 13:38:00 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: patch on hold?!6 Message-ID: <8bvlbo$m2v$1@mailint03.im.hou.compaq.com>  [ In article <8bsv7l$t4h$1@info.service.rug.nl>, helbig@astro.rug.nl (Phillip Helbig) writes:O  4   re: dec-axpvms-vms721_sys-v0400--4.pcsi-dcx_axpexe7       dec-axpvms-vms721_update-v0100--4.pcsi-dcx_axpexeo  5 :The requested file is either unavailable or on hold.s :.E :What's the story?  Will it be available soon?  Looks like something $ :everyone should have!  H   At least one of the ECOs went on hold (and is either now off hold, or I   is coming off hold shortly) due to reports of problems during the PCSI nG   installation -- and not due to problems with the files and images in cH   the kit.  Folks here are looking into the cause of and the history of H   the particular (site-specific?) PCSI database corruptions that is the I   trigger for the installation problem, and at the stackdump that occurs sC   when the PCSI database corruption is encountered during the PCSI M   installation of the kit.  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  $ Date: Thu, 6 Apr 2000 20:42:40 -0400+ From: "David Turner" <d_b_turner@yahoo.com>l% Subject: Personal Workstation A to AUy/ Message-ID: <seqbj9a6rfg132@corp.supernews.com>g   People  9 I ahev been asked this so many times, now I am posting it*  + Upgrade for All PWS 433a 500a and 600a - AU / All include a NEW VMS or UNIX license (licence)e   OpenVMS Base license 2mb cache module MSC01-ABt RRD46-AB 12x SCSI CD-ROM KZPBA-CA (QLA1040) ELSA Synergy 8mb PCI non-vivos* 9Gb 7200rpm disk or 2 x 4.5gb 7200rpm disk% Narrow and Wide SCSI Solid wire Cable  Terminator for Wide Buse' SRM and AlphaBIOS Vers. 7.0.11 and 5.69n  I Takes about  40 minutes to do the upgrade including the AlphaBIOS and SRM  u/gl  C We'll send you an ISO9660 Burnt CD with it if required at no charge     " Price $2000 or 995 without license   Everything in stock G Everything ships same day for $50 in the US or $70 to EU. Australia and)
 Japan $110           -- David Turner Island Computers US CorporationS 2700 Gregory Street  Savannah GA 31404o Tel: 912 447 6622u Fax:912 201 0096 sales@islandco.com   ------------------------------    Date: 30 Mar 2000 08:20:36 -05004 From: "Robert Deininger" <rdeininger@mindspring.com> Subject: Re: PMAD-AA+ Message-ID: <B508BEDA-275A9@165.247.42.221>   B On Thu, Mar 30, 2000 5:18 AM, Jim Tsetsos <jim@tsetsos.net> wrote: >Hi, >DF >Does anyone know where I can buy a network card for a DEC3000/300 AXP from?  >   E Not offhand, but I suppose used equipment dealers could find you one.s  E We have a 3000/300 with a PMAD card in it, but it has a problem.  ThelF console doesn't recognize it, and it is not possible to satellite bootE over that network path.  (Booting over the bult-in ethernet works.) IiF haven't been able to find out whether this is a hardware problem or a  gap in VMS device support.       ---------------------------  Robert Deininger rdeininger@mindspring.comk   ------------------------------  # Date: Thu, 30 Mar 2000 15:39:03 GMTI= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)s# Subject: Re: PowerChute for OpenVMSn0 Message-ID: <009E7DBC.A18D9EA4@SendSpamHere.ORG>  a In article <8bvrhk$qi3@gap.cco.caltech.edu>, mathog@seqaxp.bio.caltech.edu (David Mathog) writes:e[ >In article <8bv2av$4d6$1@supernews.com>, "Colin Scott" <cps@lineone.nothisbit.net> writes:M >>L >>Does anyone know of any "similar" software either commercial or home grown2 >>that would perform the same task as PowerChute?  >sK >I've got software to control Tripp-Lite UPS's.  Probably you can adapt it lC >to control the APC if you have access to sufficient technical datai/ >on the cable/UPS. It's free.  Pick it up from:o >e: >  http://seqaxp.bio.caltech.edu/pub/SOFTWARE/TCONTROL.ZIP >tM >If you do get it to work send me the instructions/changes and I'll add them.s >e+ >There is the readme from the distribution:y > = >************************************************************  >a- >This directory contains the following files:. > F >  TCONTROL.EXE  Alpha executable for tcontrol, a command line programI >                which can be used to control a terminal line attached tow >                a UPS+ >  TCONTROL.C    source code for tcontrol, nI >  UPS_INFO.TXT  Information on Tripplite UPS lan 2.1 and lan 2.2 controli> >                signals.  Others may be added at future time.F >  UPS_MONITOR.COM  A procedure designed to be run detached which usesI >                TCONTROL to monitor a UPS attached to a terminal line.  oO >                It will shut down the system after N minutes if a powerfail is-Z >                detected.  If the power comes back before then, the shutdown is canceled.? >                Edit it to fit your system.  Start it up like:i, >                $ run sys$system:loginout -+ >                   /detach /uic=[SYSTEM] -e9 >                   /input= sys$manager:ups_monitor.com - A >                   /output= sys$manager:powerfail_shutdown.log -i! >                   /priority=6 -o/ >                   /process_name="UPS_MONITOR"i >t >  >TCONTROL usage: >r# >First configure the terminal with:  >l* >$ set term/perm/nobroadcast/nomodem tta0: >n >Then to use TCONTROLa >,# >$ tcontrol :== $disk:[dir]tcontrol 2 >$ tcontrol device bitset waithigh waitlow maxwait4 >   device:   tta0: or other serial device (not LAT)B >   bitset:   bits to set high from {DTR,RTS,SecTX}, others go lowH >   waithigh: wait until these bits go high, from {CTS,DSR,CARRIER,RING}G >   waitlow:  wait until these bits go low, from {CTS,DSR,CARRIER,RING}g6 >      if more than one in a list separate with spaces6 >   maxwait   wait this many seconds before timing out  G The APC UPS uses an protocol which does not involve hardware signals ons+ lines (other than TXD/RXD) to change state.m   --N VAXman- OpenVMS APE certification number: AAA-0001           VAXman@TMESIS.COM   ------------------------------  # Date: Thu, 30 Mar 2000 15:36:00 GMTs= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)i# Subject: Re: PowerChute for OpenVMS-0 Message-ID: <009E7DBC.34571297@SendSpamHere.ORG>  S In article <8bvnn1$1oi$1@news.inav.net>, "William Ford" <wford@bwrcorp.com> writes:-H >I have an alpha sever 800. I have it working with 7.1 and here are some >points to consider. >.. >1. ttbo: = com1, tta0: = com2. silly isn't it >hL >2. com1 does not support modem controls unless you have compaq come out andF >move a jumper. (they will not tell you over the phone - it's all veryM >hush-hush) and when it is set it will only run if set at 38400. really sillyh >now.h >e- >3. no modem control = no powerchute or modemS   ^^H @@ The UPS connects via a bizarre cable which uses only TXD,RXD and GND.  