1 INFO-VAX	Thu, 04 Oct 2001	Volume 2001 : Issue 551       Contents: Re: Alternate vendor memory. Re: Alternate vendor memory. ANN: JUMP utility updated  Re: Ben Rosen the real culprit?  Re: Ben Rosen the real culprit?  Re: Ben Rosen the real culprit?  Re: Ben Rosen the real culprit?  Re: Ben Rosen the real culprit? 6 Re: BIG differences in Vax vs ALpha compile/link timesE Re: Compaq revenue goes down $1 Billion, warns of loss in 3rd Quarter  Re: Default password in RA3000- Re: DLT IV Tapes - How long should they last? 6 Re: Fabbing new VAX? (was: Re: VAX disaster tolerance)6 Re: Fabbing new VAX? (was: Re: VAX disaster tolerance)6 Re: Fabbing new VAX? (was: Re: VAX disaster tolerance)6 RE: Fabbing new VAX? (was: Re: VAX disaster tolerance)6 Re: Fabbing new VAX? (was: Re: VAX disaster tolerance)L Re: FAQ updates (was: RE: OpenVMS Frequently Asked Questions (FAQ), Part 4/5 Free Home Appraisal  Re: Good bye VAX-6400's C Re: I WILL QUIT USING VMS WHEN THE MILITARY DOES - IN 2015 OR MAYBE C Re: Itanium and Bi-endianism (was: Re: Ben Rosen the real culprit?)   Re: Letter from Michael Cappelas  Re: Letter from Michael Cappelas  Re: Letter from Michael Cappelas  Re: Letter from Michael Cappelas  Re: Letter from Michael Cappelas  Re: Letter from Michael Cappelas  Re: Letter from Michael Cappelas, Re: Locate Filename containing logical block, Re: Locate Filename containing logical block, Re: Locate Filename containing logical blockA Re: Microsoft Addresses Security Fears: Chance To Market VMS/CSWS = Microsoft Addresses Security Fears: Chance To Market VMS/CSWS A Re: Microsoft Addresses Security Fears: Chance To Market VMS/CSWS & Need help with $QIO calls in C program* Re: Need help with $QIO calls in C program Re: Need help with RMS# Re: NETSCAPE 6(.1)NEXT on OpenVMS ?  Re: NT or w2k on alpha 1200 ?  Re: ODS-2 for FreeBSD 6 RE: OpenVMS Frequently Asked Questions (FAQ), Part 4/5 Pathworks on Win20009 PC164, OVMS7.2, isacfg and graphic cards (MACH64, S3 968) = Re: PC164, OVMS7.2, isacfg and graphic cards (MACH64, S3 968)  PGP for OpenVMS??  Re: PGP for OpenVMS?? : Re: Problem With VMS Backup Save-Set Spanning Tape Volumes: Re: Problem With VMS Backup Save-Set Spanning Tape Volumes. Re: Problem with vms712_update-v0300 download.' Project Manager - EMC Storage Architect " Re: Public domain VMS clustering ?" Re: Public domain VMS clustering ? Purveyor  RE: Reset error account on Alpha  RE: Reset error account on Alpha  RE: Reset error account on Alpha  Re: Reset error account on Alpha  Re: Reset error account on Alpha  Re: Reset error account on Alpha  Re: Reset error account on Alpha' Re: Starting DECwindows on DEC 3000-600  Re: TCPIP 5.1 Telnet Problem Re: VAX disaster tolerance( Re: VAX-Alpha Migration: COBOL/BASIC/FMS Re: VMS 7.3 versus 7.2-2 Re: VMS 7.3 versus 7.2-2/ Re: Warranty on ES45 reduced from 3 years to 1? / Re: Warranty on ES45 reduced from 3 years to 1? . Re: [AMDS V7.3A] Problem on OpenVMS VAX V7.3 ?  Re: [LYRIS] OpenVMS Patches List  F ----------------------------------------------------------------------   Date: 3 Oct 2001 21:39:51 GMT * From: bleau@umtof.umd.edu (Lawrence Bleau)% Subject: Re: Alternate vendor memory. ) Message-ID: <9pg0j7$lp5$1@hecate.umd.edu>   u In article <ed1713eb.0109301405.15d7d546@posting.google.com>, graham@the-shades.demon.co.uk (Graham Harrison) writes:  > ' >Has anyone got any experience of using & >Kingston compatable memory?  I put a ( >purchase request in for some memory for( >a DS20E and the bean counters tried to 
 >get me this.   L I used Kingston memory in some older workstations.  It has never failed me. M They have a lifetime warranty, compared to Compaq's which is only a few years  at best.  7 >If Compaq don't support it, or it will cancel/increase 7 >my service contract then fair enough and Compaq memory 9 >it is - otherwise how do I justify it to the big cheese?   M As for the service contract, what I did on the newer workstations is buy them O with a small amount of Compaq memory, then load up the rest of the memory space O with third party memory.  If there's ever a question, the FE can remove the 3rd K party stuff and reproduce the problem.  Besides, you can always get a third N party service firm instead of Compaq that'll give you better prices all around> *and* service 3rd party equipment as well.  That's what we do.   Lawrence Bleau University of Maryland" Physics Dept., Space Physics Group 301-405-6223 bleau@umtof.umd.edu    ------------------------------  % Date: Thu, 04 Oct 2001 07:51:25 +0200  From: Dirk Munk <munk@home.nl>% Subject: Re: Alternate vendor memory. ' Message-ID: <3BBBF8DD.F17F2896@home.nl>   ; The problem is, there is no such thing as "Kingston memory" ? Kingston buys memory from several sources, and sticks their own  brandname & partnumber on it. H So a certain Kingston partnumber may contain memory modules from brand X' one day, and from brand Y the next day. E Go for modules from a 'real' manufacturer, like Camintonn or Dataram, H these companies were even supported by Compaq (once). Camintonn has a UK office.        Graham Harrison wrote:   > Hi,  > & > Last question this evening - honest. > ( > Has anyone got any experience of using& > Kingston compatable memory?  I put a) > purchase request in for some memory for ( > a DS20E and the bean counters tried to > get me this. > . > I wouldn't normally credit the bean counters2 > cost saving with much worth - but the difference. > is staggering.  5548 versus 400 (yes 400) > 8 > If Compaq don't support it, or it will cancel/increase8 > my service contract then fair enough and Compaq memory: > it is - otherwise how do I justify it to the big cheese? > 	 > Thanks. 	 > Graham.    ------------------------------  # Date: Wed, 03 Oct 2001 18:21:35 GMT - From: goathunter@goatley.com (Hunter Goatley) " Subject: ANN: JUMP utility updated0 Message-ID: <3bbb54e0.25633108@news.process.com>  C Jonathan Ridler's JUMP V4.0 lets authorized users look like another C user or login a la HGLOGIN.  Unlike HGLOGIN, JUMP includes a number @ of security features, allowing the use of JUMP to be documented,= logged, etc.  JUMP V4.0 includes a number of new features and ! considerably more error-checking.    You can get it here:   http://www.process.com/openvms/   4 ftp://ftp.process.com/vms-freeware/fileserv/jump.zip9 http://vms.process.com/ftp/vms-freeware/fileserv/jump.zip 0 ftp://ftp.tmk.com/vms-freeware/fileserv/jump.zip5 http://www.tmk.com/ftp/vms-freeware/fileserv/jump.zip   . And on the other mirrors in the next 24 hours.   Hunter ------9 Hunter Goatley, Process Software, http://www.process.com/ 9 goathunter@goatley.com     http://www.goatley.com/hunter/    ------------------------------  $ Date: Wed, 3 Oct 2001 14:31:04 -0400' From: "Bill Todd" <billtodd@foo.mv.com> ( Subject: Re: Ben Rosen the real culprit?( Message-ID: <9pflde$c6s$1@pyrite.mv.net>  3 "Peter da Silva" <peter@abbnm.com> wrote in message  news:9pdq8q$83q@web.nmti.com... . > In article <3BB7D087.D609816A@videotron.ca>,1 > JF Mezei  <jfmezei.spamnot@videotron.ca> wrote: K > > Consider that unless proven otherwise, Tru64 is as good as dead. Once a I > > customer accepts that, he also accepts that full migration to another  platformC > > will be necessary at which point endianness becomes irrelevant.  > " > I think endianness is overrated. > C > I have been heavily involved in porting a major (millions of LOC)  application G > from 16-bit little endian and 36-bit 1s-complement hardware to 32-bit L > big-endian to 64-bit little-endian... with detours to 32-bit little-endianI > and back to 32-bit big-endian... with some of the development on 16-bit J > big-endian and 16-bit word-oriented machines. Word size is a much biggerL > issue than byte order, if you've paid any attention at all to portability.  K And that's the point:  without examining the code, you don't know.  Even if K the code *seems* to run fine, there's no guarantee that you've executed all 
 its paths.  H Endianness is indeed a relatively minor problem when porting.  The issueK here is not the difficulty of the port, but whether the port can in any way K be considered transparent - and unless you define examination of every line F of code (even if you find nothing needs changing) as 'transparent', it can't.   - bill   ------------------------------   Date: 4 Oct 2001 02:16:26 GMT & From: peter@abbnm.com (Peter da Silva)( Subject: Re: Ben Rosen the real culprit?% Message-ID: <9pggpq$msn@web.nmti.com>   = In article <WFvu7.68493$vq.12912073@typhoon.ne.mediaone.net>, 3 Terry C. Shannon <terryshannon@mediaone.net> wrote: N > The easier approach would be to graft TruCluster and Tru64 RAS features ontoM > HP-UX. Problem is, HP-UX is IIRC the last of the old-line monolithic kernel N > UNIX implementations. Gonna have to upgrade it to a modular kernel sometime;' > this might be a good time to do so...   M I wasn't aware of that... there's no kernel module mechanism at all in HP/UX?    --  +  `-_-'   In hoc signo hack, Peter da Silva. E   'U`    "A well-rounded geek should be able to geek about anything." L                                                        -- nicolai@esperi.org          Disclaimer: WWFD?   ------------------------------   Date: 4 Oct 2001 02:19:47 GMT & From: peter@abbnm.com (Peter da Silva)( Subject: Re: Ben Rosen the real culprit?% Message-ID: <9pgh03$mt0@web.nmti.com>   , In article <3BBABA98.2B4C0D77@videotron.ca>,/ JF Mezei  <jfmezei.spamnot@videotron.ca> wrote: N > This is why I strongly suspect that HP-UX will survive and will just receive > new stuff from Tru64.   . Including, as Terry implies, the Tru64 kernel?  0 Tru64 has a very BSD-ish API. What's HP/UX like?  J > Does VMS still share parts of the TCPIP stack with Tru64 ?  With Tru64'sJ > death, will VMS engineering have to take over all of the TCPIP stack andP > maintain/improve it by themselves, or would they choose to use the HP-UX TCPIP > stack for VMS ?   O I would hope they would use the 4.3-Reno-based Tru64 stack, since it was almost N the last fork from the original BSD networking implementation before the final 4.4 release from CSRG.   --  +  `-_-'   In hoc signo hack, Peter da Silva. E   'U`    "A well-rounded geek should be able to geek about anything." L                                                        -- nicolai@esperi.org          Disclaimer: WWFD?   ------------------------------  $ Date: Wed, 3 Oct 2001 20:17:18 -0700( From: "Greg Cagle" <gregc@gregcagle.com>( Subject: Re: Ben Rosen the real culprit?/ Message-ID: <trnl5tio3hfd43@corp.supernews.com>   S "Peter da Silva" <peter@abbnm.com> wrote in message news:9pggpq$msn@web.nmti.com... ? > In article <WFvu7.68493$vq.12912073@typhoon.ne.mediaone.net>, 5 > Terry C. Shannon <terryshannon@mediaone.net> wrote: P > > The easier approach would be to graft TruCluster and Tru64 RAS features ontoO > > HP-UX. Problem is, HP-UX is IIRC the last of the old-line monolithic kernel P > > UNIX implementations. Gonna have to upgrade it to a modular kernel sometime;) > > this might be a good time to do so...  > O > I wasn't aware of that... there's no kernel module mechanism at all in HP/UX?   L Not as such; there's a thing called Dynamically Loadable Kernel Modules thatI is just coming online but is nowhere near what is available on other Unix  implementations.   --
 Greg Cagle gregc at gregcagle dot com   ------------------------------  $ Date: Wed, 3 Oct 2001 20:21:40 -0700( From: "Greg Cagle" <gregc@gregcagle.com>( Subject: Re: Ben Rosen the real culprit?/ Message-ID: <trnle2svpcrf90@corp.supernews.com>   S "Peter da Silva" <peter@abbnm.com> wrote in message news:9pgh03$mt0@web.nmti.com... . > In article <3BBABA98.2B4C0D77@videotron.ca>,1 > JF Mezei  <jfmezei.spamnot@videotron.ca> wrote: P > > This is why I strongly suspect that HP-UX will survive and will just receive > > new stuff from Tru64.  > 0 > Including, as Terry implies, the Tru64 kernel? > 2 > Tru64 has a very BSD-ish API. What's HP/UX like?  K HP-UX 11 is SVR4 but has lots of BSD legacy as well. Previous versions were 
 more BSD-ish.    --
 Greg Cagle gregc at gregcagle dot com   ------------------------------  $ Date: Wed, 3 Oct 2001 22:12:48 -0400) From: "Neil Rieck" <n.rieck@sympatico.ca> ? Subject: Re: BIG differences in Vax vs ALpha compile/link times ; Message-ID: <HwPu7.17744$4l5.2396609@news20.bellglobal.com>   < "Robert Young" <rcyoung@aliconsultants.com> wrote in message7 news:91437ce6.0110011336.3fcc1364@posting.google.com... / > Two systems: Vax 4000-600 & Alphaserver 2100a  > E > Compaq/DEC C compiler on both.  V6.4 on Alpha, V6.0 on Vax. We just H > installed the Alpha this weekend so that is the reason for the version > mismatch.  > B > Vax (VMS 7.1, ?MHz, 128Mb memory, RF-74 DSSI drives)compiles andE > links program in 14 minutes. On Alpha (VMS 7.3, 250MHz-2 processor, C > 512Mb memory, RZ29D-VWs) it takes 90 minutes for same EXACT code,  > same linker file, etc. > G > Code is written/compiled/linked first on Vax, then copied on Alpha to F > be recompiled/linked there so we have versions for both platforms at > this time. > E > Any suggestions/recommendations would be helpful where we could get J > the work on Alpha done faster? It seems that the discrepancy is just too/ > large between the systems. Thanks in advance.  >  > Respond to email if possible  K Compile times can be reduced by lowering the optimization level (but I only I do this on my final compile before moving the program into production). A 5 similar switch can be found on the Compaq-C compiler.   L I also found that adjusting "sysgen" and "authorize" parameters could make aK world of difference. For example, compiling a 40k line Compaq-BASIC program F on an old AlphaServer-4100 (EV5 / 300MHz) yielded the following times:  : 1. before tuning; over 24 hours using optimization level 47 2a. after tuning; 42 minutes using optimization level 4 7 2b. after tuning; 11 minutes using optimization level 0   L p.s. When developing code I always compile at level 1 which yields a greaterE number of useful "informational" and "warning" messages than level 0.    See this link for more info:J http://www3.sympatico.ca/n.rieck/docs/alpha_diary.html#compiler_benchmarks  
 Neil Rieck Kitchener/Waterloo/Cambridge,  Ontario, Canada.! http://www3.sympatico.ca/n.rieck/    ------------------------------  # Date: Wed, 03 Oct 2001 18:45:19 GMT 4 From: "Terry C. Shannon" <terryshannon@mediaone.net>N Subject: Re: Compaq revenue goes down $1 Billion, warns of loss in 3rd Quarter= Message-ID: <31Ju7.73330$vq.13559095@typhoon.ne.mediaone.net>   @ "Doc.Cypher" <doc_cypher@redneck.gacracker.org> wrote in message0 news:20011003163031.17382.qmail@gacracker.org...$ > -----BEGIN PGP SIGNED MESSAGE----- > L > On Wed, 03 Oct 2001, "Terry C. Shannon" <terryshannon@mediaone.net> wrote:C > >"Doc.Cypher" <doc_cypher@redneck.gacracker.org> wrote in message  > B > >> I always thought that one was CEA (Chief Excrement Agitator). > >> > >  > >Oops, forgot that one... ;-}  > L > Quite a few people in here would like to agitate some excrement at Compaq. >   
 No doubt! ;-}    ------------------------------  $ Date: Thu, 4 Oct 2001 08:41:42 +05004 From: Valentin Likoum <valentin.likoum@ncc.volga.ru>' Subject: Re: Default password in RA3000 3 Message-ID: <842684074.20011004084142@ncc.volga.ru>   : On 03.10.2001 www.islandco.com <sales@islandco.com> wrote:  J > Was wandering  what is the default "admin" password for a RZ3000 (HSZ22) > when new in from COmpaq     DECRAID   --     Valentin Likoum    valentin.likoum@ncc.volga.ru   ------------------------------  % Date: Wed, 03 Oct 2001 22:22:42 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> 6 Subject: Re: DLT IV Tapes - How long should they last?' Message-ID: <3BBBD602.4A08A99F@fsi.net>    Paul Repacholi wrote:  > 5 > "David J. Dachtera" <djesys.nospam@fsi.net> writes:  > D > > Bulk erase them. VHS degaussers show work very nicely, if you're. > > persistent. Watch trhe duty cycle, though. > G > That needs to be a very good professional unit to erase DLT IV tapes.   D I've had good luck with a VHS eraser that Radio Shack sells (or soldG about 8 years ago). Not much of a duty cycle, I've not needed it *THAT* 4 often or for much volume (one or two cart.'s, tops).   -- V David J. Dachtera	 dba DJE Systemsu http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Wed, 03 Oct 2001 20:12:58 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>e? Subject: Re: Fabbing new VAX? (was: Re: VAX disaster tolerance)i, Message-ID: <3BBBA989.3AF07189@videotron.ca>   Paul Repacholi wrote:oC > One of the sad facts of life is that your digital design can fail @ > on a faster process :( So you need to verify that it will work/ > and be able to tweek it if a wheel falls off.   J It is my understanding that a "faster process" is just a much more preciseE process that allows extreme precision and smaller/thinner "wires" andeL transistors. But woudln't these machines still be able to stamp a VAX designM on a chip that would run at the same speed as it had originally been designed L for ? Wouldn't that design simply yield very few defects because it is beingJ built by a fab that is more precise than what was available in the 1980s ?   ------------------------------  % Date: Wed, 03 Oct 2001 15:51:10 -0400i- From: JF Mezei <jfmezei.spamnot@videotron.ca>s? Subject: Re: Fabbing new VAX? (was: Re: VAX disaster tolerance)l, Message-ID: <3BBB6C2B.3AC8CD2B@videotron.ca>   Hoff Hoffman wrote:sH >   around) but the semiconductor manufacturing processes that were usedK >   to fabricate the VAX chips such as NVAX+ are themselves no longer used.f  N But if HP were to have a garage sale and I were to buy all of the VAX patents,J designs etc for $1, couldn't I go to Intel, Motorolla etc and tell them to& make me a few thousand customs chips ?  K Surely the microprocessors used to drive microwaves or watches aren't builteL with 13 micron copper technology ? There must be some FABs around that could still build VAX chips ?r  C Also, if IBM has a FAB capable of 13 micron, couldn't that same FABkN technically build a VAX with a design that only required 30 micron precision ?  L Wouldn't the VAX become incredibly cheap because it wouldn't be built on theM latest and greatest FABs and not cost much more than the small CPU on my bikes