J Bizarre in that the pins at the UPS end are not EIA or IBM 9 pin standard.   Modem control is NOT necessary.t  / APC UPS communicates at 2400 baud -- NOT 38400.c   --N VAXman- OpenVMS APE certification number: AAA-0001           VAXman@TMESIS.COM   ------------------------------  # Date: Thu, 06 Apr 2000 17:49:08 GMT0L From: winston@SSRL.SLAC.STANFORD.EDU ("Alan Winston - SSRL Admin Cmptg Mgr") Subject: Re: Pro*C on VMSe8 Message-ID: <009E833E.32DCB189@SSRL04.SLAC.STANFORD.EDU>  7 In article <38ECB5AB.7D2FAB17@bbc.co.uk>, Tim Llewellyn ! <tim.llewellyn@bbc.co.uk> writes:    >  >8 > E >Hey, WNT doesn't have a suported Posix interface last time I looked.  >s  D It's got a check-the-box-standard Posix API, and I have a friend who? worked for Softway Systems, a company whose purpose in life wassE to put a real Unix environment - shells, utilities, compilers - undersE NT.  I say "was" because they sold out to Microsoft and their productb( is supposed to be incorporated into W2K.   -- Alan   O ===============================================================================t0  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUM  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:  650/926-3056lM  Physical mail to: SSRL -- SLAC BIN 69, PO BOX 4349, STANFORD, CA  94309-0210 O ===============================================================================r   ------------------------------  $ Date: Thu, 6 Apr 2000 14:15:23 -0400' From: "Bill Todd" <billtodd@foo.mv.com>e Subject: Re: Pro*C on VMS?( Message-ID: <8cik56$6bi$1@pyrite.mv.net>  L "Alan Winston - SSRL Admin Cmptg Mgr" <winston@SSRL.SLAC.STANFORD.EDU> wrote= in message news:009E833E.32DCB189@SSRL04.SLAC.STANFORD.EDU... 9 > In article <38ECB5AB.7D2FAB17@bbc.co.uk>, Tim Llewellyn-# > <tim.llewellyn@bbc.co.uk> writes:A >2 > >1 > >1 > >0G > >Hey, WNT doesn't have a suported Posix interface last time I looked.h > >i >oF > It's got a check-the-box-standard Posix API, and I have a friend whoA > worked for Softway Systems, a company whose purpose in life was G > to put a real Unix environment - shells, utilities, compilers - underoG > NT.  I say "was" because they sold out to Microsoft and their producte* > is supposed to be incorporated into W2K.  K I would say that constituted considerably more than 'check-the-box' supporty7 on the part of W2K, then - if/when it actually appears.   H And I would also say that if POSIX were implemented in considerably moreL than just 'check-the-box' fashion on VMS, it would likely enjoy considerablyL less derision and considerably more use (though how much it was ever used isK debatable, since it was possible to download the documentation from the Web'A and use it without ever making DEC aware that you were doing to).l  F I've heard whispers, other than just the VMS fork support, that such aH usable POSIX might in fact be in the works, though not on a particularlyL accelerated schedule.  Gee, then it might not be that big an additional stepI to support Alpha Linux binaries on VMS - which would go a long way toward L erasing the growing VMS application deficit compared to more popular systemsK as well as allow near-complete masquerading as a Unix environment for those D whose needs were mostly in that area but who also valued things likeK performance, availability, and scalability.  The mind boggles - I hope it'siK not just a fantasy, and if not I hope it doesn't arrive too late to do much  good.    - bill   >t	 > -- Alan  >j >:L ============================================================================ ===g2 >  Alan Winston --- WINSTON@SSRL.SLAC.STANFORD.EDUA >  Disclaimer: I speak only for myself, not SLAC or SSRL   Phone:f 650/926-3056C >  Physical mail to: SSRL -- SLAC BIN 69, PO BOX 4349, STANFORD, CAe
 94309-0210 >gL ============================================================================ ===u >b   ------------------------------   Date: 6 Apr 2000 19:46:23 -0500 / From: jlauret@?.chem.sunysb.edu (Jerome LAURET)o0 Subject: Probably a stupid OPERATOR question .... Message-ID: <38ed21cf_2@dilbert.ic.sunysb.edu>  F 	Since my upgrade to OpenVMS7.2 from 6.2, the Operator (OPCOM) behavesM differently comparing to what it used to. I said "porbably a stupid question"iI in my subject because it sounds like maybe I have been missing some notesiO about it but this is what I observe now (and never before) and made me realizednN that I have a problem : OPCOM no longer reports messages from satellite nodes N and especially the login failures. Yes, I do have REPLY/ENABLE  turned on ...   . 	Any idea of what the problem is ? Any hints ?     -- b6                   Jerome LAURET S.U.N.Y. @ Stony Brook$        ,,,,,      Dept. of Chemistry+       ( o o )     Stony Brook NY 11794-3400-;   ---m---U---m---------------------------------------------s&   E-mail: jlauret@mail.chem.sunysb.edu<   URL   : http://nucwww.chem.sunysb.edu/jlauret/jlauret.html   ------------------------------  # Date: Fri, 07 Apr 2000 01:57:36 GMTw7 From: "David J. Dachtera" <djesys.nospam@earthlink.net>c3 Subject: Re: Problem with upgrade from 7.1 -> 7.2-1 - Message-ID: <38ED413E.C5306AFC@earthlink.net>p   Charlie Hammond wrote: > / > In article <sep2ms2qrfg7@corp.supernews.com>,   > "paul" <Paul@pwrr.com> writes: > M > >Our Alpha is currently running AXPvms 7.1 and we are attempting to upgrade M > >to 7.2-1. The upgrade went fine on our testbed/development machine (mirrorrI > >image of production machine), but when we go to upgrade the productionoK > >machine it asks about decnet, and mentions the new tcpip stack, and then N > >when it gets to the "do you always want detailed descriptions?" question itL > >fails on a "%noname-e-nomsg, message number 1652800a" error. Is there any > >way to debug this?e > M > Are the DECnet and TCP/IP questions resonable with respect to the installedLM > versions of DECnet and/or TCP/IP?  Please read the Upgrade and Installationa? > manual for more information, if you have not already done so.p > R > Regarding "%noname-e-nomsg, message number 1652800a" -- Without more informationM > it is difficult to provide any answers.  Please try to capture and post theeJ > EXACT output for a dozen lines or so up to the point where this error isG > displayed.  (And be  VERY sure you provide the message number EXACTLYt > as it displays.) > @ > Also, consider contacting your Compaq representative for help.  D There is known problem with the V7.1-2 upgrade - it can't be applied twice to the same system disk.  = This is due to a problem in PCSI (Ah, for good, old, reliable 
 VMSINSTAL!). b  F If this is in anyway related, you'd have to BACKUP/IMAGE the V7.2-1 CDF to a regular disk, apply the PCSI patch, then upgrade from the patched
 installation.i  = I don't know for sure - this is just a guess based on related- information / experience.o   --   David J. Dachterae dba DJE Systemsc" http://home.earthlink.net/~djesys/  : Unofficial Affordable OpenVMS Home Page and Message Board:+ http://home.earthlink.net/~djesys/vms/soho/y   ------------------------------   Date: 30 Mar 2000 15:50:40 GMT* From: helbig@astro.rug.nl (Phillip Helbig)Y Subject: Re: should BACKUP automatically dismount the first volume of a two-volume backupx. Message-ID: <8bvt4g$t46$1@info.service.rug.nl>  E In article <2000Mar30.084301.1@eisner>, koehler@eisner.decus.org (BobT Koehler) writes: h  ] > In article <8bv7a0$mdu$1@info.service.rug.nl>, helbig@astro.rug.nl (Phillip Helbig) writes:eJ > > When using BACKUP to write a multivolume saveset ON TWO DIFFERENT TAPEH > > DRIVES, should it automatically dismount the first volume when it isF > > finished (I can see the logic in this if only ONE drive is used).  > G > Does using a separate mount/foreign/nounload come close enough to the< > bahaviour you'ld like?   Thanks---I'll try it out.<   ------------------------------   Date: 30 Mar 2000 17:43:32 GMT* From: helbig@astro.rug.nl (Phillip Helbig)Y Subject: Re: should BACKUP automatically dismount the first volume of a two-volume backupd. Message-ID: <8c03o4$1p1$1@info.service.rug.nl>  5 In article <38E38884.9E8C6573@dnv.com>, Arne Bergsetho  <Arne.Bergseth@dnv.com> writes:    > Yes,F > The original benefit of this behaviour was appearent in the old days7 > when 9 track tape with 1600 fpi density was the rule.tD > An RP07 disk with 700 MByte could require nearly 20 reels of tape,@ > so with 2 tape drives you could mount reel number 3 as soon asK > the first the first reel had unloaded, while the second reel was rolling.e > And so on.........   Right.  I suspected as much.  H If I'm away for a while, I like to keep daily backups going.  Even if noF one is there to change the tape, I can just overwrite the previous oneH IF THE TAPE IS NOT UNLOADED FROM THE DRIVE.  Not as good as keeping (forF a while) the daily incremental backups, but better than going without.H (Note that since I do /SINCE=BACKUP but only do /RECORD with the monthlyD /IMAGE BACKUP, I only need the image and one incremental to restore,H rather than a whole series of them.  Thus, keeping daily incrementals isD only relevant for accidentally deleted files, not for disk crashes. H This scheme seems the best to me if not a huge amount of data is created% AND RETAINED between image backups.) i   ------------------------------  % Date: Thu, 30 Mar 2000 13:53:19 -0500a0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca> Subject: Re: Sign of the times.u. Message-ID: <38E3A29D.213BB5C@vl.videotron.ca>   Malcolm Dunnett wrote:E >    Unless something has changed recently, their continued existence.I > prevents Compaq customers in Canada from participating in a users groups" > which represents their interest.  M Their web site says that they are looking for speakers at a national event torI be held in MAY 2000. Obviously, they must be doing really well since theys aren't doing much publicity...L (They must have more than enough speakers and enough regostrants to fill all
 the rooms :-)f  I >    I am ( was? ) a paid-up member of the former DECUS Canada, yet at no-L > time did this group make any attempt to solicit my views on their proposedH > restructuring, nor have I heard anything about this new group prior to > JFs posting.    N I had covert discussions with a few people on starting an informal underground DECUS CanadaC (i'd call it DEQUS (tm) and would probably start off as just a mailrJ distribution list and if I can find a way to get a fixed IP, then we could have a web site etc etc.  L While that new entity seems fixated on a symposium, my goal for a user groupK would be to put the members in touch with each other for communicatiosn all0? the time (and perhaps foster get togethers at the local level).u  M What would it take to convince Compaq to recognize such a group ? Would it bec; possible to get some form of support from VMS engineering ?    ------------------------------  % Date: Thu, 30 Mar 2000 16:32:44 -0400 + From: Tim Shoppa <shoppa@trailing-edge.com>M Subject: Re: Sign of the times.M1 Message-ID: <38E381AC.1B134F32@trailing-edge.com>4   JF Mezei wrote:b > C > What is left of DECUS Canada is now called something like CANACU.d > : > The web site is www.canacu.org, and guess what it runs ? > R > Nop, not a webserver on VMS. Not a WEB server on True64, Not a web server on NT. > . > But it is running APACHE ON SOLARIS !!!!!!!!  @ The [US] DECUS FTP server, ftp.decus.org, has been an NT machine7 for the past year or so.  See my rants and raves from asC few months ago about how all the VMS-specific filetypes got mangled  in the process.i   Tim.   ------------------------------  $ Date: Thu, 6 Apr 2000 14:26:57 -0400' From: "Bill Todd" <billtodd@foo.mv.com> " Subject: Re: So who will buy VMS ?( Message-ID: <8cikqu$6pr$1@pyrite.mv.net>  8 Tim Llewellyn <tim.llewellyn@bbc.co.uk> wrote in message# news:38ECB661.B7934B4D@bbc.co.uk...DA > I think I should clarify my statement somewhat, after receiving A > email from a fellow comp.os.vms inhabitant with no A levels :-)h > = > I didn't mean to sound superior or anything. Its not that Io? > don't respect the skills of people without or with irrelevante > qualifications,0  # Er, I hope you meant 'credentials'.    - bill  (  more that the IT industry has become so= > "dumbed down" that people who really know very little abouto* > IT end up controlling the people who do. >e > Tim Llewellyn wrote: >h > > Arne Vajhj wrote: > >l@ > > >  But CS students are a relative small fraction of studentsB > > > and even a small fraction of students which after graduation' > > > are going to work with computers.0 > > >9 > >MC > > Yup, one ex-colleage uesd to joke that all you needed to get anuG > > IT job  these days is a history degree :-). The fact that he didn'ta> > > have any qualifications at all beyon A level should not be > > ignore in this context.i > >h > > >t@ > > > At least in Denmark other academic areas like engineering,D > > > business administration, economics etc. provide more people to, > > > the IT industry than computer science. > >t! > > And the natural sciences :-).e > >e > > >  > > >o* > > > I got a master degree in economics ! > > B > > Well, at least that is relevant to the maketplaces for IT :-). > >t > > >l > > > 