 computer ?   ------------------------------  # Date: Wed, 03 Oct 2001 18:54:38 GMT_2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)? Subject: Re: Fabbing new VAX? (was: Re: VAX disaster tolerance)i2 Message-ID: <O9Ju7.1028$YP.25869@news.cpqcorp.net>  e In article <3BBB4A6D.30EF7E1C@cableinet.co.uk>, Tim Llewellyn <tim.llewellyn@cableinet.co.uk> writes:n :p :Fabio Cardoso wrote:o   :> A VAX...e :> eH :> Why Compaq dont give the VAX license to be built by another vendor ? G :> If the VAXes are really needed in the market. Example: Mentec shouldr  :> build a 1U VAX for rackmount.  F   I am aware of no vendor that is seeking to fab a VAX microprocessor,F   which is at (no pun intended) the core of this discussion.  Yes, we F   have the VAX designs around (I'd expect to be able to find both the H   logical layouts and the masks, and I know the verification suites are G   around) but the semiconductor manufacturing processes that were used mI   to fabricate the VAX chips such as NVAX+ are themselves no longer used.y  E   We stopped fabbing the VAX chips a number of years ago, and we then0E   used the chip stockpile that had been built up before shutting downaE   the VAX chip production to continue to build VAX processor modules.KF   We then attempted to predict the VAX processor module sales volumes,D   and built the NVAX+ chips we had into the VAX modules accordingly.  H   If a vendor wanted to create a (insert prefered new form-factor) VAX, H   the vendor would have to use an emulation, they would have to restart I   the semiconductor fabrication processes, or they would have to harvest  &   VAX chips from existing VAX systems.  I   This can get a even more interesting, too.  Various system designs can  K   expect specific to use "glue chips", and these supporting semiconductors eH   may or may not be available.  This means that a vendor might well haveJ   to fab a new microprocessor, fab new "glue chips", and/or might have to H   alter the microprocessor design and/or alter the VAX system design to J   replace discontinued chips with the available supporting semiconductors.E   (And these newer chips may or may not be compatible with the older h   semiconductor designs.)e  K   If you don't choose to design a new VAX microprocessor layout, you would 3K   also only expect to see the performance improvements that are due to the  I   semiconductor process (and the layout efforts involved in getting into e   this new process).  H   I will assume that folks now begin to get a picture of the complexity.  :   So a vendor would build a new VAX exactly how?  And why?  G   (A xylinx-based implementation of a VAX would be a real trip.  A slowK    trip, too.  But I digress. :-)  I :Isn't the main reason for people retaining VAXen due to legacy hardware d
 :interfaces? a  G   There are probably as many different reasons given for maintaining anpJ   old system as there are reasons given for maintaining and driving older    vehicles.i  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  $ Date: Wed, 3 Oct 2001 17:37:38 -0700# From: "Tom Linden" <tom@kednos.com>l? Subject: RE: Fabbing new VAX? (was: Re: VAX disaster tolerance)a9 Message-ID: <CIEJLCMNHNNDLLOOGNJIMEDODGAA.tom@kednos.com>g   Isn't it 0.13 micron?    > -----Original Message-----6 > From: JF Mezei [mailto:jfmezei.spamnot@videotron.ca], > Sent: Wednesday, October 03, 2001 12:51 PM > To: Info-VAX@Mvb.Saic.ComdA > Subject: Re: Fabbing new VAX? (was: Re: VAX disaster tolerance)R >e >  > Hoff Hoffman wrote:eJ > >   around) but the semiconductor manufacturing processes that were used@ > >   to fabricate the VAX chips such as NVAX+ are themselves no > longer used. >,C > But if HP were to have a garage sale and I were to buy all of thes > VAX patents,L > designs etc for $1, couldn't I go to Intel, Motorolla etc and tell them to( > make me a few thousand customs chips ? >u@ > Surely the microprocessors used to drive microwaves or watches > aren't builtC > with 13 micron copper technology ? There must be some FABs aroundr > that could > still build VAX chips ?m >oE > Also, if IBM has a FAB capable of 13 micron, couldn't that same FABt= > technically build a VAX with a design that only required 30: > micron precision ? > A > Wouldn't the VAX become incredibly cheap because it wouldn't be4 > built on the@ > latest and greatest FABs and not cost much more than the small > CPU on my bike > computer ? >@   ------------------------------    Date: 04 Oct 2001 05:53:21 +0800, From: Paul Repacholi <prep@prep.synonet.com>? Subject: Re: Fabbing new VAX? (was: Re: VAX disaster tolerance)t- Message-ID: <87n13875m6.fsf@prep.synonet.com>s  / JF Mezei <jfmezei.spamnot@videotron.ca> writes:   @ > Surely the microprocessors used to drive microwaves or watchesD > aren't built with 13 micron copper technology ? There must be some0 > FABs around that could still build VAX chips ?  aE > Also, if IBM has a FAB capable of 13 micron, couldn't that same FABfD > technically build a VAX with a design that only required 30 micron
 > precision ?o  A One of the sad facts of life is that your digital design can failb> on a faster process :( So you need to verify that it will work- and be able to tweek it if a wheel falls off.1   -- o< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.b@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Wed, 03 Oct 2001 20:48:28 GMTi2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)U Subject: Re: FAQ updates (was: RE: OpenVMS Frequently Asked Questions (FAQ), Part 4/5e2 Message-ID: <wQKu7.1037$YP.25612@news.cpqcorp.net>  B In article <00A02FCB.F29E2053.1@decus.de>, zessin@decus.de writes:  H   Please compose a brief explanation of what should be added or updated,H   and submit it to the FAQ maintainer per the recommendations in the FAQL   introductory section, and the text of the request will likely be included H   into the next edition of the FAQ.  Random postings made to newsgroups G   concerning the FAQ may or may not ever be seen by the FAQ maintainer..  ; :I have made a special EXPAT package for the current Pythonk# :for OpenVMS release (V2.1.1-V001).  :e2 :See: http://www.decus.de/~zessin/python2/ext.html  D   There is a Python section in the FAQ.  If there should be a PythonC   EXPAT section (or a reference to the availability of EXPAT in the C   existing Python section), that obviously can be added to the FAQ.   B   Assuming the FAQ maintainer sees (and understands) the request,    of course.  N  ---------------------------- #include <rtfaq.h> -----------------------------N       For additional, please see the OpenVMS FAQ -- www.openvms.compaq.com    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  $ Date: Wed, 3 Oct 2001 17:11:55 -08003 From: loancompetitor.com <sentfrom@marketsplus.org>  Subject: Free Home Appraisal9 Message-ID: <iss.7142.3bbba9d6.184a7.2@mx2.west.saic.com>H  K If you are considering a home purchase, refinance, home improvement or debthJ consolidation, we are offering a free home appraisal when you close a loan) with us now through the month of October.e  K Visit our website at www.loancompetitor.com to find out how you can benefitoK from one of our many available lending programs while interest rates remaina at a historic low.  I Please note that you are only being sent this message once and you do not F need to reply in order to be removed from our mailing list. Thank you.   ------------------------------  $ Date: Wed, 3 Oct 2001 20:21:21 -0400) From: "Neil Rieck" <n.rieck@sympatico.ca>a  Subject: Re: Good bye VAX-6400's; Message-ID: <bTNu7.12990$Ve4.2080188@news20.bellglobal.com>2  3 "ChrisQ" <lightwork@aerosys.co.uk> wrote in message.' news:3BB6427A.DE9FBD9E@aerosys.co.uk...p [snip] >gI > It does have a second TU58 on the side, but both the microcode and boot. filesoH > are on a single cassette and thus only a single cassette is needed for boot.2E > Bootstrap (BSD) is not that bad timewise and once it's running, then
 machine isI > quite usable for interactive edits etc. The 725 is in fact a repackagedu 730,( > with fewer slots than the big box 730. >lE > Wasn't much of the early X development done on a 725 ?. I think the  machine wasdH > called a VS100 (Vaxstation 100) and used a unibus graphics card, maybe from an J > external vendor like Evans & Sutherland. Anyone know anything about this ?. SomeoC > of the very early X distibutions make reference to this hardware.  >t > Chris   J I recall making a "fast boot" TU58 console tape to speed up my VAX-11/730.I You use a console listing of the previous boot to create a new tape which-9 contains "only" the necessary files in the correct order.G  
 Neil Rieck Kitchener/Waterloo/Cambridge,: Ontario, Canada.! http://www3.sympatico.ca/n.rieck/r   ------------------------------  $ Date: Wed, 3 Oct 2001 14:10:09 -0400' From: "Bill Todd" <billtodd@foo.mv.com> L Subject: Re: I WILL QUIT USING VMS WHEN THE MILITARY DOES - IN 2015 OR MAYBE( Message-ID: <9pfk62$bbo$1@pyrite.mv.net>  J "Wayne Sewell" <wayne@tachysoft.xxx.374515.killspam.015f> wrote in message( news:vWSYrrkNfftf@tachxxsoftxxconsult...I > In article <hXPt7.57837$vq.10689520@typhoon.ne.mediaone.net>, "Terry C.?, Shannon" <terryshannon@mediaone.net> writes:   ...l  L > > It gets worse. Seems that the next US aircraft carrier (due 2007/2008 orH > > so), CVN such-and-such, will sport a Windoze-based battle management system. K > > I am sure that rivals and potential enemies of the USA are delighted bya this > > Stupid Strategy Trick. >gK > Especially when the thing gets confused and starts sinking the members ofo itsl > own carrier battle group.s  K But since a reboot will likely be required between volleys, there should beV time to intercede.   - bill   ------------------------------  % Date: Wed, 03 Oct 2001 20:23:15 +0100s% From: Alan Greig <a.greig@virgin.net>sL Subject: Re: Itanium and Bi-endianism (was: Re: Ben Rosen the real culprit?)* Message-ID: <3BBB65A3.441056A8@virgin.net>   Nic Clews wrote:  H > The more and more I understand about the 'port', the more I realise itJ > was the right thing to do. I also appreciate this is the time to get our  J The port is undeniably the right thing to do. It's about two years since IH posted that VMS must be ported to IA64 or die when Compaq killed Alpha -J something I always expected after the termination of Alpha/NT. Just didn'tJ expect the announcement quite this soon. It's the ending of Alpha that was wrong - not the VMS IA64 port.   --  * > Regards, Nic Clews CSC Computer Sciences > nclews at csc dot come   --
 Alan Greig   ------------------------------  $ Date: Wed, 3 Oct 2001 14:07:55 -0400' From: "Bill Todd" <billtodd@foo.mv.com>o) Subject: Re: Letter from Michael CappelasC( Message-ID: <9pfk1s$bbf$1@pyrite.mv.net>  ? "Terry C. Shannon" <terryshannon@mediaone.net> wrote in messaged7 news:qpFu7.16540$xG6.6973985@typhoon.ne.mediaone.net...    ...   F > One of Eckhard's primary goals was to make CPQ a $50B USD company by Y2K01.A > Based on 3FQ01 projections, the current run rate is about $30B.i  K Ah, but last year's *actual total* was just over $40 billion, IIRC.  So oldoL Eckhard wasn't so far off the mark, and if the non-Intel side of the company had been capitalized upon...   - bill   ------------------------------  % Date: Wed, 03 Oct 2001 15:23:06 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca>c) Subject: Re: Letter from Michael Cappelase, Message-ID: <3BBB6599.A18EB7E8@videotron.ca>   Bob Koehler wrote:E >    HP thinks they already do the enterprise bussiness and that theyeI >    can handle the VMS market place better than Compaq (the PC company).n > G >    So why would they make a statement that they're trying to buy intod >    the enterprise bussines?3  M HP used to be very enterprise and high quality. But is today's HP like that ?aK Wall street seems to view HP as a wintel outfit. And when wall street views M your company as a wintel outfit, you have to concentrate on making the winteln7 portion work to please the wall street casino analysts.d  H On the day Compaq announced its suicide, Carly said that once Compaq is I eliminated^H^H^H^H^H^Hmerged,  when enterprise software companies such as M Oracle want to sell into the industry standard market, they will have to talkx/ to HP because HP will be the only shop in town.-  M So I was given a definite impression that HP/Carly have a strong focus on theL wintel market.   ------------------------------   Date: 3 Oct 2001 18:38:17 -0700g, From: _firstname_@lr_dot_los-gatos_dot_ca.us) Subject: Re: Letter from Michael Cappelass% Message-ID: <9pgei9$15so$1@idiom.com>w  F A few personal comments on the great set of marketroid-platitudes from& Mr. Capellas, delivered here recently.  ? I hope that some Cpq staff who monitor this newsgroup send some-F feedback to management: Cpq has a very bad reputation among the people= who use, administer, plan/specify/recommend, and purchase VMS B machines.  Many of Cpq's own customers think (whether justified or@ not) that Cpq is a bunch of habitual liers. Being distrusted andF disliked by your own customers is a big problem.  Cpq/HP needs to makeE some tangible moves and actual long-term commitment to the enterprise 0 space, rather than jerking its customers around.  C But Cpq is following a time-honored DEC tradition in de-committing;iD let me give a few examples of the lies I've been fed in my time as a
 DEC customer: E - The DEC-10 is for the mainframe marketplace, and will co-exist withV>   the VAX.  Jupiter will soon bring a big performance upgrade.1 - Both VMS and Ultrix will be supported on VAXes.wB - Ultrix will be supported on MIPS for a long time; DEC has made aE   major investment in MIPS to guarantee that. The ACE initiative willt'   increase availability and lower cost.wB - VMS can never be run on RISC machines.  DEC proved that with theB   with the first prototype RISC system (I forget the name, the oneC   before the Alpha).  RISC machines just don't have the support foro<   atomic updates and ASTs.  VAX can be pushed to much higherC   performance, and is here to stay.  Just look at the VS4000/90 ands   the 6000-6xx. ? - Alpha and VAX will coexist for a long time; you can invest inh	   either.sD - The VAX 9000 has 40 VUPs per CPU (we bought one and benchmarked it2   at 12 or 13; that caused a lawsuit against DEC).4 - (At this point, I stopped buying any Digital gear)@ - Alpha has a long-term future (we heard that one just up to the   demise a few months ago).   D For any one or two of these mis-steps, one can come up with excuses.> The real problem is that DEC (and later Cpq) continuously tell/ customers one thing, and then do another thing.F   So what's new now?   --  @ The address in the header is invalid for obvious reasons. Please@ reconstruct the address from the information below (look for _).C Ralph Becker-Szendy      _firstname_@lr _dot_ los-gatos _dot_ ca.us    ------------------------------  # Date: Thu, 04 Oct 2001 01:58:05 GMTg4 From: "Terry C. Shannon" <terryshannon@mediaone.net>) Subject: Re: Letter from Michael Cappelaso= Message-ID: <NmPu7.76287$vq.13913437@typhoon.ne.mediaone.net>s  9 <_firstname_@lr_dot_los-gatos_dot_ca.us> wrote in messagew news:9pgei9$15so$1@idiom.com...rH > A few personal comments on the great set of marketroid-platitudes from( > Mr. Capellas, delivered here recently. >pA > I hope that some Cpq staff who monitor this newsgroup send some H > feedback to management: Cpq has a very bad reputation among the people? > who use, administer, plan/specify/recommend, and purchase VMSeD > machines.  Many of Cpq's own customers think (whether justified orB > not) that Cpq is a bunch of habitual liers. Being distrusted andH > disliked by your own customers is a big problem.  Cpq/HP needs to makeG > some tangible moves and actual long-term commitment to the enterpriseg2 > space, rather than jerking its customers around. >cE > But Cpq is following a time-honored DEC tradition in de-committing;bF > let me give a few examples of the lies I've been fed in my time as a > DEC customer:oG > - The DEC-10 is for the mainframe marketplace, and will co-exist with @ >   the VAX.  Jupiter will soon bring a big performance upgrade.3 > - Both VMS and Ultrix will be supported on VAXes. D > - Ultrix will be supported on MIPS for a long time; DEC has made aG >   major investment in MIPS to guarantee that. The ACE initiative willo) >   increase availability and lower cost. D > - VMS can never be run on RISC machines.  DEC proved that with theD >   with the first prototype RISC system (I forget the name, the oneE >   before the Alpha).  RISC machines just don't have the support forl> >   atomic updates and ASTs.  VAX can be pushed to much higherE >   performance, and is here to stay.  Just look at the VS4000/90 andp >   the 6000-6xx.wA > - Alpha and VAX will coexist for a long time; you can invest ind >   either.tF > - The VAX 9000 has 40 VUPs per CPU (we bought one and benchmarked it4 >   at 12 or 13; that caused a lawsuit against DEC).6 > - (At this point, I stopped buying any Digital gear)B > - Alpha has a long-term future (we heard that one just up to the >   demise a few months ago).r >   I If nothing else, the foregoing illustrates the massive disconnect between  engineering and marketing.  L One example: VMS never running on RISC. I don't know in what year that claimG was made, but the first DEC RISC system, TITAN (circa early 80's) was aoK UNIX-only machine. Successor RISC designs (SAFE and PRISM) were designed tolL run VMS or a VMS-like OS. Ken Olsen spoke of a "RISCY VAX" as early as 1988.  I During a 39-month period (beginning with CVAX 1 circa 1987, IIRC) the VAX L architecture experienced a thirteenfold performance improvement, which ain't bad at all.n  K But 40 VUPs on an Aridus machine is a huge stretch. Even if you cranked thenJ clock from 15ns to 9ns (which could be done on the system console), no way- would you get 40 anything out of that turkey!a   ------------------------------  % Date: Wed, 03 Oct 2001 22:30:58 -0500y1 From: "David J. Dachtera" <djesys.nospam@fsi.net>p) Subject: Re: Letter from Michael Cappelasb' Message-ID: <3BBBD7F2.14C8B234@fsi.net>    "Terry C. Shannon" wrote:  > [snip]I > HP actually has more credibility in the enterprise space than does CPQ.a  E These days, I would think that either Bill Clinton or Dick Nixon have @ more credibility than CPQ. (Dick Nixon, before Nixon dicks you!)   -- a David J. Dachterab dba DJE Systemse http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/r   ------------------------------  % Date: Wed, 03 Oct 2001 22:31:59 -0500.1 From: "David J. Dachtera" <djesys.nospam@fsi.net>n) Subject: Re: Letter from Michael Cappelasm' Message-ID: <3BBBD82F.76A6F195@fsi.net>-   "Terry C. Shannon" wrote:- > > > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message# > news:3BBA88A5.E2FF2009@fsi.net...  > > Paul Sture wrote:- > > > I > > > And watching the soccer this weekend, they are _still_ blasting thel
 > airwavesN > > > with PC adverts - with a PIII logo in the bottom left hand corner, would1 > > > you believe, so even _that_ is out of date.  > >nJ > > Remember what happened to the last Compaq product that got any kind ofB > > advertising in Europe or the UK? (Hint: Think OpenVMS, Alpha). > >V > G > Yes, the peecee ads featuring the Presario and the ever-popular IntelrG > Xylophone Orchestra have reared their ugly heads on US teevee as wello6 > (they're running several times a night on Fox News). > A > Could this mean that the Presario will share the fate of Alpha?    We can only hope...e   -- g David J. Dachterao dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/m   ------------------------------  % Date: Thu, 04 Oct 2001 01:26:06 -0400d- From: JF Mezei <jfmezei.spamnot@videotron.ca>i) Subject: Re: Letter from Michael Cappelas , Message-ID: <3BBBF2EC.C128D169@videotron.ca>  - _firstname_@lr_dot_los-gatos_dot_ca.us wrote:pH > disliked by your own customers is a big problem.  Cpq/HP needs to makeG > some tangible moves and actual long-term commitment to the enterprisee2 > space, rather than jerking its customers around.  L My impression is that Compaq has taken satisfactory steps towards the subsetE of VMS customers it intends to keep and doesn't give a shit about the-H remainder of customers who don't spend that much on Compaq gear anyways.  K By downsizing the VMS customer base, keeping only the most profitable ones,sN Compaq can then claim on having higher profit per VMS customer, and perhaps byM keeping only a few customers, it can also downsize its support infrastructurehK and provide more personalised service to the few remaining customers. ThoseAJ customers would get better service and Compaq wouldn't have to worry about tons of small customers.  L In other words, the way I see it, Compaq is turning VMS into an NSK businessD model with a very small but highly profitable customer base that get specialised service.   ------------------------------  % Date: Wed, 03 Oct 2001 15:44:22 -0400a- From: JF Mezei <jfmezei.spamnot@videotron.ca> 5 Subject: Re: Locate Filename containing logical block , Message-ID: <3BBB6A93.A4B7ECFA@videotron.ca>   Paul Repacholi wrote: A > Possible,  but as Keith said, it would be in INDEXF.SYS anyway.eA > As it *IS* a shadow set, dismounting the DK device, re-mount it = > private and see where the error is, then shove it back into- > the set with a full copy.c  M In the case of a shadowset, , I realise that the error log would give details:M about the read error on the actual physical disk, but would the user actually N be told about such an errior if the boock can be read succesfully from another drive in the set ?  I The original poster mentioned something about backup. Does Backup somehoweM bypass some of the shadowing layer to access more information from the actualn
 hard drives ?    ------------------------------    Date: 04 Oct 2001 01:12:43 +0800, From: Paul Repacholi <prep@prep.synonet.com>5 Subject: Re: Locate Filename containing logical blockm- Message-ID: <874rpg8x6c.fsf@prep.synonet.com>r  " norm.raphael@jamesbury.com writes:  < > The current speculation is that this block is an alternate? > homeblock.  It is not in a file, but is read by BACKUP/IMAGE,hE > producing the error.  It is marked Force-error, but unless the diskAE > is restored it will not go away.  Any support or refutation of thisi > speculation is welcome.v  ? Possible,  but as Keith said, it would be in INDEXF.SYS anyway.e? As it *IS* a shadow set, dismounting the DK device, re-mount it-; private and see where the error is, then shove it back into- the set with a full copy.l   -- e< Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.-@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------   Date: 3 Oct 2001 20:52:27 -0700 1 From: keithparris_NOSPAM@yahoo.com (Keith Parris)f5 Subject: Re: Locate Filename containing logical blockm= Message-ID: <cf15391e.0110031952.63f9d5c9@posting.google.com>a  E Paul sent me an e-mail showing the extents of INDEXF.SYS, and the bade? block was nowhere in there, so it can't be a backup home block.o  - JF Mezei <jfmezei.spamnot@videotron.ca> wrotenO > In the case of a shadowset, , I realise that the error log would give detailstO > about the read error on the actual physical disk, but would the user actuallyrP > be told about such an errior if the boock can be read succesfully from another > drive in the set ?  A If Shadowing finds an uncorrectable read error on one member of acC shadowset, and it can read the data successfully from another valid2C member, it revectors the bad sector and overwrites it with the goodsD data from the other disk at that time, which clears the forced-error< flag, so there would be no forced-error flag left around for subsequent reads.n  D The fact that a shadowset contains a sector with a Forced-Error FlagD would normally indicate that at least at one point in the past thereD was only one member in the shadowset, and there was an uncorrectable> read error on that sector at that time (because the odds of anF uncorrectable read error on two or more separate physical disks at the same time are quite low).a  E Then how could a multi-member shadowset contain a forced-error flag? mA If Shadowing is doing a full copy operation from a single disk tonB another disk and comes across a forced-error-flagged sector on the: source disk, it dutifully replicates that sector, with theB forced-error flag set, on the output volume.  That way a shadowsetD always provides the semantics of a single disk: always returning theE same data (and error, in this case) regardless of which member of thei shadowset is read.  K > The original poster mentioned something about backup. Does Backup somehow-O > bypass some of the shadowing layer to access more information from the actualt > hard drives ?    I doubt that this is the case.  A Could it be the file is marked-for-delete, and thus DFU SEARCH is@C ignoring it?  Could the block be incorrectly marked allocated?  For.C either of these cases, an ANALYZE/DISK/REPAIR pass followed by somenB operation to overwrite all free space (e.g. by writing a file thatB takes up all free blocks on the disk) might clear the forced-error flag on that block. C --------------------------------------------------------------------C Keith Parris | parris at encompasserve dot org | VMS consulting on:mC Clusters, Disaster Tolerance, Internals, Performance, Storage & I/O    ------------------------------  % Date: Wed, 03 Oct 2001 16:39:16 -0700u' From: David Mathog <mathog@caltech.edu>tJ Subject: Re: Microsoft Addresses Security Fears: Chance To Market VMS/CSWS+ Message-ID: <3BBBA1A4.75CAFA1F@caltech.edu>    Jerry Leslie wrote:    >-G > Someone in HP/Compaq should install an OpenVMS/CSWS at Mr. Osbourne's E > site, and let him see that there are web servers that don't require ! > "frequent patches and updates".t > I > If HP/Compaq don't make such an offer, perhaps Andrew Harrison will :-)eI > will offer a Solaris solution.                                      :-)i  J Well, by strange coincidence, there was a bug in Sun's own web server that	 caused ittI to crash when probed by Code Red (or was it Nimda?)  It wasn't a security  hole per se, moreiJ of a DOS sort of thing.  Sun issued a patch for that.   Sun issues patches like a meter maidsJ issues parking tickets.  Even so, Sun could certainly march in with a well configured ApacheoG server on Solaris and make a good case for that.  The only problem that  code red et al. causeeI for that combination is that it tends to make the  log files grow really,0 really big.42 Which is really not very much of a problem at all.  @ However, it is inconceivable that Compaq will capitalize on this opportunity because:  K 1.  Q management wouldn't recognize a business opportunity if it was shoved6 up theirA collective rear ends with a ten foot pole (to mangle a metaphor).sJ 2.  Acting on this opportunity would anger Microsoft..  Since Q management has its lipsK firmly affixed to that other company's posterior they would never make such$ a move evenI9 if by some miracle they thought of it in the first place.e  F Both of which go a long way towards explaining how the Q got to be the shambles it is today.-   Regards,   David Mathog mathog@caltech.edu   ------------------------------   Date: 3 Oct 2001 21:40:25 GMTe) From: leslie@clio.rice.edu (Jerry Leslie)6F Subject: Microsoft Addresses Security Fears: Chance To Market VMS/CSWS' Message-ID: <9pg0k9$6ld$1@joe.rice.edu>t2 Keywords: microsoft,security,fears,market,vms,csws  4    http://news.cnet.com/news/0-1003-200-7396207.htmlF    Microsoft attempts to allay security fears -  Tech News -  CNET.com  -   "Microsoft attempts to allay security fears     By Joe Wilcox    Staff Writer, CNET News.com!    October 3, 2001, 10:50 a.m. PTl  G    Microsoft on Wednesday launched a major security initiative intendedhF    to address concerns that its software is prone to security problems    and virus attacks.s  D    The world's largest software maker is looking to allay fears overH    ongoing security problems and recent worm attacks from Nimda and CodeB    Red, which have led experts to warn that the company's InternetB    Information Server Web server software is not secure enough for
    customers.g  G    On Tuesday, Microsoft contacted about 1,000 of its largest customerslE    to outline its program, called the Strategic Technology Protection F    Program, and address their concerns about recent worm attacks, said-    analysts and sources close to the company.5    .    .    .F    Peter Osbourne, manager of advanced technology and decision supportF    services at Dollar Rent-A-Car's headquarters, in Tulsa, Okla., saidE    installing frequent patches and updates are a necessity when using H    Microsoft's products. "Because Windows and IIS are so prevalent, they    are easy targets," he said."n  B Mr. Osbourne needs to understand that they're easy targets because% they have software defects, aka bugs.n  E Someone in HP/Compaq should install an OpenVMS/CSWS at Mr. Osbourne'sPC site, and let him see that there are web servers that don't requireo "frequent patches and updates".h  G If HP/Compaq don't make such an offer, perhaps Andrew Harrison will :-) G will offer a Solaris solution.                                      :-)   4 --Jerry Leslie     (my opinions are strictly my own)   ------------------------------  # Date: Thu, 04 Oct 2001 05:40:11 GMTi From: LBohan@dbc.spam_less..comaJ Subject: Re: Microsoft Addresses Security Fears: Chance To Market VMS/CSWS8 Message-ID: <b7tnrtk2qku776np2qkrj0uiqmvc96ejg3@4ax.com>  F On 3 Oct 2001 21:40:25 GMT, leslie@clio.rice.edu (Jerry Leslie) wrote:  G >   Peter Osbourne, manager of advanced technology and decision supportiG >   services at Dollar Rent-A-Car's headquarters, in Tulsa, Okla., said F >   installing frequent patches and updates are a necessity when usingI >   Microsoft's products. "Because Windows and IIS are so prevalent, theyc  >   are easy targets," he said." >tC >Mr. Osbourne needs to understand that they're easy targets becausen& >they have software defects, aka bugs. > F >Someone in HP/Compaq should install an OpenVMS/CSWS at Mr. Osbourne'sD >site, and let him see that there are web servers that don't require  >"frequent patches and updates".  8 While Netcraft shows www.dollar.com as running IIS/W2;K,> it was my understanding that Dollar does run some VMS systems.   eg.n  E  http://www.dollar.com/company_information/display_emp_opp.asp?id=661l  ! MANAGER, TECHNICAL SUPPORT I.T.S.u Posting #01-053m  E Responsible for evaluating, developing, recommending and implementing F system security procedures; researching and resolving system problems;> researching and recommending hardware acquisitions; installing@ software; analyzing the system usage; providing on-site support;D researching various vendor system capabilities; evaluating strategic? plans and developing strategy for hardware database. Supervises:  Administrators and Coordinators.  B       Bachelors degree in Computer Science or related discipline.@       5 years of experience with UNIX, and/or Open VMS operating systems.>       Thorough knowledge of computer networking and protocols.6       Excellent oral and written communication skills.3       Strong problem solving and analytical skills.t#       Ability to program computers.a1       Ability to build critical business systems.nC       Proficiency in the use of computer systems (UNIX & Open VMS).h>       Ability to work with all levels of personnel and outside vendors.       Ability to type 35 w.p.m.oC Ability to operate proficiently Digital equipment (Alpha), personale+ computer, fax, copy machine and calculator.   @ If you are interested in pursuing this unique opportunity with aA strong international company that offers a competitive package ofpC compensation and benefits please send resume with salary history. . E When applying for this position please include the Posting number for 	 this job.          Human Resources        5330 E. 31st Streetw       Tulsa, OK 74135 %       or fax resume to (918) 669-3414 )       or E-mail resume to Jobs@dollar.comi   EOE/M/F/D/Vr   ------------------------------  % Date: Wed, 03 Oct 2001 22:45:40 -0500a. From: Lyndon Bartels <lbartels@pressenter.com>/ Subject: Need help with $QIO calls in C programa. Message-ID: <3BBB9514.3C1A2B4B@pressenter.com>  , This is a multi-part message in MIME format.  $ --------------175779287418F641EED7E6* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bitd   Hello,  / Alpha VMS v7.2-1 and v7.2-2 (with various ECOs) ) TCP/IP v5.0a and v5.1 (with various ECOs)c
 DEC C v6.2  H I'm looking for some help in writing a tcp/ip client server application.  > And, before anything else, no, this is NOT a homework project.  F I started by writing a couple simple programs, a client and server, to prove > the concept and to test my logic. (I've tried to attach them.)  F My idea is have the server listen, whenever a connection request comes in, anF AST fires off, takes the connection and gets the data from the client, andeG sends back an acknowledgement. Otherwise, the server is in HIBER state.nH Consuming minimal resources. The client is the active part; it fires up, getsF the server's address, and gets the user input. It sends the request to the % server and waits for acknowledgement.s  F I've written both applications, and they both work. And they do what I want themG to do. But I'm concerned. I'm not sure I've written the applications toe be$ as efficient and robust as possible.  ( I chose to use the AST approach so I canF handle multiple clients. I thought that by keeping the listener highlyC available, and by firing off an AST to handle the actual connectionrB communication, I could have multiple clients. Keeping the listener available as@ much as possible is important. I want to be certain that the AST	 instancese? won't interfere with one another. (Assuming multiple copies are. simultaneously active.)  F I don't want to keep i/o connections open and "locked" any longer thenF necessary. Nor do I want to waste resources unnecessarily creating and deleting i/o connections either.   G I looked through TCPIP$EXAMPLES:*.C files, and I've been reading at thexE manuals. Particularly the TCP/IP Services for OpenVMS Sockets API andu System Services Programming manual.  < http://www.openvms.compaq.com:8000/73FINAL/6529/6529pro.html  D In Chapter 2: I used figure 2-1 for a model but added looping on the client" <side, and AST on the server side.  @ I guess my base questions are: How did I do? Did I call the QIOs correctly? AndF in the correct sequences? Am I doing too much? Not enough? Am I on the rightn track?  E Anyway, using the examples as starting a starting, I came up with the  attached programs.     Thank-you for looking at them.   Lyndon   --  G My opinions are mine and mine alone. They seldom align with those of my 	 employer.     H The only good thing about putting the cart before the horse is you don't have to look at the horse's butt.  $ --------------175779287418F641EED7E6I Content-Type: text/plain; charset=us-ascii; name="tcpip-client-sock.c.27"n Content-Transfer-Encoding: 7bitp> Content-Disposition: inline; filename="tcpip-client-sock.c.27"   /* tcpip-client-sock.c  9 This program is the client portion. It does the followingo   1. Creates socket. 2. Binds socket. 3. Starts looping. 4. Get's user input. 5. sends connection request. 6. Writes data to socket.f 7. Waits for acknowledgment. 8. Goes back to loop beginning 9. Deletes socket.