 > > > Arne > >. > > --: > > Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project4 > > MedAS at the BBC, Whiteladies Road, Bristol, UK.E > > Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.uk  > >eE > > I speak for myself only and my views in no way represent those of* > > MedAS or the BBC.  >a > --8 > Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project2 > MedAS at the BBC, Whiteladies Road, Bristol, UK.C > Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.ukn >tC > I speak for myself only and my views in no way represent those oft > MedAS or the BBC.F >X >I   ------------------------------  % Date: Thu, 06 Apr 2000 15:05:28 -0400s/ From: "Joseph Pomeroy" <jpomeroy@holycross.edu>UM Subject: solved! Re: How to determine via batch if another node is	reachable? / Message-ID: <s8eca7c7.093@dudley.holycross.edu>i  F Mr. Briggs has graciously pointed out that MU PING    WILL  respond if  C /number_of_packets  is specified.   How dumb can I get? It's in thec  
 HELP file!   Thanks,w   Joe Pomeroy    ------------------------------  " Date: Thu, 6 Apr 2000 20:27:16 GMT From: briggs@eisner.decus.orgiQ Subject: Re: solved! Re: How to determine via batch if another node is	reachable?y& Message-ID: <2000Apr6.162716.1@eisner>  a In article <s8eca7c7.093@dudley.holycross.edu>, "Joseph Pomeroy" <jpomeroy@holycross.edu> writes: H > Mr. Briggs has graciously pointed out that MU PING    WILL  respond if > E > /number_of_packets  is specified.   How dumb can I get? It's in them > " I forgot to tell you how I use it.  + $       multinet ping /number=3 /quiet 'p1'o $       save_status = $statuswA $       if save_status .ne. 1  .and.  save_status .ne. %x100022E0  $       then- $           call report "''p2'" "PING failed"w $       else+ $           call report "''p2'" "All clear"R
 $       endif   E That %x100022E0 is SS$_DATALOST which is a warning condition reported C if some packets made it but others were lost.  You may want to look ) at the other documented error conditions.   6 I always do these kinds of procedures under $ SET NOON   	John Briggs   ------------------------------   Date: 6 Apr 2000 19:39:42 -0500P/ From: jlauret@?.chem.sunysb.edu (Jerome LAURET)@ Subject: Re: SSH for VMS. Message-ID: <38ed203e_2@dilbert.ic.sunysb.edu>  H 	Current SSH client for VMS is available at http://www.free.lp.se/fish/.3 I have a redistributed client version on my site at9H http://nucwww.chem.sunysb.edu/htbin/software_list.cgi which is the FISH M (ssh client for VMS) original version with some patched applied (as discussedaK in the related mailing list but did not got in the offcial distribution due  to lack of time).2  ; 	OpenSSL is available at http://www.free.lp.se/openssl/ ...u   -- o6                   Jerome LAURET S.U.N.Y. @ Stony Brook$        ,,,,,      Dept. of Chemistry+       ( o o )     Stony Brook NY 11794-3400o;   ---m---U---m---------------------------------------------d&   E-mail: jlauret@mail.chem.sunysb.edu<   URL   : http://nucwww.chem.sunysb.edu/jlauret/jlauret.html   ------------------------------  % Date: Thu, 30 Mar 2000 05:36:38 -0800M5 From: "Larry D Bohan, Jr" <LBohan@dbc.spam_less..com> % Subject: Re: Suggestion for authorizeT2 Message-ID: <xFXjOKHo+=yUXdBO4RULKxzrIc0b@4ax.com>  2 On Thu, 30 Mar 2000 13:18:31 +0100, "POWERS, John" <John.POWERS@sema.co.uk> wrote:h  < >- So - **WHAT WE REALLY NEED** is a proper extension to the, >SETUAI to allow record creation / deletion.  > Agreed.  while at our site we have very few users, it would be? useful if there were ways to 'extend'  the UAI functions,  with  callouts or somesuch.l  A I guess I'm thinking of ways to 'add' per-user parameters perhapse: like UAI$_USER_DATA, that otherwise we would, say, add to ? job logicals tables in each login script, or do with right-ids.r  A And while *we* can use/modify the user data fields, I'm not sure  9 every site can, as they may have 3rd party s/w that does.T  > just a thought.  I imagine there are many more pressing issues that ought be worked on firstt   ------------------------------  % Date: Thu, 30 Mar 2000 15:29:30 +0100=- From: Tim Llewellyn <tim.llewellyn@bbc.co.uk> % Subject: Re: Suggestion for authorizes) Message-ID: <38E364CA.ADA7A0F4@bbc.co.uk>o   JF Mezei wrote:0   >3 >l> > ALL-IN-1 has it now. (ok, so it is called office server....) >   @   So, will it make you happy, JF, if Compaq bundle Office Server with base VMS licence :-)  --6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.ukm  A I speak for myself only and my views in no way represent those ofA MedAS or the BBC.n   ------------------------------  % Date: Thu, 30 Mar 2000 12:51:46 -0500<- From: "Peter Weaver" <peter.weaver@stelco.ca> % Subject: Re: Suggestion for authorize . Message-ID: <se7528cdni731@corp.supernews.com>  ! POWERS, John wrote in message ...k > ... E >The absence of this facility makes a terrible mess of any controllede= >way of allowing departmental managers the ability to add andnB >remove employees. It is particularly relevant, as the managerial,B >people-oriented staff tend not to be the confident tecchie-types.? >These people need a nice captive environment that will ask the.. >right questions and do all the work for them.  @ You should take a look at that OpenVMS Management Station that ID mentioned in a previous post. It is a supported product from DEC andD it is designed for "the managerial, people-oriented staff [who] tend' not to be the confident tecchie-types."s    C >                                              This can do whatever @ >the local requirements demand - add new UICs at end/fill in the >UIC blanks,  B It does have the problem I mentioned in the other post of startingD with member 0 if you tell it to pick the first available member. But it tries to do what you want.n  8 >            auto-create/delete directories if required,    