 10. exits.   */    7 #include <descrip.h>		/* define OpenVMS descriptors		*/e  8 #include <efndef.h>		/* define 'EFN$C_ENF' event flag	*/  : #include <in.h>			/* define internet related constants,	*/$ 				/* functions, and structures		*/5 #include <inet.h>		/* define network address info		*/14 #include <iodef.h>		/* define i/o function codes		*/  8 #include <lib$routines.h>	/* define library routines		*/  > #include <netdb.h>		/* define network database library info */  ; #include <rmsdef.h>		/* define rms functions, structures	*/g  < #include <ssdef.h>		/* define system service status codes	*/8 #include <starlet.h>		/* define system service calls		*/7 #include <stdio.h>		/* define standard i/o functions	*/a< #include <stdlib.h>		/* define standard library functions	*/; #include <string.h>		/* define string handling functions	*/h8 #include <stsdef.h>		/* define condition value fields	*/  B #include <tcpip$inetdef.h>	/* define tcpip/ip network constants	*/$ 				/* structures, and functions		*/   /*  *  NAMED CONSTANTSv  */o  0 #define BUFSZ		1024	/* user input buffer size	*/3 #define SERV_PORTNUM	12345	/* server port number	*/m4 #define max_msg_len	256	/* maximum message length */    3 #define $ARRAY_DESCRIPTOR(name,size,array_name)			\n"     char array_name[ size ];					\(     struct dsc$descriptor_s name = 				\3 	{ size, DSC$K_DTYPE_T, DSC$K_CLASS_S, array_name }-     /*  *  Structure definitions0  */    struct iosb0!     {					/* i/o status block 	*/o7     unsigned short status;		/* i/o completion status */ A     unsigned short bytcnt;		/* bytes transferred if read/write */l9     void *details;			/* address of buffer or parameter */d     };   struct itemlist_2 .     {					/* item-list 2 descriptor/element */,     unsigned short length;		/* length			  */2     unsigned short type;		/* parameter type;		  */2     void *address;			/* address of item list		  */     };   struct sockchard.     {					/* socket characteristics buffer  */,     unsigned short prot;		/* protocol			  */)     unsigned char type;			/* type				  */ /     unsigned char af;			/* address format		  */e     };       /*  *  FORWARD REFERENCES  */n  ( int main(void);				/* client main			  */> void get_serv_addr( void * );		/* get server host address	  */     int main(void) {   + struct iosb iosb;			/* i/o status block		*/OC struct sockchar conn_sockchar;		/* connect socket char buffer	   */ E struct itemlist_2 serv_itemlst;	/* server item-list 2 descriptor   */4D struct sockaddr_in serv_addr;		/* server socket address structure */  9 unsigned int status;			/* system service return status	*/e7 unsigned short conn_channel;		/* connection channel		*/P  ? $DESCRIPTOR( inet_device,		/* string descriptor with logical */e= 	     "TCPIP$DEVICE:");		/* name of internet pseudodevice  */r  V $DESCRIPTOR(prompt_string_desc,"DATA TO SEND TO SERVER (<CTRL/Z> to terminate) >>> ");> $ARRAY_DESCRIPTOR(write_buffer_desc,max_msg_len,write_buffer);' int done = FALSE;			/* Loop Flag			  */    char inbuffer[BUFSZ];h unsigned int inbuflen=BUFSZ;   /*1  *  init connection socket characteristics buffer   */   ! conn_sockchar.prot = TCPIP$C_TCP;9$ conn_sockchar.type = TCPIP$C_STREAM;% conn_sockchar.af   = TCPIP$C_AF_INET;    /*&  *  init server's item-list descriptor  */W  1 memset( &serv_itemlst, 0, sizeof(serv_itemlst) );e* serv_itemlst.length = sizeof( serv_addr );" serv_itemlst.address = &serv_addr;   /**  *  init server's socket address structure  */i  + memset( &serv_addr, 0, sizeof(serv_addr) );:' serv_addr.sin_family = TCPIP$C_AF_INET;m- serv_addr.sin_port   = htons( SERV_PORTNUM );t# get_serv_addr(&serv_addr.sin_addr);-   /*  *  assign device socket  */S  E status = sys$assign( &inet_device,	/* device name                  */b8 		     &conn_channel,	/* i/o channel                  */. 		     0,			/* access mode                  */. 		     0);		/* not used                     */    if ( !(status & STS$M_SUCCESS) )     {o?     printf( "Failed to assign i/o channel to TCPIP device\n" );c     exit( status );u     }m   /*F  *  Loop forever, getting user input, and sending it across the TCPIP   *  connection.o  */    while ( !(done) )      {e1     write_buffer_desc.dsc$w_length = max_msg_len;r     status = lib$get_input(v 		&write_buffer_desc,o 		&prompt_string_desc,# 		&write_buffer_desc.dsc$w_length);o       /*  /      * If at end of file (user typed <ctrl/z> )       * send exit signala      */h       if (status == RMS$_EOF)s 	{ c0 	strcpy(write_buffer_desc.dsc$a_pointer,"exit");$ 	write_buffer_desc.dsc$w_length = 4;
 	done = TRUE;  	}   /*  *  create connection socket  */a  1 status = sys$qiow( EFN$C_ENF,		/* event flag			*/ & 		   conn_channel,	/* i/o channel			*/+ 		   IO$_SETMODE,		/* i/o function code		*/a$ 		   &iosb,		/* i/o status block		*/$ 		   0,			/* ast service routine		*/ 		   0,			/* ast parameter		*/2 		   &conn_sockchar,	/* p1 - socket char buffer	*/! 		   0,0,0,0,0);		/* p2 - p6			*/i   if ( status & STS$M_SUCCESS )r     status = iosb.status;r    if ( !(status & STS$M_SUCCESS) )     {r*     printf( "Failed to create socket\n" );     exit( status );y     }o       /*%      *  Initiate connection to serveri      */s  ;     printf( "Initiated connection to host: %s, port: %d\n",o: 	inet_ntoa(serv_addr.sin_addr),ntohs(serv_addr.sin_port));  4     status = sys$qiow( EFN$C_ENF,	/* event flag			*/* 		       conn_channel,	/* i/o channel			*/- 		       IO$_ACCESS,	/* i/o function code		*/ ( 		       &iosb,		/* i/o status block		*/' 		       0,		/* ast service routine		*/f! 		       0,		/* ast parameter		*/  		       0,		/* p1				*/ 		       0,		/* p2				*/5 		       &serv_itemlst,	/* p3 - remote socket name	*/K! 		       0,0,0);		/* p4 - p6			*/r        if (status & STS$M_SUCCESS ) 	status = iosb.status;  $     if ( !(status & STS$M_SUCCESS) ) 	{* 	printf( "Failed to connect to server\n"); 	exit( status ); 	}       /*+      *  Connection established with server;e)      *  now attempt to send user's input.g      */n  4     status = sys$qiow(	EFN$C_ENF,		/* event flag		*/$ 			conn_channel,		/* i/o channel		*/* 			IO$_WRITEVBLK,		/* i/o function code	*/" 			&iosb,			/* i/o status block	*/! 			0,			/* ast service routine	*/: 			0,			/* ast parameter	*/r, 			write_buffer_desc.dsc$a_pointer,	/* p1	*/, 			write_buffer_desc.dsc$w_length,		/* p2	*/ 			0,0,0,0);		/* p3 - p6		*/  !     if ( status & STS$M_SUCCESS )V 	status = iosb.status;  $     if ( !(status & STS$M_SUCCESS) ) 	{- 	printf( "Failed to write to connection.\n");c 	exit(status); 	}         /*:      *  Message sent. Wait for acknowledgement from server      */   4     status = sys$qiow(	EFN$C_ENF,		/* event flag		*/$ 			conn_channel,		/* i/o channel		*/) 			IO$_READVBLK,		/* i/o function code	*/b" 			&iosb,			/* i/o status block	*/! 			0,			/* ast service routine	*/i 			0,			/* ast parameter	*/ ) 			&inbuffer,		/* input buffer address */ ' 			inbuflen,		/* input buffer length	*/s 			0,0,0,0);		/* p3 - p6		*/  !     if ( status & STS$M_SUCCESS )  	status = iosb.status;  $     if ( !(status & STS$M_SUCCESS) ) 	{9 	printf( "Failed to read data from server connection\n");a 	exit(status); 	}  -     printf( "Data received: %s\n", inbuffer);m   /*  *  shutdown connection socket  */g  1 status = sys$qiow( EFN$C_ENF,			/* event flag		*/.& 		   conn_channel,		/* i/o channel		*/9 		   IO$_DEACCESS|IO$M_SHUTDOWN,  /* i/o function code	*/ $ 		   &iosb,			/* i/o status block	*/$ 		   0,				/* ast service routine	*/ 		   0,				/* ast parameter	*/ 		   0,0,0,			/* p1 - p3		*/5 		   TCPIP$C_DSC_ALL,		/* p4 - discard all packets */M 		   0,0);			/* p5 - p5		*/t   if ( status & STS$M_SUCCESS )      status = iosb.status;i    if ( !(status & STS$M_SUCCESS) )     {I6     printf( "Failed to shutdown server connection\n");     exit(status);c     }l  , printf( "Connection to server shutdown.\n");   /*  *  close connection socketd  */   1 status = sys$qiow( EFN$C_ENF,			/* event flag		*/-& 		   conn_channel,		/* i/o channel		*/+ 		   IO$_DEACCESS,		/* i/o function code	*/l$ 		   &iosb,			/* i/o status block	*/$ 		   0,				/* ast service routine	*/ 		   0,				/* ast parameter	*/" 		   0,0,0,0,0,0);		/* p1 - p6		*/   if ( status & STS$M_SUCCESS )h     status = iosb.status;1    if ( !(status & STS$M_SUCCESS) )     { )     printf( "Failed to close socket.\n");T     exit(status);t     }t   printf( "Socket Closed.\n");   }    /*  *  deassign device socket  */   $ status = sys$dassgn( conn_channel );    if ( !(status & STS$M_SUCCESS) )     {c)     printf( "Failed to close socket.\n");o     exit(status);t     }      return SS$_NORMAL; }      void get_serv_addr( void *addrptr ) {e char buf[BUFSZ]; struct in_addr val;e struct hostent *host;c   while ( TRUE )     { $     printf( "Enter remote host: " );  1     if ( fgets(buf, sizeof(buf), stdin) == NULL )  	{) 	printf( "Failed to read user input\n" );w 	exit( EXIT_FAILURE ); 	}       buf[strlen(buf)-1] = 0;a  "     val.s_addr = inet_addr( buf );  $     if ( val.s_addr != INADDR_NONE ) 	{1 	memcpy( addrptr, &val, sizeof(struct in_addr) );  	break;m 	}  &     if ( (host = gethostbyname(buf)) ) 	{9 	memcpy( addrptr, host->h_addr, sizeof(struct in_addr) );  	break;. 	}     }P }r  $ --------------175779287418F641EED7E6I Content-Type: text/plain; charset=us-ascii; name="tcpip-server-sock.c.41"  Content-Transfer-Encoding: 7bit > Content-Disposition: inline; filename="tcpip-server-sock.c.41"   /*  N This program is a tcpip server process. It starts. Creates a output file. ThenJ goes to sleep. While it is sleeping, any tcpip messages that come into theO computer on the port 12345, awakens an ast. Any non-null text sent to that porttK is printed to the file. Then the process goes back to sleep again. Once thewJ exit signal is recieved, the process awakens, closes the output file, then exits.   */   /*  * INCLUDE FILES:A  */   N #include <descrip.h>                /* define OpenVMS descriptors           */  N #include <efndef.h>                 /* define 'EFN$C_ENF' event flag        */  N #include <in.h>                     /* define internet related constants,   */N                                     /* functions, and structures            */N #include <inet.h>                   /* define network address info          */N #include <iodef.h>                  /* define i/o function codes            */   #include <lib$routines.h>o  N #include <netdb.h>                  /* define network database library info */  N #include <ssdef.h>                  /* define system service status codes   */N #include <starlet.h>                /* define system service calls          */N #include <stdio.h>                  /* define standard i/o functions        */N #include <stdlib.h>                 /* define standard library functions    */N #include <string.h>                 /* define string handling functions     */N #include <stsdef.h>                 /* define condition value fields        */   #include <tcpip$inetdef.h>     /*  * NAMED CONSTANTS:g  */W  4 #define MBUFSIZE	128		/* size of a message buffer	*/ #define MBUFQUO		3842 #define SERV_BACKLOG	1		/* server backlog 		    */N #define SERV_PORTNUM    12345           /* server port number               */     /*  * Structure definitions  */    struct iosbd      {					/* i/o status block */7     unsigned short status;		/* i/o completion status */e@     unsigned short bytcnt;		/* bytes transfered if read/write */9     void *details;			/* address of buffer or parameter */      };     struct itemlst_2N     {                                   /* item-list 2 descriptor/element   */N     unsigned short length;              /* length                           */N     unsigned short type;                /* parameter type                   */N     void *address;                      /* address of item list             */     };   struct itemlst_3N     {                                   /* item-list 3 descriptor/element   */N     unsigned short length;              /* length                           */N     unsigned short type;                /* parameter type                   */N     void *address;                      /* address of item list             */N     unsigned int *retlen;               /* address of returned length       */     };%                                      e struct sockcharcN     {                                   /* socket characteristics buffer    */N     unsigned short prot;                /* protocol                         */N     unsigned char type;                 /* type                             */N     unsigned char af;                   /* address format                   */     };    G struct itemlst_3 client_itemlst;	/* client item-list 3 descriptor    */r: struct iosb iosb;			/* i/o status block                 */G struct sockaddr_in client_addr;		/* client socket address structure  */ E struct sockaddr_in serv_addr;		/* server socket address structure  */nF struct itemlst_2 serv_itemlst;		/* server item-list 2 descriptor    */  D unsigned short conn_channel;		/* connect inet device i/o channel  */F unsigned short listen_channel;		/* listen inet device i/o channel   */  . char inbuffer[MBUFSIZE];		/* message buffer */? unsigned int inbuflen=MBUFSIZE;		/* length of message buffer	*/sA char outbuffer[MBUFSIZE];		/* server data buffer               */oJ unsigned int  outbuflen = MBUFSIZE;	/* length of server data buffer     */   FILE *message_file;,   void connectast();   int main (void) {e    2 fprintf(message_file,"I'm going to sleep now.\n");, int one = 1;				/* reuseaddr option value	*/  9 unsigned int status;			/* system service return status */r  C unsigned int client_retlen;		/* returned length of client socket */k 					/* address structure */  @ struct sockchar listen_sockchar;	/* listen socket char buffer	*/    F struct itemlst_2 sockopt_itemlst;	/* sockopt item-list 2 descriptor */G struct itemlst_2 reuseaddr_itemlst;	/* reuseaddr item-list 2 element */e      ? $DESCRIPTOR( inet_device,		/* string descriptor with logical */l= 	     "TCPIP$DEVICE:" );		/* name of internet pseudodevice */k     /*-  *  init listen socket characteristics buffern  */'  # listen_sockchar.prot = TCPIP$C_TCP;s& listen_sockchar.type = TCPIP$C_STREAM;' listen_sockchar.af   = TCPIP$C_AF_INET;    /*&  *  init reuseaddr's item-list element  */y  * reuseaddr_itemlst.length  = sizeof( one );. reuseaddr_itemlst.type    = TCPIP$C_REUSEADDR;! reuseaddr_itemlst.address = &one;u   /*'  *  init sockopt's item-list descriptor   */u  6 sockopt_itemlst.length  = sizeof( reuseaddr_itemlst );* sockopt_itemlst.type    = TCPIP$C_SOCKOPT;- sockopt_itemlst.address = &reuseaddr_itemlst;n   /*&  *  init client's item-list descriptor  */t  5 memset( &client_itemlst, 0, sizeof(client_itemlst) ); . client_itemlst.length = sizeof( client_addr );& client_itemlst.address = &client_addr;' client_itemlst.retlen = &client_retlen;i   /**  *  init client's socket address structure  */l  / memset( &client_addr, 0, sizeof(client_addr) );    /*&  *  init server's item-list descriptor  */o  + serv_itemlst.length  = sizeof( serv_addr );l) serv_itemlst.type    = TCPIP$C_SOCK_NAME;o" serv_itemlst.address = &serv_addr;   /**  *  init server's socket address structure  */   + memset( &serv_addr, 0, sizeof(serv_addr) );s, serv_addr.sin_family      = TCPIP$C_AF_INET;2 serv_addr.sin_port        = htons( SERV_PORTNUM );/ serv_addr.sin_addr.s_addr = TCPIP$C_INADDR_ANY;a   /*  *  assign device socketst  */d  4 status = sys$assign( &inet_device,	/* device name	*/) 		     &listen_channel,	/* i/o channel	*/n 		     0,			/* access mode	*/  		     0);		/* not used	*/    if ( !(status & STS$M_SUCCESS) )     {rB     printf( "Failed to assign i/o channel(s) to TCPIP device\n" );     exit( status );8     }7    4 status = sys$assign( &inet_device,	/* device name	*/' 		     &conn_channel,	/* i/o channel	*/c 		     0,			/* access mode	*/o 		     0);		/* not used	*/    if ( !(status & STS$M_SUCCESS) )     { B     printf( "Failed to assign i/o channel(s) to TCPIP device\n" );     exit( status );n     }e   /*  * create a listener sockete  */u  2 status = sys$qiow( EFN$C_ENF,		/* event flag 			*/( 		   listen_channel,	/* i/o channel			*/+ 		   IO$_SETMODE,		/* i/o function code		*/.$ 		   &iosb,		/* i/o status block		*/$ 		   0,			/* ast service routine		*/ 		   0,			/* ast parameter		*/4 		   &listen_sockchar,	/* p1 - socket char buffer	*/ 		   0,0,0,		/* p2 - p4			*/1 		   &sockopt_itemlst,	/* p5 - socket options		*/i 		   0);			/* p6 				*/e   if ( status & STS$M_SUCCESS )u     status = iosb.status;     if ( !(status & STS$M_SUCCESS) )     {s3     printf( "Failed to create listener socket\n" );.     exit( status );      }e   /*5  *  bind server's internet address and port number too1  *  listen socket; set socket as a passive socket   */m  1 status = sys$qiow( EFN$C_ENF,		/* event flag			*/f( 		   listen_channel,	/* i/o channel			*/+ 		   IO$_SETMODE,		/* i/o function code		*/ $ 		   &iosb,		/* i/o status block		*/$ 		   0,			/* ast service routine		*/ 		   0,			/* ast parameter		*/ 		   0,0,			/* p1 - p2			*/d0 		   &serv_itemlst,	/* p3 - local socket name	*/0 		   SERV_BACKLOG,	/* p4 - connection backlog	*/ 		   0,0);		/* p5 - p6			*/e   if ( status & STS$M_SUCCESS )A     status = iosb.status;f    if ( !(status & STS$M_SUCCESS) )     {/1     printf( "Failed to bind listener socket\n" );*     exit( status );e     }*      < if ((message_file = fopen("socket-output.txt","w")) == NULL)     {a'     perror ("fopen socket-output.txt");	     exit (EXIT_FAILURE);     }a  < fprintf(message_file, "Program started. Awaiting input.\n");   /*%  *  accept a connection from a client   */t  8 printf( "Waiting for a client connection on port: %d\n",#         ntohs(serv_addr.sin_port));*  F fprintf( message_file,"Waiting for a client connection on port: %d\n",$          ntohs(serv_addr.sin_port));  0 status = sys$qio( EFN$C_ENF,		/* event flag			*/( 		   listen_channel,	/* i/o channel			*/4 		   IO$_ACCESS|IO$M_ACCEPT,	/* i/o function code	*/$ 		   &iosb,		/* i/o status block		*/- 		   &connectast,		/* ast service routine		*/e 		   0,			/* ast parameter		*/ 		   0,0,			/* p1 - p2			*/ 3 		   &client_itemlst,	/* p3 - remote socket name	*/	0 		   &conn_channel,	/* p4 i/o channel for new	*/ 					/* connection			*/d 		   0,0);		/* p5 - p6			*/	   if ( status & STS$M_SUCCESS )W     status = iosb.status;  /*  if ( !(status & STS$M_SUCCESS) )     {t7     printf( "Failed to establish listener channel.\n");      exit( status );      }t */$ printf("I'm going to sleep now.\n");2 fprintf(message_file,"I'm going to sleep now.\n");   /* Go into hibernation */s status = SYS$HIBER();r if (status != SS$_NORMAL)      LIB$SIGNAL(status);s     /* After being awakened */ /* Close test file */i  : printf("Received termination Signal... Shutting down.\n");I fprintf(message_file, "Received termination Signal... Shutting down.\n");i fclose(message_file);o   /*  *  close listen socketP  */C  % printf("Closing listener socket.\n");*  1 status = sys$qiow( EFN$C_ENF,		/* event flag			*/ ( 		   listen_channel,	/* i/o channel			*/+ 		   IO$_DEACCESS,	/* i/o function code		*/,$ 		   &iosb,		/* i/o status block		*/$ 		   0,			/* ast service routine		*/ 		   0,			/* ast parameter		*/" 		   0,0,0,0,0,0);	/* p1 - p6			*/   if ( status & STS$M_SUCCESS )      status = iosb.status;P    if ( !(status & STS$M_SUCCESS) )     {A)     printf( "Failed to close socket\n" );      exit( status );'     }l   /*  *  deassign all device socketst  */,  & printf("Deassigning i/o channels.\n");$ status = sys$dassgn( conn_channel );   if ( (status & STS$M_SUCCESS) ) *     status = sys$dassgn( listen_channel );    if ( !(status & STS$M_SUCCESS) )     {aD     printf( "Failed to deassign i/o channel(s) to TCPIP device\n" );     exit( status );N     }    exit( EXIT_SUCCESS );s }a   void connectast(void) {v  9 unsigned int status;			/* system service return status */i  : char exitstr[5];			/* Exit string to signal end of exec */  - printf("Entering Connection AST routine.\n");o; fprintf(message_file,"Entering Connection AST routing.\n");   7 printf("Accepted connection from host: %s, port: %d\n",r> 	inet_ntoa(client_addr.sin_addr),ntohs(client_addr.sin_port));E fprintf(message_file,"Accepted connection from host: %s, port: %d\n",n> 	inet_ntoa(client_addr.sin_addr),ntohs(client_addr.sin_port));   /*)  *  connection established with a client;_/  *  now attempt to read data on this connectiont  */	  1 status = sys$qiow( EFN$C_ENF,		/* event flag			*/e& 		   conn_channel,	/* i/o channel			*/+ 		   IO$_READVBLK,	/* i/o function code		*/z$ 		   &iosb,		/* i/o status block		*/$ 		   0,			/* ast service routine		*/ 		   0,			/* ast parameter		*/+ 		   &inbuffer,		/* p1 - buffer address		*/w) 		   inbuflen,		/* p2 - buffer length		*/  		   0,0,0,0);		/* p3 - p6			*/    if ( status & STS$M_SUCCESS )      status = iosb.status;_    if ( !(status & STS$M_SUCCESS) )     {O=     printf( "Failed to read data from client connection\n" );k     exit( status );t     }e   inbuffer[0,iosb.bytcnt] = 0;' printf("Data received: %s\n",inbuffer);-5 fprintf(message_file,"Data received: %s\n",inbuffer);*  ! strcpy(outbuffer,"acknowledged");  outbuflen = sizeof(outbuffer);   /*  *  Write data to client  */r  / status = sys$qiow( EFN$C_ENF,		/* event flag */x$ 		   conn_channel,	/* i/o channel */+ 		   IO$_WRITEVBLK,	/* i/o function code */ # 		   &iosb,		/* i/o status block */o# 		   0,			/* ast service routine */s 		   0,			/* ast parameter */d, 		   &outbuffer,		/* p1 - buffer address		*/* 		   outbuflen,		/* p2 - buffer length		*/ 		   0,0,0,0);		/* p3 - p6			*/    if ( status & STS$M_SUCCESS )	     status = iosb.status;/    if ( !(status & STS$M_SUCCESS) )     {v<     printf( "Failed to write data to client connection\n" );     exit( status );      }	    I printf( "Data sent: %s\n", outbuffer ); /* output server's data buffer */0V fprintf(message_file,"Data sent: %s\n", outbuffer ); /* output server's data buffer */   /*  *  Shutdown connection socket  */p  1 status = sys$qiow( EFN$C_ENF,		/* event flag			*/u& 		   conn_channel,	/* i/o channel			*/  		   IO$_DEACCESS|IO$M_SHUTDOWN, 					/* i/o function code		*/t$ 		   &iosb,		/* i/o status block		*/$ 		   0,			/* ast service routine		*/ 		   0,			/* ast parameter		*/ 		   0,0,0,		/* p1 - p3			*/4 		   TCPIP$C_DSC_ALL,	/* p4 - discard all packets	*/ 		   0,0);		/* p5 - p6			*/:   if ( status & STS$M_SUCCESS )	     status = iosb.status;i    if ( !(status & STS$M_SUCCESS) )     {l7     printf( "Failed to shutdown client connection\n" );(     exit( status );S     }s  ( printf("Connection Socket Shutdown.\n");6 fprintf(message_file,"Connection Socket Shutdown.\n");   /*  *  Close Connection Socket   */   0 status = sys$qiow( EFN$C_ENF,		/* event flag		*/% 		   conn_channel,	/* i/o channel		*/s* 		   IO$_DEACCESS,	/* i/o function code	*/# 		   &iosb,		/* i/o status block	*/$# 		   0,			/* ast service routine	*/  		   0,			/* ast parameter	*/	! 		   0,0,0,0,0,0);	/* p1 - p6		*/*   if ( status & STS$M_SUCCESS )      status = iosb.status;f    if ( !(status & STS$M_SUCCESS) )     {e)     printf( "Failed to close socket\n" );      exit( status );T     }C  & printf("Connection Socket Closed.\n");4 fprintf(message_file,"Connection Socket Closed.\n");   strcpy(exitstr,"exit");c  " if (strcmp(inbuffer,exitstr) == 0)     { $     printf("strcmp equals zero.\n");     status = SYS$WAKE(0,0);o1     if (status != SS$_NORMAL) LIB$SIGNAL(status);e     }	     else     {h,     printf("strcmp does not equal zero.\n");         /*)      *  accept a connection from a client	      */a  Y     printf( "Waiting for a client connection on port: %d\n", ntohs(serv_addr.sin_port) );0f     fprintf(message_file,"Waiting for a client connection on port: %d\n", ntohs(serv_addr.sin_port) );       /*      *  Create another listenert      */.  3     status = sys$qio( EFN$C_ENF,		/* event flag		*/ + 		      listen_channel,		/* i/o channel		*/o7 		      IO$_ACCESS|IO$M_ACCEPT,	/* i/o function code	*/ ' 		      &iosb,			/* i/o status block	*/ / 		      &connectast,		/* ast service routine	*/N  		      0,			/* ast parameter	*/ 		      0,			/* p1			*/a 		      0,			/* p2			*/S7 		      &client_itemlst,		/* p3 - remote socket name */ 5 		      &conn_channel,		/* p4 i/o channel for new  */	 						/* connection		*/	 		      0,			/* p5			*/  		      0);			/* p6			*/  !     if ( status & STS$M_SUCCESS )t 	status = iosb.status;         /*$     if ( !(status & STS$M_SUCCESS) ) 	{4 	printf( "Failed to establish listener channel.\n"); 	exit( status ); 	}      */*     }e return;v }   & --------------175779287418F641EED7E6--   ------------------------------  % Date: Thu, 04 Oct 2001 01:47:25 -0400 - From: JF Mezei <jfmezei.spamnot@videotron.ca> 3 Subject: Re: Need help with $QIO calls in C programd, Message-ID: <3BBBF7EA.8F5B7DE7@videotron.ca>   re: programming TCPIP wit ASTs.o   This applies also to DECNET.    K When you have a server that uses ASTs to serve multiple clients at the same=N time, you have to be careful about the sequence of events, especially for very short transactions.n  N I had a server in a previous life which tested 100% fin with a heavy load on aL microvax II. But at the customer, at random times, the software would crash.  I Turns out that my software was receiving the "decnet link closed" messagedI prior to having received the data message (eg: instead of: open link, get7A data, close link, I was getting open link, close link, get data).a  K When I got the "close link" message, I would deallocate the structures. ThetH problem occured when QIO READ completed, it would drop the data into theN memory location it had been told to put it in, but that memory was now used by5 some other portion of the program, having been freed.e  I Solved the problem by putting a timer to ask to deallocate the structuresp@ later if they are still empty after the close has been received.   ------------------------------   Date: 3 Oct 2001 15:40:24 -0700s+ From: hobbesnet@hotmail.com (Scott Squires)* Subject: Re: Need help with RMSe= Message-ID: <ff921edf.0110031440.6918b5ad@posting.google.com>   5 I've narrowed down some more specifics to this bug...E  E 1. If I GETKEY (or FIND then GETSEQ) the record I want to update, anddF then modify the buffer written by GET, and then UPDATE using that same, buffer, it works, and the record is updated.  B 2. If I GET the record, and then use a different buffer (than thatE which was written by GET) to UPDATE the record, it fails (the file iscF not changed to reflect the UPDATE); however, it returns a RMS$_NORMAL.  @ Well, now that I write this it seems even more odd.  Because theC "different" buffer I talk about in statement 2 is actually the same E buffer that GET wrote to when the program started running, just later E in the program execution.  So statement 1 is only true when the callsf immediately follow one another.l   ------------------------------   Date: 3 Oct 2001 20:11:07 +0200c* From: eplan@kapsch.net (Peter LANGSTOEGER), Subject: Re: NETSCAPE 6(.1)NEXT on OpenVMS ?* Message-ID: <3bbb54bb$1@news.kapsch.co.at>  W In article <3bb7330d$1@news.kapsch.co.at>, eplan@kapsch.net (Peter LANGSTOEGER) writes:dK >btw: I'm on MOZILLA 0.9.4 and on JAVA131/FASTVM131 (beta) and I had also afJ >little success with a page containing a small JAVA applet (before MOZILLA$ >crashed - running out of memory ?).  K Sorry folks, no success. One short test is surely not enough (Hi, CJL, RIP)	E I was on the wrong webserver (embedded webserver in ENTERASYS networktN components Vertical-Horizon vs. Matrix E7). No JAVA involved, only JavaScript.N Non-working-pages in NETSCAPE3-VMS are unfortunately fairly common nowadays...  I Seems I was too fast to believe the MOZILLA release notes and hence post:n  M http://www.openvms.compaq.com/openvms/products/ips/mozilla_relnotes.html#java   : There is not even a JAVA console window opening/working...  O btw: Is it only me seeing 0.9.4 worse than 0.9.3 (which was worse than 0.9.2) ?l   --  < Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888 < <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------  % Date: Wed, 03 Oct 2001 22:14:56 -0500t1 From: "David J. Dachtera" <djesys.nospam@fsi.net> & Subject: Re: NT or w2k on alpha 1200 ?& Message-ID: <3BBBD430.D09C9EA@fsi.net>   Tom Linden wrote:t > > > Try posting your query on alphant@lyris.sunbelt-software.com  G C'mon, guys, enough with the guessing games - is this a mailing list oroE what? How does one "post a query", or is it YAML (yet another mailing  list)?   --   David J. Dachtera  dba DJE Systemsc http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/_   ------------------------------  # Date: Thu, 04 Oct 2001 03:50:04 GMTi! From: "ClaudeVMS" <unix@dev.null>  Subject: Re: ODS-2 for FreeBSD@ Message-ID: <M%Qu7.39718$w62.24916121@news1.denver1.co.home.com>  	 Hi Chuck.v  : If you have a website I'll be happy to point to it from my http://www.freevms.orgJ site. I'm looking for any code that comes from the VMS world and is ported toJ the "pagan" OSes ( ;-D). If you send me the code I'll try to get it posted on theH FreeVMS site. Also, lots of projects use Source Forge as a place to park code.	: Then again they might point back to the authors' websites?  L By the way FreeVMS site is not just a Holly Grail go reinvent VMS site. It'sJ mission is to transplant as much VMS concepts and implementations from VMSJ to other OSes. I have been looking hard at using FreeBSD as the basis of a FreeVMS I since it does have a crude concept of IPL in its kernel design. Since I'my marriedi5 I spend more time looking than coding unfortunetly...   ( Check out the site for an email address.   Thanks,t  	 ClaudeVMS   ; "Chuck McCrobie" <mccrobie@cablespeed.com> wrote in message/( news:3BBA6CEF.4C957B8A@cablespeed.com...J > After reading the posts from Hunter Goatley about the ODS-2 reader, I'veF > decided to dust off my NATIVE FreeBSD ODS-2 file system (read-only).H > Its in working order, but needs improvement.  I plan on donating it toJ > the FreeBSD project under the BSD license, so if anyone wishes to try itD > / use it in a product, please contact me.  Source is available andH > binaries can be had.  Currently, it supports FreeBSD I386 v3.3 and 4.? >rF > I've started some work on an RMS library for FreeBSD which correctlyC > handles the record structures and potentially file organizations.tG > However, since this is a hobby project, not much has been done on it.*G > Perhaps some enterprising programmer can modify the ODS-2 reader fromm? > Mr. Goatley and friends for use with this file system driver.d >iI > Sorry, its for FreeBSD only :(  (Maybe portable to other Unix flavors -iE > A Windows NT/2000/XP file system driver is possible, but that wouldeE > require incentive ;)  Just to be complete, there is a product whichlD > wraps a FreeBSD file system with the Windows NT file system driverJ > stuff, so it might be less complicated to use this file system driver to& > wedge in an NT file system driver... > / > If anyone is interested, please contact me...e >i > Chuck McCrobie >P > -- > --o >e   ------------------------------  $ Date: Wed, 3 Oct 2001 19:42:06 +0200 From: zessin@decus.det? Subject: RE: OpenVMS Frequently Asked Questions (FAQ), Part 4/5 * Message-ID: <00A02FCB.F29E2053.1@decus.de>   Hoff Hoffman wrote:  >XML:v >d9 >  Source code of an XML Parser is available from Oracle.d >c >  Also see: > ( >    http://www.python.org/sigs/xml-sig/ >e >m@ >  An XML parser is available as part of OpenVMS V7.3 and later.  : I have made a special EXPAT package for the current Python" for OpenVMS release (V2.1.1-V001).  1 See: http://www.decus.de/~zessin/python2/ext.html    --  
 Uwe Zessin   ------------------------------  $ Date: Thu, 4 Oct 2001 00:51:02 +01001 From: "Chris Townley" <news@townleyc.demon.co.uk>	 Subject: Pathworks on Win2000	B Message-ID: <1002157254.24747.0.nnrp-08.d4e45fa5@news.demon.co.uk>  J An interesting issue using Pathworks on Win 2000. It seems that the normal; method of getting licences is no longer supported underW2K.s  I We cannot get DecNet over TCP/IP working as our license server has an oldnI TCPWare cludge of using 2 IP addresses, so need to resort to client based 
 licensing.  ! Any ideas how this can be setup ?e   --
 Chris Townleyt chris@townleyc.demon.co.uk townleyc@spicers.ltd.uk    ------------------------------  % Date: Thu, 04 Oct 2001 02:12:12 +0200-& From: Michael Joosten <joost@c-lab.de>B Subject: PC164, OVMS7.2, isacfg and graphic cards (MACH64, S3 968)$ Message-ID: <3BBBA95C.1CFB@c-lab.de>   Hello,  @ I'm probably stuffing too many items in one posting, but anyway.  H It began when someone gave me this PC164 (with SRM 5.51) board such thatD I finally could start playing with VMS again. Unfortunately, after IE just had installed a small NetBSD, one cache chip died away and threehD bits went belly up (literally). After long experiments, I eventuallyF found out which chip this was (thanks to SROM's minidebugger) and thatD an undocumented jumper setting for the cache size disabled the cacheG completely. Thus, I could face the next challenge, installing VMS7.2 ona a castrated 500MHz CPU 8-(  H Now, while I'm still waiting to have my DECUS number passed to Mortagar,D I'm trying to find out which of my old graphic cards might work withE DECWindows. There is an ELSA GLoria-8 (nono, not a Synergy, this is aeF '95 vintage early OpenGL beast for NT workstations, with a S3 968, 8MBC and a GLINT chip), and an old ATI MACH64 (210888GX), 2MB, ISA. Botht cards work in SRM.  F If I understood the postings of Fred Kleinsorge correctly, the chances; of getting the GLoria-8 running are somewhat shakey, as theoE SYS$GQSDRIVER might just not already know about the 968 or the RAMDAClG that is used, right? I got so far to write a DECW$USER_CONFIG.DAT entrydG with the card's PCI id, such that SYSMAN IO AUTO/FULL detects the card,l2 but a show dev/full gqa0 still show it as offline: $ sh dev/fu gqa0  ? Device GQA0:, device type DECwindows output device, is offline,n record-orientede(     device, carriage control, shareable.  2     Error count                    0    Operations completed                  0F     Owner process                 ""    Owner UIC                      [SYSTEM]3     Owner process ID        00000000    Dev Prot   $ S:RWPL,O:RWPL,G:RWPL,W:RWPLs6     Reference count                0    Default buffer size                   0    ' and I saw this logged during io config:t   Configure the PCI Bus: TR 2'C BUSARRAY[ 5] id_l = 0x00021011 id_h = 0x00000000, NODE = 040, CSR =n 82FEE000, VE CTOR = 2336rC BUSARRAY[ 6] id_l = 0x88F05333 id_h = 0x00000000, NODE = 048, CSR =I 82FF0000, VE CTOR = 2304AC BUSARRAY[ 8] id_l = 0x04848086 id_h = 0x00000000, NODE = 064, CSR =r 82FF2000, VE CTOR = 2304 C BUSARRAY[ 9] id_l = 0x00011000 id_h = 0x00000000, NODE = 072, CSR =t 831F6000, VE CTOR = 2352sC BUSARRAY[11] id_l = 0x06461095 id_h = 0x00000000, NODE = 088, CSR =  831F8000, VE CTOR = 2384tC Found HW device id_l = 0x88F05333, id_h = 0x00000000 in busarray...,G matching device id_l = 0x88F05333, id_h = 0x00000000 <ELSA GLoria-8> in& dynamic  tabletF Device GQA0, TR 2, CSR 0x82FF0000, Vector -1, Node 48, Units 1, is not	 connected  Configure the ISA Bus: TR 3	C BUSARRAY[ 0] id_l = 0x00000016 id_h = 0x00000000, NODE = 000, CSR =  82FF4000, VE CTOR = 0  H Well, let's try the MACH64. As Google told me, this is supposed to work,H though I have not seen in this SRM a 'shorthand' for registering it with? isacfg. The main problem here is with isacfg: The system always;E complains that the ISA table is corrupt, inits it, but again a isacfgh -all reports it as corrupt:T *** keyboard not plugged in...- ff.fe.fd.fc.fb.fa.f9.f8.f7.f6.f5.f3.f2.f1.f0.  ef.b0.b1.b2.b3.ee.ed.  ERROR: ISA table corrupt!  Initializing table to defaults ec.f4.eb.ea.e9.e8.e7.i  @ Digital AlphaPC 164 500 MHz Console V5.5-1, Jul 14 1999 12:37:45 >>>isacfg -inits   ERROR: ISA table corrupt!/ Initializing table to defaults Initializing table to defaults >>>isacfg -initg   ERROR: ISA table corrupt!  Initializing table to defaults Initializing table to defaults >>>t    ?  Anyway, what should I have to enter in order to get the MACH64g5 recognized? I tried to just provide the minimal info,n  9 isacfg -mk -handle ISA--VGA -etyp 1 -ena 1 -slot 1 -dev 0)  