 It does this.i   >delete/forward emaild  = AFAIK it does not delete mail, but you can set the forwardingdD information as well as other mail settings by pointing and clicking.  7 >                      - ensuring that the manager onlye5 >creates/deletes users in their own group, etc. etc..   F That would be a nice addition to the product. It would also be nice toE have it restrict certain users so they can only modify certain fields D (i.e. the HELPDESK can only modify the password). Also allow certainD users to ADD/MODIFY/DELETE accounts but have other accounts that can only MODIFY.  C IIRC the thing is not smart enough to delete intrusion records whenS= you reset a password, but that would also be a nice addition.e   > ...L  ? If you do take a look at this note that we had to get a patchedAA version for our ES40's to fix some problem but I was only briefly=0 involved with that so I do not know the details.  F I wonder if anyone knows this question; does O.V.M.S. write to the UAF8 directly, or is it using some undocumented system calls?   ------------------------------  % Date: Thu, 06 Apr 2000 14:11:25 -040030 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>/ Subject: Re: Sun considers VMS a "mainframe" OSb/ Message-ID: <38ECD343.19D395A3@vl.videotron.ca>   L Ok, someone please explain to me why VMS/ALPHA is not considered in the same  class as IBM/390/MVS machines ?     Isn't Alpha a faster processor ?2 Do IBM disk drives have far superior performance ?H Is is a question of the bus which is much faster on IBM than on Alphas ?  H If VMS is at a performance disadvantage compared to the 1960s's IBM 360, 	1- how come ? 	2- will wildfire fix that ?   ------------------------------   Date: 6 Apr 2000 12:01:05 -0700o* From: dunnett@mala.bc.ca (Malcolm Dunnett)/ Subject: Re: Sun considers VMS a "mainframe" OSm, Message-ID: <B8Rte6x62XpO@malvm2.mala.bc.ca>  0 In article <38ECD343.19D395A3@vl.videotron.ca>, 4   JF Mezei <jfmezei.spamnot@vl.videotron.ca> writes:  N > Ok, someone please explain to me why VMS/ALPHA is not considered in the same" > class as IBM/390/MVS machines ?  > " > Isn't Alpha a faster processor ?4 > Do IBM disk drives have far superior performance ?J > Is is a question of the bus which is much faster on IBM than on Alphas ? > J    I think the last is the closest answer. I've always been told that it'sH the I/O architechture of "mainframes" which tends to differentiate them,  not things like processor speed.  I    However I suspect in this case "mainframe" is being used strictly in atD perjorative, rather than technical, sense ( as in "you don't want anG obsolete, legacy 'mainframe' O/S like VMS, you want a state-of-the-art,o 'open' O/S like Solaris" )   ------------------------------  $ Date: Thu, 6 Apr 2000 17:03:14 -07005 From: "cstranslations" <cstranslations@email.msn.com>  Subject: Tandem on EV7) Message-ID: <#d6XtSCo$GA.305@cpmsnbbsa04>h  K I don't keep up with all the rags and associated nonsense all that much butI anyway . . .  J . . . just got back from the LUG meeting. After listening to the splurb onK ProLiant and Windoz 2000 [when that wonderful "blue screen of death" popped L up in the middle of the alpha presentation it was good for a chuckle] we hadJ a presentation/road map for alpha/wildfire. It was some really cool stuff., But then he was preaching to the choir . . .  K In any event from what the man said EV7 is due out in about 15 to 18 months J running above 1200Mhz. When it comes out Tandem is going to run on it [notH that I'd know Tandem if I fell over it]. He also whipped out a new alphaI roadmap going all the way out to ev12 [due out about 2010 and running any I where between 7 and 9Ghz]. Now although this 9Ghz ev12 is the ultimate inaI vaporware I was wondering if any of this is being released to the generalaD public [or if it's just "feel good" stuff passed around internally]?   Jowo   ------------------------------  % Date: Thu, 30 Mar 2000 05:51:14 -0800/5 From: "Larry D Bohan, Jr" <LBohan@dbc.spam_less..com>p& Subject: Re: UCX's NTP vs. DST Changes2 Message-ID: <X1rjOIVLCctCiz2JFqhlGhbc1JcL@4ax.com>  @ On 30 Mar 2000 03:32:53 GMT, leslie@clio.rice.edu (Jerry Leslie) wrote:  9 >  o Are there any obvious problems with the procedures ?c   Don't think so.i  ; >  o Is it still necessary to disable the UCX NTP service ?n   in our humble experience, yes<      3 We too, prefer to do the DST/ST jumps in one leap,  . vs skewing the clock for some number of hours,  - I find the time-offset stuff vexing/annoying;i# I wish the world ran on UTC time;  C  ' this is our typical  Spring_forward.comI   $!'f$verify(1)
 $ set noon: $ if( f$getsyi( "CPU" ) .gt. 127 ) ! Alpha, assume UCX ... $ thenD $   if( f$search( "sys$startup:ucx$ntpd_shutdown.com;*" ) .nes. "" ) $   then* $       @sys$startup:ucx$ntpd_shutdown.com	 $   endifh $ else ! vax, assume Multinet  $   multinet configure/servers DISABLE NTP- RESTART- YES  EXIT $ endif2 $!4 $ @sys$manager:utc$configure_tdf.com SET "-420" "60" $!: $ if( f$getsyi( "CPU" ) .gt. 127 ) ! Alpha, assume UCX ... $ thenC $   if( f$search( "sys$startup:ucx$ntpd_startup.com;*" ) .nes. "" )> $   then$ $       mcr t_exe:do_modify /exact -(    sys$startup:ucx$ntpd_startup.com;0  -     "_TZ -028800"  -     "_TZ -025200" ) $       @sys$startup:ucx$ntpd_startup.com 	 $   endifa $ else ! vax, assume Multineta $   multinet configure SET TIMEZONE G  SET TIMEZONE-RULES US-Military/G EXIT $   multinet configure/servers
 ENABLE NTP RESTARTe YESs EXIT $ endif  $ exit $STATUS   ------------------------------  % Date: Thu, 30 Mar 2000 15:35:48 +0200.= From: Arne =?iso-8859-1?Q?Vajh=F8j?= <arne.vajhoej@gtech.com>l9 Subject: Re: VMS today: % users x-widnows vs VT terminalst) Message-ID: <38E35833.AD8E92F8@gtech.com>    JF Mezei wrote:E > David A Froble wrote:MT > > used for console terminals.  On the desktop, it's a PC with the biggest monitorsS > > we can find/afford running SmarTerm for accessing VMS, and mess office and suchi > > for other purposes.C > L > Ok, how come folks aren't running a X emulator on a PC to at least benefitN > from "windowed" editor (TPU) on VMS, as well as the ability to have multiple1 > windows open (multiple decterms for instance) ?m  E X gives some nice features, but as so often, then features comes at aN price.B In this case the price is memory usage and slowness. If you need 6	 DECterms,0E then a X emulator is a good choice, but if you only need 2, then 2 VTj emulator sessions is faster.e   Arne   ------------------------------  % Date: Fri, 07 Apr 2000 03:23:00 +0930 / From: Mark Daniel <mark.daniel@wasd.vsm.com.au>x Subject: VRC16 Monitor Problem/ Message-ID: <38ECCEFB.F5575D0D@wasd.vsm.com.au>u  3 Slightly off-topic (but it is used on a VMS system)e  E VRC16 monitor is "reluctant" to start.  