 >>>sh conf ISAcG Slot    Device  Name            Type         Enabled  BaseAddr  IRQ    n DMAo 0 B         0       MOUSE           Embedded        Yes     60      12A         1       KBD             Embedded        Yes     60      1cA         2       COM1            Embedded        Yes     3f8     4rA         3       COM2            Embedded        Yes     2f8     3/A         4       LPT1            Embedded        Yes     3bc     7*G         5       FLOPPY          Embedded        Yes     3f0     6        2t 1s3         0       ISA--VGA        Singleport      Yes   D Still, this beast does not get detected in the autoconfig stage. Has" anybody the correct incantations? + Or: why is the ISA table always corrupt ???e     -- i* Michael Joosten, SBS C-LAB, joost@c-lab.de* Fuerstenallee 11, 33094 Paderborn, Germany, Phone: +49 5251 606127, Fax: +49 5251 6060658 C-LAB is a cooperation of University Paderborn & SIEMENS   ------------------------------  $ Date: Wed, 3 Oct 2001 22:02:30 -04000 From: "Island Computers" <dbturner@islandco.com>F Subject: Re: PC164, OVMS7.2, isacfg and graphic cards (MACH64, S3 968)/ Message-ID: <trnganqd27fbac@news.supernews.com>	  J You know - for a few bucks you can buy a cheap S3Trio64 1MB PCI Video Card which is enough for DECwindows   It's a hell of a lot easier !    -- David Turner   We sell Alpha systems & partsS http://www.islandco.com  sales@islandco.com Island Computers US Corp.; 2700 Gregory Street  Savannah GA 31404f Tel: 912 447 6622, Fax: 912 201 0096t ICQ#: 130698221 1 Michael Joosten <joost@c-lab.de> wrote in message  news:3BBBA95C.1CFB@c-lab.de... > Hello, >/B > I'm probably stuffing too many items in one posting, but anyway. >CJ > It began when someone gave me this PC164 (with SRM 5.51) board such thatF > I finally could start playing with VMS again. Unfortunately, after IG > just had installed a small NetBSD, one cache chip died away and three*F > bits went belly up (literally). After long experiments, I eventuallyH > found out which chip this was (thanks to SROM's minidebugger) and thatF > an undocumented jumper setting for the cache size disabled the cacheI > completely. Thus, I could face the next challenge, installing VMS7.2 on) > a castrated 500MHz CPU 8-( > J > Now, while I'm still waiting to have my DECUS number passed to Mortagar,F > I'm trying to find out which of my old graphic cards might work withG > DECWindows. There is an ELSA GLoria-8 (nono, not a Synergy, this is ahH > '95 vintage early OpenGL beast for NT workstations, with a S3 968, 8MBE > and a GLINT chip), and an old ATI MACH64 (210888GX), 2MB, ISA. Botha > cards work in SRM. >0H > If I understood the postings of Fred Kleinsorge correctly, the chances= > of getting the GLoria-8 running are somewhat shakey, as the G > SYS$GQSDRIVER might just not already know about the 968 or the RAMDACrI > that is used, right? I got so far to write a DECW$USER_CONFIG.DAT entryeI > with the card's PCI id, such that SYSMAN IO AUTO/FULL detects the card,=4 > but a show dev/full gqa0 still show it as offline: > $ sh dev/fu gqa0 >)A > Device GQA0:, device type DECwindows output device, is offline,  > record-orientedr* >     device, carriage control, shareable. > 4 >     Error count                    0    Operations > completed                  03 >     Owner process                 ""    Owner UICn
 > [SYSTEM]2 >     Owner process ID        00000000    Dev Prot > S:RWPL,O:RWPL,G:RWPL,W:RWPLv8 >     Reference count                0    Default buffer > size                   0 >s >q) > and I saw this logged during io config:  >t > Configure the PCI Bus: TR 2oE > BUSARRAY[ 5] id_l = 0x00021011 id_h = 0x00000000, NODE = 040, CSR =s > 82FEE000, VE
 > CTOR = 2336	E > BUSARRAY[ 6] id_l = 0x88F05333 id_h = 0x00000000, NODE = 048, CSR =a > 82FF0000, VE
 > CTOR = 2304*E > BUSARRAY[ 8] id_l = 0x04848086 id_h = 0x00000000, NODE = 064, CSR =s > 82FF2000, VE
 > CTOR = 2304hE > BUSARRAY[ 9] id_l = 0x00011000 id_h = 0x00000000, NODE = 072, CSR =  > 831F6000, VE
 > CTOR = 2352	E > BUSARRAY[11] id_l = 0x06461095 id_h = 0x00000000, NODE = 088, CSR =u > 831F8000, VE
 > CTOR = 2384uE > Found HW device id_l = 0x88F05333, id_h = 0x00000000 in busarray...lI > matching device id_l = 0x88F05333, id_h = 0x00000000 <ELSA GLoria-8> in-	 > dynamic4 > table7H > Device GQA0, TR 2, CSR 0x82FF0000, Vector -1, Node 48, Units 1, is not > connected  > Configure the ISA Bus: TR 3tE > BUSARRAY[ 0] id_l = 0x00000016 id_h = 0x00000000, NODE = 000, CSR =- > 82FF4000, VE
 > CTOR = 0 >nJ > Well, let's try the MACH64. As Google told me, this is supposed to work,J > though I have not seen in this SRM a 'shorthand' for registering it withA > isacfg. The main problem here is with isacfg: The system alwayscG > complains that the ISA table is corrupt, inits it, but again a isacfg  > -all reports it as corrupt:a  > *** keyboard not plugged in.../ > ff.fe.fd.fc.fb.fa.f9.f8.f7.f6.f5.f3.f2.f1.f0.  > ef.b0.b1.b2.b3.ee.ed.m > ERROR: ISA table corrupt!"  > Initializing table to defaults > ec.f4.eb.ea.e9.e8.e7.d > B > Digital AlphaPC 164 500 MHz Console V5.5-1, Jul 14 1999 12:37:45 > >>>isacfg -inito >l > ERROR: ISA table corrupt!t  > Initializing table to defaults  > Initializing table to defaults > >>>isacfg -initn >O > ERROR: ISA table corrupt!p  > Initializing table to defaults  > Initializing table to defaults > >>>  >o >wA >  Anyway, what should I have to enter in order to get the MACH64d7 > recognized? I tried to just provide the minimal info,l >a; > isacfg -mk -handle ISA--VGA -etyp 1 -ena 1 -slot 1 -dev 0l >  > >>>sh conf > ISA E > Slot    Device  Name            Type         Enabled  BaseAddr  IRQm > DMAs > 0oD >         0       MOUSE           Embedded        Yes     60      12C >         1       KBD             Embedded        Yes     60      1eC >         2       COM1            Embedded        Yes     3f8     4 C >         3       COM2            Embedded        Yes     2f8     3rC >         4       LPT1            Embedded        Yes     3bc     7nC >         5       FLOPPY          Embedded        Yes     3f0     6f > 2t > 1e5 >         0       ISA--VGA        Singleport      Yese >dF > Still, this beast does not get detected in the autoconfig stage. Has# > anybody the correct incantations? - > Or: why is the ISA table always corrupt ???u >t >  > --, > Michael Joosten, SBS C-LAB, joost@c-lab.de, > Fuerstenallee 11, 33094 Paderborn, Germany. > Phone: +49 5251 606127, Fax: +49 5251 606065: > C-LAB is a cooperation of University Paderborn & SIEMENS   ------------------------------  # Date: Thu, 04 Oct 2001 01:47:52 GMT-/ From: "Tom Simpson" <simpsont@xxx.mediaone.net>@ Subject: PGP for OpenVMS??G Message-ID: <cdPu7.28998$d44.6665004@typhoon.jacksonville.mediaone.net>@  A I'm looking for PGP for OpenVMS Alpha.  I've checked the VMS faq,nC NAI's site and Deja News and had no luck locating a current versiona> of this utility.  Most of the VMS links to PGP sites are dead.  E The only thing I found is v2.6.3.  Is that a useable revision?  Is itr/ compatible with the PC versions 6.5 and higher?s  ? We are willing to purchase the software if a current version ise available...  D As it stands now, I have to put a PC in the production loop in orderB to encrypt/decrypt files for transmission.  I'd rather not do that if possible.   Regards, Tomu   ------------------------------   Date: 3 Oct 2001 20:36:55 -0700h1 From: nothome@spammers.are.scum (Malcolm Dunnett)w Subject: Re: PGP for OpenVMS??, Message-ID: <Qu68FETYvESV@malvm6.mala.bc.ca>  H In article <cdPu7.28998$d44.6665004@typhoon.jacksonville.mediaone.net>, 5     "Tom Simpson" <simpsont@xxx.mediaone.net> writes:9  C > I'm looking for PGP for OpenVMS Alpha.  I've checked the VMS faq,-E > NAI's site and Deja News and had no luck locating a current version @ > of this utility.  Most of the VMS links to PGP sites are dead. > @ > The only thing I found is v2.6.3.  Is that a useable revision?  2    I'm using 2.6.3 and it seems to work just fine.   ------------------------------  % Date: Wed, 03 Oct 2001 22:46:56 +0100c5 From: Martyn <mpatt644-dropthis@netscapeonline.co.uk>.C Subject: Re: Problem With VMS Backup Save-Set Spanning Tape Volumeso4 Message-ID: <3BBB8750.B1114EFC@netscapeonline.co.uk>   "Richard D. Piccard" wrote:  > Q > My experience with 9-track tapes, many moons ago, was that if you do a straighteG > restore (NOT /IMAGE), and put the second tape in, it will generate anwP > INFORMATIONAL error message about not being the start of the saveset, but willQ > proceed.  It will likely restore a fragment of whatever specific file spans theiQ > volume break, so you may well want to do a BACKUP/LIST and print the first pageM+ > to identify the file that may be partial.. > S > I don't know whether $ SET MAGTAPE/SKIP will work to get past a bad spot on a DATwL > tape.  On the old TS-11 ("TapeStretcher"), when the BACKUP command startedO > "polishing" the heads (as it kept trying to read the bad part of the tape), ItP > would CONTROL-Y, then SPAWN, then SET MAGTAPE/SKIP to get it past the problem,  F I remember when I was an operator and we needed to restore from a duffH 2400 foot 9 track tape on a TE16, it kept bombing out due to a bad spot.H In the end we had to take the drive offline just before the bad spot andA manually wind the tape on past the duff bit, reload the drive andBE continue the backup which complained about missing blocks but we onlyPH lost 1 file, you can't do that with these darn new fangled drives..... IH don't know if set mag/skip wasn't available then or whether we were just  so dumb we didn't know about it!   ------------------------------    Date: 04 Oct 2001 05:35:07 +0800, From: Paul Repacholi <prep@prep.synonet.com>C Subject: Re: Problem With VMS Backup Save-Set Spanning Tape Volumesd- Message-ID: <87r8sk76gk.fsf@prep.synonet.com>a  / "Richard L. Dyson" <rickdyson@home.com> writes:m  F > I have been asked to lend a hand in trying to restore some data fromC > a multi-tape DAT backup (4 volumes I believe) where we may have at3 > spot part way through tape #1 that can't be read.a  F > Is it going to be possible to plow through the bad spot and continueB > on through the other 3 tapes, etc.?  Or will we just be screwed?  eB You really need Save Set Utility, SSU, for this. Copy the save set$ to real media and restore from that.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda.s@                                              West Australia 6076. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------   Date: 03 Oct 2001 21:24:00 GMT' From: "Jim Strehlow" <jims@data911.com>e7 Subject: Re: Problem with vms712_update-v0300 download.e0 Message-ID: <9pfvlg$274@dispatch.concentric.net>  / Do you FTP directly from your OpenVMS computer?y  - using tcp/ip services?o  - using other vendor?  F or do you FTP to a pc workstation first and then copy to OpenVMS etc.?  / Be certain to use binary instead of ASCII copy.s   Jim Strehlow, JimS@data911.com Data911 Systems Manager, Alameda, CAe  / "Let them do their worst. We will do our best."f    - <Steve.Spires@yellgroup.com> wrote in messageP) news:00256ADA.005453F4.00@quegw01.btyp...s >e, > Problem with vms712_update-v0300 download.I > When I download this file from the .au ftp site via the Compaq website,. the file- > I get has the following characteristics ...y   ------------------------------   Date: 3 Oct 2001 13:16:34 -0700~# From: pkluesen@techmate.com (Patty) 0 Subject: Project Manager - EMC Storage Architect= Message-ID: <37e9bb33.0110031216.46d63e63@posting.google.com>n   Salary:     Negotiable Experience: 4 years or moresO Required:   EMC Storage Architectural Design, Project Management, Unix. OS/390 i             a plus!o             K             The EMC Storage Architect will manage projects, people and the eN             allocation, performance, capacity and distribution configurations.O             The qualified candidate must have strong hands on knowledge of EMC  H             Storage Architectural Design.  Experience with various Unix J             platforms a must.  Excellent written and verbal communication E             skills.  Excellent customer service skills as well.      e     Travel:     60%uC Benefits:   Comprehensive benefits package, 401k and bonus program.s! Contact:    pkluesen@techmate.com    ------------------------------  $ Date: Wed, 3 Oct 2001 13:53:22 -0400' From: "Bill Todd" <billtodd@foo.mv.com>1+ Subject: Re: Public domain VMS clustering ? ( Message-ID: <9pfj6j$aik$1@pyrite.mv.net>  : "JF Mezei" <jfmezei.spamnot@videotron.ca> wrote in message& news:3BBA61DC.3E3A4756@videotron.ca...B > Hoff said something interesting in another thread about building applicationsL > with a generic layer for inter-process communications to allow for changes8 > over time (was discussin how event flags have limits). >oH > This got me to think. Wouldn't it be great is some of the VMS specific stuff E > were to become "public domain" and used by the industry as de-facto, standards ?t  I It would be (as has all too often been the case over the years) too late.sL IBM opened its 'Phoenix' clustering architecture to the world years ago, andK Linux people are already busily working on standardizing a bunch of clusterTF components (e.g., membership and connection managers) for higher-level product use.   > K > For instance, what if the SCS protocols, including the ICC (intra clusteroE > comms) were to become available to anyone, in the same way as Java,  wouldn'tC > it eventually allow clusters to be formed with multiple differentC	 operatingn > systems ?e  J Hardly.  The details of how all the pieces of a cluster work together varyH widely among implementations, from the very low level (the communicationI protocols) through the intermediate building blocks (who's in, who's out,eK how you know, what you do) to the application-visible pieces (do you really I expect the rest of the world to adopt ODS-2 file systems, or VMS to adopt 	 theirs?).    > I > Imagine if IBM were to adopt the VMS clustering protocols, and you'd ber ableL > to cluster a VMS system with an MVS system or integrate a Sun cluster with au > VMS cluster.  J Beyond the obstacles mentioned above, you've now introduced the additional, difficulty of endian-heterogeneous clusters.   > H > It would bring VMS into the limelight and outline the fact that VMS is still  > the leader in clustering.   J You really need to take some time after indulging before posting.  The VMSC concept of clustering is far too closely-coupled to be easy to makeUK heterogeneous in any manner that would make all participants happy.  A more H loosely-coupled implementation (such as the shared file systems and comm? mechanisms that are already starting to appear in heterogeneous0I environments - and which *may* be the direction the Linux standardization5J efforts are heading in) makes sense and will likely appear, but, as usual,K VMS will likely be left out in the cold rather than shine in the limelight.r   - bill   ------------------------------  # Date: Thu, 04 Oct 2001 04:01:44 GMT ! From: "ClaudeVMS" <unix@dev.null>e+ Subject: Re: Public domain VMS clustering ?t@ Message-ID: <IaRu7.39758$w62.24929966@news1.denver1.co.home.com>   Hey Hey Hey!!!  ? VMS specific "stuff" is making it into the public domain. Go toyE http://www.freevms.org and you will see the projects I have uncovereda2 that are in fact moving VMS features into Linux!!!  	 Examples:g  % One group has ported DECnet to Linux.5H Another group was working on a Distributed Lock Manager for Linux as the baisis of a more general web file system.  J And the best for last: IBM developed a DLM clone and have been using it in their clusteringI software for AIX. They built their DLM to work just like the VMS DLM. Bute wait there's/ more!!! IBM ported the DLM from AIX to Linux!!!n  K I have read the VAXClusters book by Davis and the SCS stuff looks just like, DECnetH which DECnet on Linux would provide the source code for. The VAXClusters bookH provides the key information for how the cluster works. The IBM DLM code provides theI source code to allow one to truely integrate the DLM into an OS kernel. Ia believe the implementation is all user mode.S  L I believe there are enough pieces including an ODS-2 implementation by Chuck McCrobieJ for FreeBSD that one could actually approach building a VMS like operating system.   ? check out the sight and see the links to what others are doing.    Later,  	 ClaudeVMS   @ "Tim Llewellyn" <tim.llewellyn@cableinet.co.uk> wrote in message) news:3BBA8295.DCD9844A@cableinet.co.uk...  >  >  > JF Mezei wrote:  > > D > > Hoff said something interesting in another thread about building applicationsF > > with a generic layer for inter-process communications to allow for changesI: > > over time (was discussin how event flags have limits). > >BJ > > This got me to think. Wouldn't it be great is some of the VMS specific stuff7G > > were to become "public domain" and used by the industry as de-factoS standards ?- >-F > Wasn't the Sun Fortran compiler's compatibility with VAX fortran non > stantdardrJ > extensions one of the reasons they (Sun) cleaned up in academia a decade > or more back?  >)+ > Anyway, did you look at any NT internals?c >	 > >oE > > For instance, what if the SCS protocols, including the ICC (intrad clustersG > > comms) were to become available to anyone, in the same way as Java,n wouldn'tE > > it eventually allow clusters to be formed with multiple differentm	 operatingm
 > > systems ?; > >0K > > Imagine if IBM were to adopt the VMS clustering protocols, and you'd be  ableI > > to cluster a VMS system with an MVS system or integrate a Sun cluster5 with a > > VMS cluster. > E > well, I guess Tru64 and linux and HP/UK would be a step forward, to   > demonstrate the possibilities.% > But then, why not keep it in house.  >t > > J > > It would bring VMS into the limelight and outline the fact that VMS is stilla > > the leader in clustering.c > 6 > Wouldn't some decent marketting be more appropriate? >e	 > regardsr > -- > Tim.Llewellyn@cableinet.co.uk  >pD > Standard disclaimer applies. My views in no way represent those of# > my employers or service provider.a   ------------------------------  $ Date: Thu, 4 Oct 2001 19:22:44 -0400! From: "Airnews" <Kuff@Tessco.Com>M Subject: PurveyoroO Message-ID: <713981D6627AEAE3.BA0724D90CE94486.60F52888B33960E2@lp.airnews.net>i  L     Does anyone know the max post size under purveyor? I understand that theG buffer comes in at 256 bytes per qio until the buffer is exhausted, butlK what's the max size of the message that can be passed by the mbox? If thereaH is no size in the crembx service than I belive that the sysgen parameterH default mailbox max quota defmbxquo is used.... we're seeing an issue at greater than 1024....i   ------------------------------  % Date: Wed, 03 Oct 2001 12:38:31 -0500r0 From: arturo saavedra <arturo.saavedra@wcom.com>) Subject: RE: Reset error account on AlphaSC Message-ID: <MOEAJKGGEIMGCCPEPJBHOECLFLAA.arturo.saavedra@wcom.com>U  K I tried that and it has crashed every single one of my Alphas .. careful :):       -----Original Message-----D From: norm.raphael@jamesbury.com [mailto:norm.raphael@jamesbury.com]) Sent: Wednesday, October 03, 2001 9:51 AMh To: Info-VAX@Mvb.Saic.Como) Subject: Re: Reset error account on Alphao        A There is an unsupported utility from Hardware Support called ZDEC > (Zero Device Error Count) that will do this, device by device.        3 P.Hansford@elsevier.co.uk on 10/03/2001 09:29:59 AM   + Please respond to P.Hansford@elsevier.co.uk    To:   Info-VAX@mvb.saic.coma cc:s& Subject:  Reset error account on Alpha         HiG I was wondering if there was any official or even unofficial utility or0L piece of Freeware to reset the error count on Alphas apart from the possiblyA perilous process of SDA and Delta debugger to zap the contents of  UCB+UCB$L_ERRCNT.  Cheers Paul **Paul Hansford  **Elsevier Science UK2 **p.hansford@elsevier.co.uk    ------------------------------  $ Date: Wed, 3 Oct 2001 17:59:14 -0400  From: norm.raphael@jamesbury.com) Subject: RE: Reset error account on Alpha04 Message-ID: <C2256ADA.00787CD0.00@jklh21.valmet.com>  