When switched on the power-on ? indicator briefly lights, then there's a click and off it goes.s? Several iterations later it stays on.  Worse in colder weather.h Getting annoying.h  / Anyone encounter anything like it?  Simple fix?a    As always, thanks for your time.  E +-------------------------------------------------------------------+ D  Mark Daniel            Opinions my own ... and on loan from others.E  mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au)tE +-------------------------------------------------------------------+U   ------------------------------  % Date: Fri, 07 Apr 2000 03:25:10 +0930i/ From: Mark Daniel <mark.daniel@wasd.vsm.com.au>g Subject: VRC16 Monitor Problem/ Message-ID: <38ECCF7E.9EAC905E@wasd.vsm.com.au>e  3 Slightly off-topic (but it is used on a VMS system)o  E VRC16 monitor is "reluctant" to start.  When switched on the power-on0? indicator briefly lights, then there's a click and off it goes.u? Several iterations later it stays on.  Worse in colder weather.u Getting annoying.o  / Anyone encounter anything like it?  Simple fix?r    As always, thanks for your time.  E +-------------------------------------------------------------------+tD  Mark Daniel            Opinions my own ... and on loan from others.E  mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au)eE +-------------------------------------------------------------------+    ------------------------------  $ Date: Thu, 6 Apr 2000 18:01:35 -0400= From: "Mike Foley" <mike.foley@technologist-dot-com-spamspam> " Subject: Re: VRC16 Monitor Problem. Message-ID: <seq2a22qrfg86@corp.supernews.com>  G     Sounds like it's starting to go.. When they do that, it's somethinga@     internal. (Can't remember) It's cheaper to buy a new monitor$     nowadays than to fix an old one.     mike  < "Mark Daniel" <mark.daniel@wasd.vsm.com.au> wrote in message) news:38ECCF7E.9EAC905E@wasd.vsm.com.au...v5 > Slightly off-topic (but it is used on a VMS system)  >iG > VRC16 monitor is "reluctant" to start.  When switched on the power-on A > indicator briefly lights, then there's a click and off it goes.dA > Several iterations later it stays on.  Worse in colder weather.l > Getting annoying.i >t1 > Anyone encounter anything like it?  Simple fix?d >p" > As always, thanks for your time. >oG > +-------------------------------------------------------------------+tF >  Mark Daniel            Opinions my own ... and on loan from others.G >  mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au) G > +-------------------------------------------------------------------+  >w >y   ------------------------------  # Date: Fri, 07 Apr 2000 03:30:27 GMTr* From: "Jack Peacock" <peacock@simconv.com>" Subject: Re: VRC16 Monitor ProblemB Message-ID: <nDcH4.3987$y4.73254@newsread1.prod.itd.earthlink.net>  : Mark Daniel <mark.daniel@wasd.vsm.com.au> wrote in message) news:38ECCF7E.9EAC905E@wasd.vsm.com.au...i5 > Slightly off-topic (but it is used on a VMS system)a >aG > VRC16 monitor is "reluctant" to start.  When switched on the power-oncA > indicator briefly lights, then there's a click and off it goes.eA > Several iterations later it stays on.  Worse in colder weather.  > Getting annoying.o > 1 > Anyone encounter anything like it?  Simple fix?  >nK What resolution are you using to drive it?  I've seen them shut down if theeJ horizontal is too fast, I think the horz. oscillator crowbars to keep fromJ burning up.  Try it on a PC or multia set to 640x480 and see what happens.K Or it could be dying, mine has ever-increasing jitter, just retired it to adK place on the sacrificial components shelf for bench testing relatively more@D flaky boxes, most likely a Multia of indefinite age and reliability.     Jack Peacock   ------------------------------  % Date: Fri, 07 Apr 2000 11:56:52 +0800u- From: David B Sneddon <dbsneddon@bigpond.com>t" Subject: Re: VRC16 Monitor Problem+ Message-ID: <38ED5C84.37C32772@bigpond.com>a   Jack Peacock wrote:  > < > Mark Daniel <mark.daniel@wasd.vsm.com.au> wrote in message+ > news:38ECCF7E.9EAC905E@wasd.vsm.com.au...-7 > > Slightly off-topic (but it is used on a VMS system)s > >mI > > VRC16 monitor is "reluctant" to start.  When switched on the power-oneC > > indicator briefly lights, then there's a click and off it goes.cC > > Several iterations later it stays on.  Worse in colder weather.  > > Getting annoying.e > > 3 > > Anyone encounter anything like it?  Simple fix?n > >tM > What resolution are you using to drive it?  I've seen them shut down if the-L > horizontal is too fast, I think the horz. oscillator crowbars to keep fromL > burning up.  Try it on a PC or multia set to 640x480 and see what happens.M > Or it could be dying, mine has ever-increasing jitter, just retired it to adM > place on the sacrificial components shelf for bench testing relatively morerF > flaky boxes, most likely a Multia of indefinite age and reliability. >     Jack Peacock  C The VRC16 uses three BNC connectors (R, G and B) at the monitor endyH and a rather non-standard connector at the video card end... not readily connectable to your average PC.f -- _ Regards, Dave.nI -------------------------------------------------------------------------pI David B Sneddon (dbs)  OpenVMS Systems Programmer   dbsneddon@bigpond.comrI DBS software at ...   http://www.users.bigpond.com/dbsneddon/software.htmnI "Life is what happens to you while you're busy making other plans" Lennonn   ------------------------------  $ Date: Thu, 6 Apr 2000 13:07:32 -0500 From: "Gil" <hamiltone@fwi.com> 1 Subject: Re: Warning: circuitsurgeon is a ripoff!-. Message-ID: <sepkm7ktrfg13@corp.supernews.com>  A The unit was working PERFECTLY when it left here on July 20 1999.eD DAN then screwed up the hard drive, which I replaced at no cost evenA though DAN did the damage. $35 to send hard drive FEDEX. This wasr  - August 13,1999.  He then stated it works now.wB And as for stating a repair cost of $500, get real. The machine isC almost ten years old. And currently not worth that much as a whole.eG  As for a warrantee, The ad stated all sales final. That means when youoE get it and it works, it is yours permanently. You said it arrived andlB worked. And not know what Dans capabilities are, I am not going to    replace everything DAN dorks up.F  And as for not answering DANS emails, I did. I did not however answerC any more of DANS VMS based questions, I simply do not have time fore that.cG  DAN called several times a day for VMS questions. This went on for twot  D weeks. DAN said he was plugging in console ports to log in and could  B not. DAN then wanted to know where certain VMS files were located.D I do have other things to do than TEACH some body who claims to be a  ? SYSTEM ADMINISTRATOR about how VMS works on an Alpha. I stopped"  1 answering his VMS questions and he went away mad.W2  To email me 8 (eight) months later is ridiculous.G  As for his BROKEN MACHINE, I offered a replacement board for DANS unita  D at $200. There would be NO warrantee for this. It is in good working  D order as I would have to pull it out of a working machine. (This was  G when I got the email from wetboy telling me about DANS absurd post) Nowt  C seeing that he is posting LIES I retract this offer to him and willt  E never deal with him again and advise all to be aware that if you deali  E with him you may get a posting from him on the newsgroups like I did.   D There are some people in this world that do NOT want to be helped or  I fixed, these kind of people just want to spread misery amonst the rest ofh the world.  I have  K sent a lot of stuff to Canada with out any troube at all. So it is not  allt of  $ Canada that is screwed up, Just DAN.   ------------------------------  % Date: Thu, 06 Apr 2000 12:28:35 -0700k+ From: Tom Crabtree <gbsci@gbscientific.com> 1 Subject: Re: Warning: circuitsurgeon is a ripoff!f0 Message-ID: <38ECE563.60997FE0@gbscientific.com>  , This is a multi-part message in MIME format.& --------------A98552D40AC3AF13561E5DE0* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bitm   In defense of Circuitsurgeon,y  @ We have purchased 2 systems from this individual and have had noH problems whatsoever.  Admittedly, they were used systems which were fullF of dust elephants (larger untamed relatives of dust bunnies), but they2 were cheap and they worked, and are still working!  G I'm hoping that Dan just had a one-off problem, which the good folks atlG Circuitsurgeon will come to their senses and fix.  Perhaps a compromiseaH is in order here?  How about Dan offer to pay for shipping, or something
 in this vein?e   Tom C. GB Scientific, Inc.s    
 wetboy wrote:  > ) > In comp.os.vms Dan <dan@vrx.net> wrote:d > 
 > < snip > > ? > I have sent a copy of Dan's post to the "circuitsurgeon", andf4 > invited him to respond with his side of the story. >  > -- Wetboyh& --------------A98552D40AC3AF13561E5DE0- Content-Type: text/x-vcard; charset=us-ascii;n  name="gbsci.vcf"i Content-Transfer-Encoding: 7bite* Content-Description: Card for Tom Crabtree  Content-Disposition: attachment;  filename="gbsci.vcf"    begin:vcard  n:Crabtree;Tom tel;fax:415-898-7514 tel;work:415-898-7606n x-mozilla-html:TRUE  url:http://www.gbscientific.comn org:GB Scientific, Inc.t version:2.1m% email;internet:gbsci@gbscientific.comt title:Tom CrabtreeG adr;quoted-printable:;;701 DeLong Ave.=0D=0ASuite K;Novato;CA;94945;USAe	 end:vcardr  ( --------------A98552D40AC3AF13561E5DE0--   ------------------------------  # Date: Fri, 07 Apr 2000 03:04:35 GMTn From: Dan <dan@vrx.net>s1 Subject: Re: Warning: circuitsurgeon is a ripoff! ' Message-ID: <38ED50B1.ECE5E553@vrx.net>m  
 Gil wrote:  C > The unit was working PERFECTLY when it left here on July 20 1999. F > DAN then screwed up the hard drive, which I replaced at no cost evenC > though DAN did the damage. $35 to send hard drive FEDEX. This wasE  L LIAR! I did NOT screw up the drive, and you know it. You sent me a defectiveE one. You offered to replace it but ONLY if I paid shipping both ways!e  / > August 13,1999.  He then stated it works now.rD > And as for stating a repair cost of $500, get real. The machine isE > almost ten years old. And currently not worth that much as a whole. I >  As for a warrantee, The ad stated all sales final. That means when you=G > get it and it works, it is yours permanently. You said it arrived andkD > worked. And not know what Dans capabilities are, I am not going to  ? LIE again. I was given that cost by many others as repair cost.-5 Remember, you charged me MORE for the unit than that!FB You stated you would replace it if it was defective, which it was.? But then you ignored my calls (hanging up on me) and my emails.o  " > replace everything DAN dorks up.  ) You mean defective equipment you sold me.a  I > Gil also has complaints on Ebay about these deals, so it's not just me.I  E > seeing that he is posting LIES I retract this offer to him and willu  ! and seeing YOUR LIES posted here.   G > never deal with him again and advise all to be aware that if you deal  > G > with him you may get a posting from him on the newsgroups like I did.-  " yeah unlike YOU who commits FRAUD!  F > There are some people in this world that do NOT want to be helped or >eK > fixed, these kind of people just want to spread misery amonst the rest ofa > the world.  I have   Yeah people you RIP OFF!M so you "retracted" your offer? but then in your email to me privately you saynJ you want to replace the board still? What kind of screwed up story are you telling  here?   M > sent a lot of stuff to Canada with out any troube at all. So it is not  alln > of >s& > Canada that is screwed up, Just DAN.  9 No, just you. Legal action is following Gil, be prepared!e  F BTW: I've also sent a complaint to DECUS suggesting your membership beJ suspended, and I'm offering all concerned your FRAUD as all the proof they need. J I've also complained to Yahoo, and any other sources I can find Gils email being used.,  A I've complained to sources in the US, including the ISP that runseH circuitsurgeon, upstream providers FWCC and FWI, I'll also be contactingK whatever you have in the US in the way of BBB, state department, etc. and I0@ will be filing a civil lawsuit against Circuitsurgeon for FRAUD.  M Gil is a ripoff artist and conman, he lies repeatedly, he promises to replaceeC defective goods he ships out, but then refuses to. He is true SCUM.n   Dan.   ------------------------------  # Date: Fri, 07 Apr 2000 03:06:26 GMTl From: Dan <dan@vrx.net>o1 Subject: Re: Warning: circuitsurgeon is a ripoff!p' Message-ID: <38ED5122.5705DCE0@vrx.net>s   You mean his lies?   Dan.  