 None of mine.         2 arturo.saavedra@wcom.com on 10/03/2001 01:38:31 PM  * Please respond to arturo.saavedra@wcom.com   To:   Info-VAX@mvb.saic.comi cc: * Subject:  RE: Reset error account on Alpha        K I tried that and it has crashed every single one of my Alphas .. careful :)        -----Original Message-----D From: norm.raphael@jamesbury.com [mailto:norm.raphael@jamesbury.com]) Sent: Wednesday, October 03, 2001 9:51 AMm To: Info-VAX@Mvb.Saic.Com,) Subject: Re: Reset error account on Alphah        A There is an unsupported utility from Hardware Support called ZDECs> (Zero Device Error Count) that will do this, device by device.        3 P.Hansford@elsevier.co.uk on 10/03/2001 09:29:59 AMu  + Please respond to P.Hansford@elsevier.co.uk.   To:   Info-VAX@mvb.saic.com2 cc:0& Subject:  Reset error account on Alpha         HiG I was wondering if there was any official or even unofficial utility or L piece of Freeware to reset the error count on Alphas apart from the possiblyA perilous process of SDA and Delta debugger to zap the contents ofe UCB+UCB$L_ERRCNT.i Cheers Paul **Paul Hansfordf **Elsevier Science UKI **p.hansford@elsevier.co.uki   ------------------------------  % Date: Wed, 03 Oct 2001 17:34:41 -0500 0 From: arturo saavedra <arturo.saavedra@wcom.com>) Subject: RE: Reset error account on AlphasC Message-ID: <MOEAJKGGEIMGCCPEPJBHIEDDFLAA.arturo.saavedra@wcom.com>   J Although back then I was running 7.1.. zeroing errors on hsj50s.. have not tried it on our 7.21h1 systems.e   arturo saavedra/vmst vnet   671 5487  did    512 684 5487e 800    800 999 4788 x5487  cell   512 663 1581  im     nuclearArte pager  1 888 mci 2way #1721372+        mailto:arturo.saavedra+page@wcom.com      -----Original Message-----D From: norm.raphael@jamesbury.com [mailto:norm.raphael@jamesbury.com]) Sent: Wednesday, October 03, 2001 4:59 PMe To: Info-VAX@Mvb.Saic.Com ) Subject: RE: Reset error account on Alpha         
 None of mine.         2 arturo.saavedra@wcom.com on 10/03/2001 01:38:31 PM  * Please respond to arturo.saavedra@wcom.com   To:   Info-VAX@mvb.saic.com  cc:d* Subject:  RE: Reset error account on Alpha        K I tried that and it has crashed every single one of my Alphas .. careful :)-       -----Original Message-----D From: norm.raphael@jamesbury.com [mailto:norm.raphael@jamesbury.com]) Sent: Wednesday, October 03, 2001 9:51 AMo To: Info-VAX@Mvb.Saic.Com-) Subject: Re: Reset error account on Alpha1        A There is an unsupported utility from Hardware Support called ZDECP> (Zero Device Error Count) that will do this, device by device.        3 P.Hansford@elsevier.co.uk on 10/03/2001 09:29:59 AM   + Please respond to P.Hansford@elsevier.co.uka   To:   Info-VAX@mvb.saic.comc cc: & Subject:  Reset error account on Alpha         HiG I was wondering if there was any official or even unofficial utility orhL piece of Freeware to reset the error count on Alphas apart from the possiblyA perilous process of SDA and Delta debugger to zap the contents ofi UCB+UCB$L_ERRCNT.  Cheers Paul **Paul Hansford  **Elsevier Science UKl **p.hansford@elsevier.co.uky   ------------------------------   Date: 3 Oct 2001 11:36:45 -0700e  From: alanb@cloud9.net (Alan B.)) Subject: Re: Reset error account on Alpha0= Message-ID: <88599d89.0110031036.17d81af0@posting.google.com>c  B Here's a macro program; I don't remember where I got it. I relieve0 myself of all responsibility for system crashes.   ;vD ; Hack to zero the system and device error counters on OpenVMS Alpha ;s ; Jur van der Burg ; Compaq ;p ; $ macro zeroerrc ; $ link/sysexe zeroerrA ; -         .library        'sys$library:lib.mlb'u           $ucbdeff         $ssdef           .entry  zeroerr,^m<>  #         $cmkrnl_s       routin=zerou         ret            .entry  zero,^m<>            movab   nocrash,(fp)         movl    g^ctl$gl_pcb,r4-         jsb     g^sch$iolockwe         clrl    r115   10$:    jsb     g^ioc$scan_iodbo         blbc    r0,20$!         clrl    ucb$l_errcnt(r10)          brb     10$i    20$:    clrl    g^exe$gl_memerrs!         clrl    g^exe$gl_mchkerrs          movl    g^ctl$gl_pcb,r4          jsb     g^sch$iounlock         setipl  #0         movzbl  #ss$_normal,r0         retN           .entry  nocrash,^m<>           setipl  #0         $exit_sT           .end    zeroerro            \ norm.raphael@jamesbury.com wrote in message news:<C2256ADA.005148A0.00@jklh21.valmet.com>...C > There is an unsupported utility from Hardware Support called ZDECC@ > (Zero Device Error Count) that will do this, device by device. >  >  >  > 5 > P.Hansford@elsevier.co.uk on 10/03/2001 09:29:59 AM  > - > Please respond to P.Hansford@elsevier.co.ukl >  > To:   Info-VAX@mvb.saic.com  > cc:c( > Subject:  Reset error account on Alpha >  >  >  >  > HiI > I was wondering if there was any official or even unofficial utility orAN > piece of Freeware to reset the error count on Alphas apart from the possiblyC > perilous process of SDA and Delta debugger to zap the contents oft > UCB+UCB$L_ERRCNT.  > Cheers > Paul > **Paul Hansfordt > **Elsevier Science UKt > **p.hansford@elsevier.co.ukl   ------------------------------  # Date: Wed, 03 Oct 2001 18:42:29 GMTe= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)t) Subject: Re: Reset error account on Alphan0 Message-ID: <00A02FA2.16F8C381@SendSpamHere.ORG>  ` In article <88599d89.0110031036.17d81af0@posting.google.com>, alanb@cloud9.net (Alan B.) writes:C >Here's a macro program; I don't remember where I got it. I relievec1 >myself of all responsibility for system crashes.W >  >;E >; Hack to zero the system and device error counters on OpenVMS Alphap >; >; Jur van der Burgd	 >; Compaq< >; >; $ macro zeroerr >; $ link/sysexe zeroerr >;. >        .library        'sys$library:lib.mlb' >  >        $ucbdef >        $ssdefI >l >        .entry  zeroerr,^m<>p >p$ >        $cmkrnl_s       routin=zero >        ret >  >        .entry  zero,^m<> >r >        movab   nocrash,(fp)   >        movl    g^ctl$gl_pcb,r4 >        jsb     g^sch$iolockw >        clrl    r11 >e  >10$:    jsb     g^ioc$scan_iodb >        blbc    r0,20$i" >        clrl    ucb$l_errcnt(r10) >        brb     10$ > ! >20$:    clrl    g^exe$gl_memerrs6" >        clrl    g^exe$gl_mchkerrs  >        movl    g^ctl$gl_pcb,r4 >        jsb     g^sch$iounlocks >        setipl  #0, >        movzbl  #ss$_normal,r0l >        ret >h >        .entry  nocrash,^m<>l >s >        setipl  #0l >        $exit_s >- >        .end    zeroerr  H This zaps more than just a specified device.  In fact, you can't specify the device.w     --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COMo            cJ   "And of course, I'm a genius, so people are naturally drawn to my fiery I   intellect.  Their admiration overwhelms their envy!" -- Calvin & HobbesS   ------------------------------   Date: 3 Oct 2001 15:49:23 -0500e/ From: brooks@cuebid.zko.dec.nospam (Rob Brooks) ) Subject: Re: Reset error account on Alphae- Message-ID: <XUHAcmP9F6OF@cuebid.zko.dec.com>u  p In article <00A02FA2.16F8C381@SendSpamHere.ORG>, system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) writes:b > In article <88599d89.0110031036.17d81af0@posting.google.com>, alanb@cloud9.net (Alan B.) writes:D >>Here's a macro program; I don't remember where I got it. I relieve2 >>myself of all responsibility for system crashes. >> >>;.F >>; Hack to zero the system and device error counters on OpenVMS Alpha >>;3 >>; Jur van der Burg
 >>; Compaq >>;g >>; $ macro zeroerr  >>; $ link/sysexe zeroerr  >>; / >>        .library        'sys$library:lib.mlb'e >> >>        $ucbdeft >>        $ssdef >> >>        .entry  zeroerr,^m<> >>% >>        $cmkrnl_s       routin=zerom
 >>        ret  >> >>        .entry  zero,^m<>o >> >>        movab   nocrash,(fp)! >>        movl    g^ctl$gl_pcb,r4  >>        jsb     g^sch$iolockwh >>        clrl    r11w >>! >>10$:    jsb     g^ioc$scan_iodbh >>        blbc    r0,20$# >>        clrl    ucb$l_errcnt(r10)l >>        brb     10$e >>" >>20$:    clrl    g^exe$gl_memerrs# >>        clrl    g^exe$gl_mchkerrs ! >>        movl    g^ctl$gl_pcb,r4   >>        jsb     g^sch$iounlock >>        setipl  #0  >>        movzbl  #ss$_normal,r0
 >>        rets >> >>        .entry  nocrash,^m<> >> >>        setipl  #0 >>        $exit_s: >> >>        .end    zeroerrl > J > This zaps more than just a specified device.  In fact, you can't specify
 > the device.i  C 	It also doesn't take into account multipath devices.  This utilitye? would only zap the primary UCB for a multipath set, leaving anyn* secondary UCB with its error count intact.   -- t  M Rob Brooks    VMS Engineering -- I/O Exec Group     brooks!cuebid.zko.dec.coms   ------------------------------  # Date: Wed, 03 Oct 2001 20:23:57 GMTt1 From: "Mark D. Jilson" <jilly@clarityconnect.com>M) Subject: Re: Reset error account on Alphac2 Message-ID: <3BBB7442.2D007825@clarityconnect.com>  ,         .TITLE  ZDEC Zero Device Error Count         .IDENT  /1.2/a     ;++a     ;B     ; Title:/     ;   ZDEC.MAR - Zero-out device error count.r     ;a     ; Version:
     ;   1-003s     ;k     ; Facility:i      ;   System Management Tools.     ;      ; Abstract:nC     ;   This program zeros-out the device error count stored in theeB     ;   UCB for a particular device. This is useful when you wish 5     ;   reset this value without booting the machine.y     ;a7     ;   THIS PROGRAM IS UNSUPPORTED AND PROVIDED AS IS,      ;   USE AT OWN RISK.     ;n     ;   Building sequence:     ;t!     ;           $ macro/list zdecs+     ;           $ link/map/full/sysexe zdect     ;                ;   Calling sequence:a     ;r,     ;           $ ZDEC := $mydev:[mydir]ZDEC*     ;           $ ZDEC [/OVERRIDE]  device     ;d     ; Environment:C     ;   CMKRNL privilege required, I/O data base is locked, programp!     ;   executed at elevated IPL.p     ;i
     ; Author: C     ;   Mark Oakley     DuPont Experimental Station     12-Nov-1984b>     ;   This program is based upon the VARY program written by.     ;   Gary Grebus of Battelle Columbus Labs.     ;e     ; Modifications:     ;oA     ;  17-Nov-1984      Mark Oakley     Added /OVERRIDE to handleo allocated devices.     ;uD     ;  25-Nov-1985      Mark Oakley     Modified to correctly secure physical terminal F     ;                           name that is associated with a virutal	 terminal.T     ; D     ;  23-May-1986      Mark Oakley     Modified to zero-out CPU and memory errors.     ;nG     ;  14-Jul-1988      Ted Nieland     Modified to work under VMS 5.0.l     ;iA     ;  15-Jul-2001  Paul Gallo  Compaq Computer Corporation, CSC,l Colorado     ; D     ;                           Modified to work under OpenVMS Alpha 7.x.     ; A     ;  19-Jul-2001  Paul Gallo  Compaq Computer Corporation, CSC,l Colorado     ; F     ;                           Modified to return errors to the user.     ;gD     ;  20-Jul-2001      Paul Gallo      Compaq Computer Corporation,
 CSC, ColoradoM     ;                   E     ;                           Modified for multipath devices, as we  need to7F     ;                           clear the errors on each device in the path.t     ;      ;--           %         .SBTTL  Symbols, Macros, Datao     -         .LIBRARY        /SYS$LIBRARY:LIB.MLB/h     1         $TPADEF         ; Symbols for LIB$TPARSE.l4         $SSDEF          ; Symbols for return status.1         $UCBDEF         ; Symbols for device ucb.o6         $STSDEF         ; Symbols for returned status.6         $DVIDEF         ; Symbols for $GETDVI service.2         $DCDEF          ; Symbols for device type.<         $MPDEVDEF       ; Symbols for multipath information.          ;e#     ; Macro to handle return codes.      ;L#         .MACRO  ON_ERR  THERE,?HEREr         BLBS    R0,HEREC         BRW     THERE,     HERE:       .ENDM   ON_ERR     3         .PSECT  ZDEC_DATA,RD,WRT,NOEXE,LONG,SHR,PICr     >     CMD_BUF:                    ; Buffer to hold command line.         .BLKB   80     CMD_BUF_SIZ = . - CMD_BUFi     E     CMD_BUF_DESC:                       ; Descriptor to command line.e         .LONG    CMD_BUF_SIZ         .ADDRESS CMD_BUF          PROMPT:t         .ASCID  /Device: /     =     PARSE_BLK:                  ; Parse block for LIB$TPARSE.g         .LONG   TPA$K_COUNT0C         .LONG   TPA$M_ABBREV    ; Permit unambiguous abbreviations.M         .BLKB   TPA$K_LENGTH0-8l     E     OVERRIDE:                   ; Zero out allocated devices, if set..         .LONG   0z     B     LOG_DEV_DESC:                       ; Descriptor for device on
 command line.p         .BLKQ   1h     F     CPU_STG:                    ; Used to compare against command line parm.m         .ASCII  /CPU/e     F     MEM_STG:                    ; Used to compare against command line parm.o         .ASCII  /MEMORY/          7     DEV_BUF:                    ; Buffer to hold devicea name.                        .BLKB   40     DEV_BUF_SIZ = . - DEV_BUF      G     DEV_BUF_DESC:                       ; Descriptor pointing to devicec name.u         .LONG    DEV_BUF_SIZ         .ADDRESS DEV_BUF     C     PID:                                ; Owner of device (if any).          .BLKL   1u     B     DEV_ITEM_LIST:                      ; Device list for $GETDVI.G         .WORD    DEV_BUF_SIZ    ; Make sure we a have a physical devices name.f         .WORD    DVI$_DEVNAM         .ADDRESS DEV_BUF         .ADDRESS DEV_BUF_DESCr@         .WORD    4              ; See if someone has this device
 allocated.         .WORD    DVI$_PIDt         .ADDRESS PID         .LONG    0         .WORD    47         .WORD    DVI$_DEVCLASS  ; Check for a terminal.          .ADDRESS DEV_CLASS         .LONG    03         .LONG    0              ; End if item list.      @     TT_ITEM_LIST:                       ; Item list if we have a	 terminal. G         .WORD    DEV_BUF_SIZ    ; Make sure we a have a physical device  name.i"         .WORD    DVI$_TT_PHYDEVNAM         .ADDRESS DEV_BUF         .ADDRESS DEV_BUF_DESCe         .LONG    0          DEV_CLASS:         .LONG   1e     G     K_ARG:                              ; Argument list for kernel-mode2 routine.4         .LONG    1              ; Just one argument.B         .ADDRESS DEV_BUF_DESC   ; Pass descriptor for device name.                   .SBTTL  State Tablet5         .PSECT  ZDEC_STATES,RD,NOWRT,EXE,LONG,SHR,PIC      -         $INIT_STATE     STATE_TABLE,KEY_TABLE               $STATE  STARTe         $TRAN   '/',QUAL.         $TRAN   TPA$_SYMBOL,DEV,,,LOG_DEV_DESC"         $TRAN   TPA$_EOS,TPA$_EXIT              $STATE  DEVt"         $TRAN   TPA$_EOS,TPA$_EXIT         $TRAN   '/',QUAL         $TRAN   ':',COLONn              $STATE  COLON-         $TRAN   '/',QUAL"         $TRAN   TPA$_EOS,TPA$_EXIT              $STATE  QUAL'         $TRAN   'OVERRIDE',,,1,OVERRIDEr$         $TRAN   'NOOVERRIDE',,CLR_OV              $STATE.         $TRAN   TPA$_SYMBOL,DEV,,,LOG_DEV_DESC"         $TRAN   TPA$_EOS,TPA$_EXIT              $END_STATE                             .SBTTL  Main program3         .PSECT  ZDEC_CODE,RD,NOWRT,EXE,LONG,SHR,PICt0         .ENTRY  ZDEC,^M<R2,R3,R4,R5,R6,R7,R8,R9>          ;o     ; Get the command line.a     ;o8         PUSHAL  CMD_BUF_DESC            ; Return length.A         PUSHAL  PROMPT                  ; Prompt for device name.t=         PUSHAL  CMD_BUF_DESC            ; Place to hold line./$         CALLS   #3,G^LIB$GET_FOREIGN         ON_ERR  ERR_EXIT     ?         PUSHAL  CMD_BUF_DESC            ; Upper case the input.n         PUSHAL  CMD_BUF_DESC         CALLS   #2,G^STR$UPCASEa         ON_ERR  ERR_EXIT     H         MOVL    CMD_BUF_DESC,-          ; Set length of string to parse.)                 PARSE_BLK+TPA$L_STRINGCNTuB         MOVL    CMD_BUF_DESC+4,-        ; Set address of string to parse.)                 PARSE_BLK+TPA$L_STRINGPTReD         PUSHAL  KEY_TABLE               ; Parse the command line for?         PUSHAL  STATE_TABLE             ; qualifier and device.:         PUSHAL  PARSE_BLKh'     ;## vax     CALLS   #3,G^LIB$TPARSEhE         CALLS   #3,G^LIB$TABLE_PARSE    ; Alpha version of LIB$TPARSE0         ON_ERR  ERR_EXIT     ;s;     ; First, check if we are zeroing-out the cpu or memory.o     ;          MOVQ    LOG_DEV_DESC,R6n8         CMPC3   R6,(R7),CPU_STG         ; Is it the cpu?         BNEQ    20$a         BRW     DO_CPU          20$:;         CMPC3   R6,(R7),MEM_STG         ; Is it the memory?h         BNEQ    40$e         BRW     DO_MEM     40$:               ;XH     ; Get the physical device name, and see if this device has an owner.     ;n         $GETDVI_S -mG                 DEVNAM=LOG_DEV_DESC,-   ; Command line has device name.i$                 ITMLST=DEV_ITEM_LIST         ON_ERR  ERR_EXIT     @         CMPL    #DC$_TERM,DEV_CLASS     ; Did we get a terminal?         BNEQ    50$.C         $GETDVI_S -                     ; Yes, make sure we get the ?                 DEVNAM=LOG_DEV_DESC,-   ; physical device name. #                 ITMLST=TT_ITEM_LISTr         ON_ERR  ERR_EXIT          50$:         TSTL    PIDD         BEQL    60$tH         TSTL    OVERRIDE                ; Device is allocated, see if we9         BGTR    60$                     ; should proceed.-H         MOVL    #SS$_DEVALLOC,R0        ; Device is allocated, exit with0         BRW     ERR_EXIT                ; error.     60$:          DO_DEV:tE         $CMKRNL_S -                     ; Enter k-mode to zero devicec error 8                 ROUTIN=ZERO_DEV,-               ; count.                 ARGLST=K_ARG         ON_ERR  ERR_EXIT         BRW     EXIT          DO_CPU: H         $CMKRNL_S -                     ; Enter k-mode to zero cpu error count.                 ROUTIN=ZERO_CPU1         ON_ERR  ERR_EXIT         BRW     EXIT          DO_MEM:iF                 $CMKRNL_S -                     ; Enter k-mode to zero memory error0                 ROUTIN=ZERO_MEM         ; count.         ON_ERR  ERR_EXIT         BRW     EXIT     
     ERR_EXIT: B         $EXIT_S R0                      ; Return the error, if any     	     EXIT:l9         RET                             ; Return normallym                         .SBTTL  ZERO_DEV Routine     ;++a     ;      ; Functional Description:e=     ;   This routine zeros out the error count in the UCB fori     ;   a specified device.r     ;j     ; Calling Sequence:..     ;   $CMKRNL_S ROUTIN=ZERO_DEV,ARGLST=K_ARG     ;0C     ;           where K_ARG is an argument list. This list containsuC     ;           the number of arguments passes (always 1), followeduC     ;           by the address of a descriptor pointing to the namec     ;           of a device.     ;v     ; Formal Parameters:(     ;   Descriptor for name of a device.     ;e     ; Implicit Inputs:     ;   I/O database.      ;:     ; Implicit Outputs:A*     ;   Device error count is set to zero.     ;o     ; Completion Status:     ;   Returned in R0.e     ;o     ; Side Effects:tG     ;   I/O database is locked (routine runs in kernel mode at elevatedu