 wetboy wrote:r  ) > In comp.os.vms Dan <dan@vrx.net> wrote:s >m
 > < snip > >t? > I have sent a copy of Dan's post to the "circuitsurgeon", andt4 > invited him to respond with his side of the story. >m > -- Wetboyn   ------------------------------  # Date: Fri, 07 Apr 2000 03:11:17 GMTt From: Dan <dan@vrx.net>l1 Subject: Re: Warning: circuitsurgeon is a ripoff! ' Message-ID: <38ED5243.E9A8ACB7@vrx.net>o  H I offered that, but Gil stated here he refuses that, however, in private email K he agrees to it later. He's saying contradictory things here, I can't trustt him.  > He lies in public and in private, how should I deal with that.  E BTW: further analysis shows the IO board is not the problem. it's thea motherboard.K I'll have to take it in to the local Compaq/DEC office to get fully checked: out.K Once I've done that, I'll know how much of a lemon I've been stuck with andn= how much it will cost to fix it (or replace defective parts).   J Gil is not technically inclined, he knows very little about vax systems beH it vms or alphas, he just knows how to make a buck and have the customerC worry about the problems later. He gives you a gaurantee, then saysT everything is "as is".  K I'd gladly accept a replacement but after his post he'll have to cross-shipfI components, in other words, after I verify he's not sending me crap, I'lli5 send the defective stuff back to him. No other deals.    Dan.   Tom Crabtree wrote:7   > In defense of Circuitsurgeon,  > B > We have purchased 2 systems from this individual and have had noJ > problems whatsoever.  Admittedly, they were used systems which were fullH > of dust elephants (larger untamed relatives of dust bunnies), but they4 > were cheap and they worked, and are still working! >tI > I'm hoping that Dan just had a one-off problem, which the good folks atuI > Circuitsurgeon will come to their senses and fix.  Perhaps a compromise)J > is in order here?  How about Dan offer to pay for shipping, or something > in this vein?. >; > Tom C. > GB Scientific, Inc._ >- > wetboy wrote:  > >s+ > > In comp.os.vms Dan <dan@vrx.net> wrote:l > >v > > < snip > > >aA > > I have sent a copy of Dan's post to the "circuitsurgeon", andi6 > > invited him to respond with his side of the story. > >T
 > > -- WetboyT   ------------------------------  " Date: Fri, 7 Apr 2000 04:38:20 GMT From: coy@eisner.decus.org1 Subject: Re: Warning: circuitsurgeon is a ripoff!o& Message-ID: <2000Apr7.003820.1@eisner>  D I think I've lost track of some things here.  Could you answer a few questions, please?  6 1.  When did you buy this equipment, and for how much?  0 2.  When did you find out that it didn't work?    E 3.  What parts have been replaced so far, and what did they cost you?m    A In article <38ED5122.5705DCE0@vrx.net>, Dan <dan@vrx.net> writes:s > You mean his lies? >  > Dan. >  > wetboy wrote:t > * >> In comp.os.vms Dan <dan@vrx.net> wrote: >> >> < snip >o >>@ >> I have sent a copy of Dan's post to the "circuitsurgeon", and5 >> invited him to respond with his side of the story.r >> >> -- Wetboy >    ------------------------------  $ Date: Thu, 6 Apr 2000 23:14:19 +0400# From: "Suslin Alex" <alxsus@aha.ru>h( Subject: Where are TXAn: terminal ports?* Message-ID: <8cio7f$1749$1@gavrilo.mtu.ru>   Hi allH I'm trying to recover old MicroVAX 3100. I've found three TTAn: terminalJ ports. But where are eight TXAn: ports? There is a small box with 8 DEC423L sockets and a thin cable. I suspect that they should be used, but don't know- how. What settings must be done? Help please.t	 Thank your Alex Suslin (alxsus@aha.ru)r   ------------------------------  $ Date: Thu, 6 Apr 2000 17:05:08 -04002 From: "Richard B. Gilbert" <DRAGON@compuserve.com>( Subject: Where are TXAn: terminal ports?7 Message-ID: <200004061705_MC2-A03F-53AD@compuserve.com>a  J         The "harmonica" connects to a connector on the rear of the MicroV= AXJ 3100 if the MicroVAX is so equipped.  I seem to recall that the TXA ports=  J were optional.  The connector is located on the rear of the machine, at t= heH upper left, and above the "Standard Ethernet Connector" (AUI Connector).  % Message text written by "Suslin Alex"0J >I'm trying to recover old MicroVAX 3100. I've found three TTAn: terminal=  J ports. But where are eight TXAn: ports? There is a small box with 8 DEC42= 3eG sockets and a thin cable. I suspect that they should be used, but don't  know- how. What settings must be done? Help please.o
 Thank you<   ------------------------------  $ Date: Thu, 6 Apr 2000 20:57:30 -0400+ From: "David Turner" <d_b_turner@yahoo.com>r; Subject: Why are there no French people in this newsgroup ? / Message-ID: <seqcfl5mrfg119@corp.supernews.com>    This is a weird deal....  I Netherlands, Germany, US UK, Ozzy's;  pretty much everyone, but no Frenchs  
 Strange huh ?n  ! Do they still use VMS en France ?e# Or PC's only  - Vive Monsieur Bill!eH Actually - if one tries to pronouce Vms with a french accent, it doesn't' sound 'arf as good as Solaris or Unix --     Vive Monsieur Bill!2  F There again, they make their own planes, build their own nuclear powerI plants, and discuss the fundamentals of other countries as if they didn'tr have running water...   J Seriously though... I wander why there are never any questions or requests posted herer  , Do they have their own "elitist" newsgroup ?  D I spent almost 4 years of college hell learning French, got into theD computer business and communicate less in French than I do in German   David      -- David Turner Island Computers US Corporation  2700 Gregory Streets Savannah GA 31404r Tel: 912 447 6622n Fax:912 201 0096 sales@islandco.com   ------------------------------  $ Date: Thu, 6 Apr 2000 20:36:47 -0500) From: "John E. Malmberg" <wb8tyw@qsl.net>e? Subject: Re: Why are there no French people in this newsgroup ?e. Message-ID: <seqekcumrfg14@corp.supernews.com>  8 David Turner <d_b_turner@yahoo.company> wrote in message) news:seqcfl5mrfg119@corp.supernews.com...  > This is a weird deal.... > K > Netherlands, Germany, US UK, Ozzy's;  pretty much everyone, but no French. >. > Strange huh ?- >-# > Do they still use VMS en France ?- <snip>L > Seriously though... I wander why there are never any questions or requests
 > posted here   K This one seem to show up a lot: http://www.multimania.com/pmoreau/ to offer 
 solutions.  I There is at least one other French Web site that has been posted to offeraG solutions, but I could not find any English translations on it's pages.o   -John  wb8tyw@qsl.network  < I also have French Canadian ancestry if that means anything.   ------------------------------  % Date: Thu, 06 Apr 2000 18:39:50 -0700s5 From: "Larry D Bohan, Jr" <LBohan@dbc.spam_less..com>r? Subject: Re: Why are there no French people in this newsgroup ?n2 Message-ID: <AzztODjHT2DShecd+hu9KKrIiYjg@4ax.com>  1 On Thu, 6 Apr 2000 20:57:30 -0400, "David Turner"i <d_b_turner@yahoo.com> wrote:n   >This is a weird deal....aJ >Netherlands, Germany, US UK, Ozzy's;  pretty much everyone, but no French   there's always  fr.comp.os.vms  s2 It doesn't look too busy/active, but I recognized % one ot two names from C.O.V, even so.    ------------------------------   End of INFO-VAX 2000.194 ************************