     ;   IPL).      ;s     ;--+          "         .ENTRY  ZERO_DEV,^M<R2,R4>     B         MOVL    G^CTL$GL_PCB,R4 ;; Our PCB address is input to SCH4                                         ;; routines.A         JSB     G^SCH$IOLOCKW           ;; Lock the I/O database.l     B         MOVL    4(AP),R1                ;;; Address of device name descriptor.pF         JSB     G^IOC$SEARCHDEV         ;;; Get UCB address for device into R1.         ON_ERR  ZERO_DEV_EXITk     D         TSTL    UCB$L_PID(R1)           ;;; Make sure device did not become6         BEQL    80$                     ;;; allocated.G         TSTL    OVERRIDE                ;;; Device is allocated, see ifb we;         BGTR    80$                     ;;; should proceed.c          MOVL    #SS$_DEVALLOC,R0         BRW     ZERO_DEV_EXIT^          80$:>         MOVL    #SS$_NORMAL,R0          ; Set a success statusE         BBS     #DEV$V_MPDEV_MEMBER,UCB$L_DEVCHAR(R1),90$ ; Multipathc device ?F     ;## vax     CLRW    UCB$W_ERRCNT(R1)        ;;; Zero out the error count.E         CLRL    UCB$L_ERRCNT(R1)        ;;; Zero out the error count.  Alpha *         BRW     ZERO_DEV_EXIT           ; E     90$:                                        ; We have a multipath  device7         MOVL    UCB$PS_SUD(R1),R4       ; R4 is our SUD G         BEQL    ZERO_DEV_EXIT           ; Should never occur, but checks anywayH         MOVL    SUD$PS_MPDEV_PRIMARY_UCB(R4),R1 ; Start with the primary UCBaG         BEQL    ZERO_DEV_EXIT           ; Should never occur, but checkE anywayG         CLRL    UCB$L_ERRCNT(R1)        ; Zero the error count, primarys path.oB     100$:       MOVL    UCB$PS_SUD(R1),R4       ; Get the next SUD6         BEQL    ZERO_DEV_EXIT           ; Exit if zero?         MOVL    SUD$PS_MPDEV_NEXT_UCB(R4),R1 ; Get the next UCBi<         BEQL    ZERO_DEV_EXIT           ; Exit if zero, done@         CLRL    UCB$L_ERRCNT(R1)        ; Clear the error count, secondaries.B         BRB     100$                    ; Go check for next device          ZERO_DEV_EXIT:<         PUSHL   R0                      ;;; Remember status.E         JSB     G^SCH$IOUNLOCK          ;;; Unlock I/O database (drop- IPL).-         POPL    R0              RETe          E         .ENTRY  CLR_OV,^M<>             ;;; Move to the code PSECT ##t         CLRL    OVERRIDE         MOVL    #SS$_NORMAL,R0         RETR           c          +         .SBTTL  ZERO_CPU, ZERO_MEM Routinesn     ;++n     ;n     ;  Functional Description:E     ;   Routines to zero-out error-count location for cpu and memory.o     ;t     ;  Calling Sequence:!     ;   $CMKRNL_S ROUTIN=ZERO_CPUd!     ;   $CMKRNL_S ROUTIN=ZERO_MEMl     ;;     ;  Input Parameters:
     ;   None.;     ;        ;  Output Parameters:r
     ;   None.      ;r     ;  Implicit Inputs:i<     ;   Global sysmbols, EXE$GL_MCHKERRS for cpu errors, and)     ;   EXE$GL_MEMERRS for memory errors.s     ;r     ;  Implicit Outputs:
     ;   None.n     ;r     ;  Procedures called: 
     ;   None.      ;      ;  Completion Status: *     ;   Returned in R0, always successful.     ;      ;  Side Effects:
     ;   None.      ;      ;--r              .ENTRY  ZERO_CPU,^M<>(>         CLRL    EXE$GL_MCHKERRS         ; Zero out cpu errors.         MOVL    #SS$_NORMAL,R0         RET               .ENTRY  ZERO_MEM,^M<>jA         CLRL    EXE$GL_MEMERRS          ; Zero out memory errors._         MOVL    #SS$_NORMAL,R0         RET               .END    ZDEC   --  D Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------    Date: 03 Oct 2001 16:46:42 -05001 From: Graham Allan <allan@physics.umn.edu.nospam>.0 Subject: Re: Starting DECwindows on DEC 3000-600/ Message-ID: <w538zespfb1.fsf@crail.spa.umn.edu>   . Paul Repacholi <prep@prep.synonet.com> writes:  5 > Graham Allan <allan@physics.umn.edu.nospam> writes:r > F > > I just added an old DEC 3000m600 to our cluster, running VMS 7.2-1@ > > (and the most recent patch bundle). It's a standard ethernetC > > satellite type of configuration, and there are similar hardware0E > > nodes already present in the cluster. For some reason this system G > > won't start DECwindows at boot time, even though WINDOW_SYSTEM=1 inyG > > MODPARAMS.DAT, and all the SYSGEN parameters (which I know of) seemoD > > ok. The video card is the standard/basic PMAGB-B, and DECwindows? > > does start ok if I run sys$startup:decw$startup manually...s > B > Standard reason is one of the sysgen parameters too low... So itD > automunts, and screws royally into a little circle of... Irony is,@ > on Alpha it is a Dynamic parameter, so it can be fixed just by > tweeking the param.  > & > SERVER on the console has no effect.  D The main reason it's puzzling me is that it's an identical system toB another satellite, on which DECwindows *does* start automatically!@ MODPARAMS.DAT is identical between the two nodes (apart from the1 node-specific parts like SCSSYSTEMID, of course).m  I Now the really odd part is that this workstation used to live in another,rH totally separate cluster, in which it always showed the same behaviour -J the window system had to be started manually after boot. So it seems as ifJ this weird behaviour has come along with the hardware, but I can't imagine where it could be!   Graham -- iI -------------------------------------------------------------------------t Graham Allan9 School of Physics and Astronomy - University of MinnesotapI -------------------------------------------------------------------------c   ------------------------------  $ Date: Wed, 3 Oct 2001 20:27:06 +0200" From: "Hans Vlems" <hvlems@iae.nl>% Subject: Re: TCPIP 5.1 Telnet Problem,( Message-ID: <9pfl79$hfa$1@news.IAEhv.nl>   Is the default gateway set?c$ What does TCPIP SHOW ROUTE tell you?   Hans> Piyush Avichal <pa@it.singer-friedlander.com> wrote in message0 news:rKEu7.14180$ev2.22832@www.newsranger.com... > Hi,1 >a: > I have just upgraded a vax to OpenVMS 7.3 and TCPIP 5.1.F > I can ping, ftp and traceroute to an ip address 10.116.2.21. HoweverC > when I try to Telnet to that address, I get the following error :t >t$ > BART([SYSTEM])$ telnet 10.116.2.217 > %TELNET-E-IVHOST, Invalid or unknown host 10.116.2.21t6 > -TCPIP-E-NO_RECOVERY, unexpected name server failure > = > I can only telnet to hosts that are in my local host table.PF > I have not enabled any bind services, nor do I want to, but it seems2 > to be trying to resolve the address and failing. > 3 > Does anyone have any ideas why I am getting this?/ >e > Many Thanks, > 	 > Piyush.  >  >l   ------------------------------  % Date: Wed, 03 Oct 2001 22:19:45 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> # Subject: Re: VAX disaster tolerancee' Message-ID: <3BBBD551.2C168521@fsi.net>a   Fabio Cardoso wrote: > 
 > A VAX... > , > Why Compaq dont give the VAX license to be3 > built by another vendor ? If the VAXes are reallyh. > needed in the market. Example: Mentec should& > build a 1U VAX for rackmount. [snip]  C Well, a VAXstation 4000/VLC is tad bigger than 1U, and needs shelf,  but...   --   David J. Dachtera  dba DJE Systemsa http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/t   ------------------------------   Date: 3 Oct 2001 10:51:45 -0700h, From: don.braffitt@compaq.com (Don Braffitt)1 Subject: Re: VAX-Alpha Migration: COBOL/BASIC/FMSa= Message-ID: <14c5ce2f.0110030951.3fbb02d1@posting.google.com>   ! For COBOL, consider these things.   D Compile with /CHECK on Alpha.  This will help catch at run-time someC of the potential problem MOVE situations that Jim Butler mentioned.k   Compile /WARN=ALL on Alpha.   @ Look at /MATH=CIT3 on Alpha.  If /WARN=ALL turns up any cases ofD possible intermediate truncation, /MATH=CIT3 is designed to give youD arithmetic results very to Compaq COBOL on OpenVMS VAX as contrasted( with /MATH=FLOAT (the default on Alpha).   - Don Braffitt&   Compaq COBOL and SORT project leader   ------------------------------   Date: 3 Oct 2001 21:55:36 -0500 + From: young_r@encompasserve.org (Rob Young)v! Subject: Re: VMS 7.3 versus 7.2-2 3 Message-ID: <zbzwSl1sFMXk@eisner.encompasserve.org>   P In article <3BBA2EEC.8B86A75F@cha.ab.ca>, Lee Y T Mah <lytmah@cha.ab.ca> writes:H > I am looking at setting up a cluster with the following configuration:H >   Data Center 1: Two ES40's and a SAN with two HSG80's supporting 47 x > 36GB drives.H >   Data Center 2: Two ES40's and a SAN with two HSG80's supporting 47 x > 36GB drives.G >   A quorum node, possibly located in either Data centre, or remote tob > either data centre.mD > This cluster will gradually be brought into an existing production > cluster of four AS1200's atRH > VMS 7.1-2, using HSJ40/50's and on CI/FDDI/Ethernet, with the eventual > aim to drop the AS1200's.EE > Can anyone with similar experience provide recommendations as which 3 > version to use for the new cluster, 7.3 or 7.2-2?  >    Lee,  B 	I am working a similar config for a potential situation.  At thisA 	time, I recommend 7.2-2.  I do know that 7.3 is in production at A 	some large customer sites.  However, the withdrawal of the first// 	XFC patch makes me more than a little nervous:l  Q http://ftp1.support.compaq.com/patches/public/vms/axp/v7.3/vms73_xfc-v0100.README   > 	When that shows back up and stays there for good, I would say@ 	7.3 is good to go.  Let me also add, I think that 7.3 is a veryC 	good release and VMS engineering keeps raising the bar on quality. @ 	That said, XFC is "in the critical path" and that is the reason> 	for heightened uneasiness and a quick Deja search will reveal" 	some folks have had "XFC" issues.  . http://groups.google.com/advanced_group_search  * 	An advanced search for "xfc rms" reveals:  V http://groups.google.com/groups?selm=3B6F86E7.BDBAC4E3%40wasd.vsm.com.au&output=gplain    B "Beware the Ides of XFC" lest your RMS integrity die an unpleasant death.  F We recently upgraded to 7.3 on two Alpha clusters (7 and 4 systems) asF well as a miscellany of stand-alone systems.  This was in part to reapD the benfits of XFC (which looked promising on the two standalones we3 experimented on - obviously not thoroughly enough).L  C From the outset we noticed (and reported) %XFCACHE-W-DATALOSS OPCOMDD messages.  This (seems) to have resulted in at least one applicationG ACCVIOing.  More recently we noticed RMS index file corruption (ANA/RMSFF reporting "Unrecoverable error encountered in structure of file") withA XFC (VIOC) turned on and none with it turned off (same data, same A application, 2 cycles of investigation).  Without reproducing the D contents of some two dozen or so emails back and forth with CSC it's6 difficult to be more specific with these descriptions.   	etc.    				RobI   ------------------------------   Date: 3 Oct 2001 22:25:05 -0700 1 From: keithparris_NOSPAM@yahoo.com (Keith Parris)h! Subject: Re: VMS 7.3 versus 7.2-2R= Message-ID: <cf15391e.0110032125.165a658f@posting.google.com>E  U Lee Y T Mah <lytmah@cha.ab.ca> wrote in message news:<3BBA2EEC.8B86A75F@cha.ab.ca>... H > I am looking at setting up a cluster with the following configuration:H >   Data Center 1: Two ES40's and a SAN with two HSG80's supporting 47 x > 36GB drives.H >   Data Center 2: Two ES40's and a SAN with two HSG80's supporting 47 x > 36GB drives.  9 Are you already running this cluster with multiple sites?I  < Do you plan to run Fibre Channel between the two sites?  (MyE recommendation at present is that until direct / MSCP-served failover_B is supported, as planned for 7.3-1, that you NOT connect the FibreF Channel fabric between sites, or at the very least not unless you haveB at two separate (redundant) fabrics between the sites.  The manualD shadowset failover actions you have to take today in the event of an> inter-site fabric failure are just too hard and error-prone in practice, I feel.)  G >   A quorum node, possibly located in either Data centre, or remote to  > either data centre.X  B A quorum node in either data centre is not really a quorum node atF all; in that case, you really have an unbalanced-votes configuration. D A node with a tie-breaking vote at a 3rd site that is independent ofE either main site would be a real quorum node.  You might want to takee@ a look at the sections on Multi-site Clusters, Disaster-TolerantD Clusters, and vote configurations for multi-site clusters in my 'VMS- Concepts: Advanced Concepts' seminar notes at W http://www.geocities.com/keithparris/decus_presentations/f2001_vmsclusters_advanced.ppti  D > This cluster will gradually be brought into an existing production > cluster of four AS1200's at H > VMS 7.1-2, using HSJ40/50's and on CI/FDDI/Ethernet, with the eventual > aim to drop the AS1200's.$  F "Gradually" is the key word here for success.  Plan on lots of testing? of induced failure scenarios, and go slowly and cautiously with @ changes.  CI, FDDI, Ethernet, and HSJs are fairly stable, provenA technology; the Fibre Channel stuff is lots faster, but much less   mature and stable at this point.  E > Can anyone with similar experience provide recommendations as which 3 > version to use for the new cluster, 7.3 or 7.2-2?T  E If your aim is very high availability, as might be inferred from your F multi-site questions, today I would use neither.  7.3 has been out forF a little while, and has some patch kits available for it, but it has aA whole lot of brand-new functionality, and new features add to the B risk.  7.2-2 has just barely hit the streets, so I feel it has too> little real-world exposure yet.  Also, unlike 7.2-1, which wasD basically 7.2 with a roll-up of patches, 7.2-2 also includes some of> the new features from 7.3 in addition to a roll-up of patches.  E For very-high-availability clusters, I like to wait at least 9 months E after a release for the major problems to be isolated (and patch kitsQ? to become available and proven) before using any new release ineF production.  I also like to wait for at least the first "dash" releaseD after a new major release (for example, waiting for 7.3-1 instead ofC using 7.3 itself).  So today I'd probably recommend 7.2-1 (or maybe  7.2-1H1) plus patches.  D If you were planning to start standalone testing on the new hardwareA now, with integration with the main cluster not planned until 6-9 B months from now, then using either 7.3 or 7.2-2 for the standalone? cluster would be fine, even desirable; and then consider moving B carefully from 7.1-2 to 7.2-1[H1] on the production cluster in the interim.C -------------------------------------------------------------------GC Keith Parris | parris at encompasserve dot org | VMS consulting on: C Clusters, Disaster Tolerance, Internals, Performance, Storage & I/Or   ------------------------------   Date: 03 Oct 2001 20:11:54 GMT) From: wkb@freebie.xs4all.nl (Wilko Bulte)i8 Subject: Re: Warranty on ES45 reduced from 3 years to 1?; Message-ID: <3bbb710a$0$19387$e4fe514c@newszilla.xs4all.nl>   E In <VA.0000046c.32dd9237@sture.ch> Paul Sture <paul@sture.ch> writes:   O >In article <3bb8b284$0$19390$e4fe514c@newszilla.xs4all.nl>, Wilko Bulte wrote: n >> In <69Kt7.14283$xG6.5530020@typhoon.ne.mediaone.net> "Terry C. Shannon" <terryshannon@mediaone.net> writes: >> >[snip]e >>  I >> >Finally came across some Sun Blades. They were in an Internet cafe int( >> >Amsterdam, down by Leidseplein (sp). >> c >> Spelled just fine Terry ;-) >>    >But how do we pronounce it? :-)   Leydsuplain    [roughly..]0   Wilko  --- |   / o / /_  _   		email: 	wilko@FreeBSD.orgc1 |/|/ / / /(  (_)  Bulte		Arnhem, The Netherlands	a   ------------------------------   Date: 03 Oct 2001 20:12:31 GMT) From: wkb@freebie.xs4all.nl (Wilko Bulte) 8 Subject: Re: Warranty on ES45 reduced from 3 years to 1?; Message-ID: <3bbb712d$0$19386$e4fe514c@newszilla.xs4all.nl>t  k In <Tzvu7.68433$vq.12907140@typhoon.ne.mediaone.net> "Terry C. Shannon" <terryshannon@mediaone.net> writes:o    5 >"Carl Perkins" <carl@gerg.tamu.edu> wrote in messageo' >news:2OCT200120185834@gerg.tamu.edu...a >> paul@sture.ch writes...K >> }In article <3bb8b284$0$19390$e4fe514c@newszilla.xs4all.nl>, Wilko Bulte  >wrote:YM >> }> In <69Kt7.14283$xG6.5530020@typhoon.ne.mediaone.net> "Terry C. Shannon"i$ ><terryshannon@mediaone.net> writes: >> }> 
 >> }[snip] >> }>eL >> }> >Finally came across some Sun Blades. They were in an Internet cafe in+ >> }> >Amsterdam, down by Leidseplein (sp).d >> }>f! >> }> Spelled just fine Terry ;-)  >> }> # >> }But how do we pronounce it? :-)d >> }___  >> }Paul Sture >> }Switzerland  >>
 >> Poorly? >> >> ("Led Zeppelin"?)  M >Guess it depends on whether or not you've dallied in a coffee shop or "smart * >shop" before visiting the Leidseplein....  ' You see? Here speaks a man in the know l   ;-) ;-)e   Wilko  --- |   / o / /_  _   		email: 	wilko@FreeBSD.orgp1 |/|/ / / /(  (_)  Bulte		Arnhem, The Netherlands	    ------------------------------   Date: 3 Oct 2001 19:53:04 +02000* From: eplan@kapsch.net (Peter LANGSTOEGER)7 Subject: Re: [AMDS V7.3A] Problem on OpenVMS VAX V7.3 ?t* Message-ID: <3bbb5080$1@news.kapsch.co.at>  W In article <3b98e2a9$1@news.kapsch.co.at>, eplan@kapsch.net (Peter LANGSTOEGER) writes:;
 >Am I alone ?e  5 Now that I haven't got any response, may I ask again. A Maybe the question got lost in the upteen hundreds posting of the 7 'the-end-is-near' threads (Alphacide/HighPrice/WTC/...)    -- i< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888P< <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------   Date: 3 Oct 2001 19:40:14 +0200 * From: eplan@kapsch.net (Peter LANGSTOEGER)) Subject: Re: [LYRIS] OpenVMS Patches ListL* Message-ID: <3bbb4d7e$1@news.kapsch.co.at>  W In article <3bb3abc4$1@news.kapsch.co.at>, eplan@kapsch.net (Peter LANGSTOEGER) writes:EX >In article <3bb3ab21$1@news.kapsch.co.at>, eplan@kapsch.net (Peter LANGSTOEGER) writes:E >>At earlier times I got emails of a mailing-list (in fact 2 of them:cE >>openvms-patches and security) after adding myself to this lists via  >>: >>	http://www.support.compaq.com/patches/mailing-list.html >>H >>Appearantly these lists are silent for quite some time now (at least IF >>didn't get anything) and the above link is also no longer reachable. >e	 >It's nowe >E: >	http://www.support.compaq.com/patches/mailing-list.shtml >0 >One problem solved ;-)    And the other one, too.T1 I've just received a mail from the security list.   I And the VMS list was really (almost) silent during Sep (I was on vacationtJ the whole Aug) and so I got this feeling. Maybe I should collect the mails% just to check again before posting...O   So, many thanks to all. & Seems the problem was only my mind ;-)   --  < Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888c< <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------   End of INFO-VAX 2001.551 ************************