1 INFO-VAX	Mon, 11 Jul 2005	Volume 2005 : Issue 383       Contents:0 Break into the Computer Industry - Get Hired Now Re: DECnet over DSSI Re: DECnet over DSSI Re: DECnet over DSSI Re: DECnet over DSSI9 Re: interchangeability of memory between various machines 9 Re: interchangeability of memory between various machines 9 Re: interchangeability of memory between various machines / Re: Less than 1 Gorham of VMS systems out there ) Re: Looking for MRU (Media Robot utility) 2 Re: Need DCPS Software for hobbyist (BHM, AL Area)D Re: Now the UK Has it's own date in the history of al quiada attacksP Re: Now the UK Has it's own date in the history of al quiada attacks attacksatta) Re: Performance issue after 7.3-2 upgrade  Re: Upgrading VMS  Re: Upgrading VMS   F ----------------------------------------------------------------------    Date: 10 Jul 2005 20:05:47 -0700! From: findjobeasydotcom@gmail.com 9 Subject: Break into the Computer Industry - Get Hired Now B Message-ID: <1121051147.134755.89340@o13g2000cwo.googlegroups.com>  F Tired of being unemployed? FindJobEasy.com is the solution to all yourD employment worries. With on quick and easy site, you can access jobs= from more than 50 major job listing sites and employer sites. F FindJobEasy.com is a convienent and simple way to search a job withoutE having to go through each site individually. So, what are you waiting 8 for? Go to FindJobEasy.com today and start working soon.   ------------------------------  % Date: Sun, 10 Jul 2005 23:24:27 +0800  From: prep@prep.synonet.com  Subject: Re: DECnet over DSSI - Message-ID: <874qb2d6qc.fsf@prep.synonet.com>   ( "H Vlems" <nospam@what.ever.com> writes:  K > Years ago it was possible to run DECnet over the CI bus. IIRC it involved + > loading the CYDRIVER to make that happen. 4 > The question is: can DECnet run over the DSSI bus? > If so, with what driver?  H CYDRIVER. CI and DSSI both use compatable port drivers. There is a shortI `how to' in the FM I seem to remember. I run it between my home 4000-700s  for some time.       --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Sun, 10 Jul 2005 23:39:30 +0800  From: prep@prep.synonet.com  Subject: Re: DECnet over DSSI - Message-ID: <87vf3ibrgt.fsf@prep.synonet.com>    How soon they forget...   D MOP, Maintainance and Operation Protocol, is a low level set of comsA protocols that work between *adjacent* nodes to do functions like % booting, dumping, and remote console.   ? DUP, Diagnostic and Utilities Protocol, is part of the MSCP/SCS ? protocol suite that is use for comunicating with controlers for A diagnostic and general utility function as opposed to storage and @ shovelling the bytes. An SCS port implements 3 type of transfer,E Datagram, Sequenced Messages, and Block Data. The DECnet classdriver, > for reason known only to the person who wrote it, if him, uses@ datagrams. Small messages, no HW assist for sequencing and accs, ie  the worst of the 3...   C Mind you, a high cost CI line is IMO almost essential in a critical 2 cluster with functions distributed over the nodes.   --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  # Date: Sun, 10 Jul 2005 21:02:22 GMT # From: Beach Runner <bob@nospam.com>  Subject: Re: DECnet over DSSI = Message-ID: <y1gAe.228037$w15.180272@tornado.tampabay.rr.com>    prep@prep.synonet.com wrote:  * > "H Vlems" <nospam@what.ever.com> writes: >  > K >>Years ago it was possible to run DECnet over the CI bus. IIRC it involved + >>loading the CYDRIVER to make that happen. 4 >>The question is: can DECnet run over the DSSI bus? >>If so, with what driver? >  > J > CYDRIVER. CI and DSSI both use compatable port drivers. There is a shortK > `how to' in the FM I seem to remember. I run it between my home 4000-700s  > for some time.      C Yes, DSSI is still supported, and you can mix, CI, DSSI, ethernet,  4 memory channel, as long as they call see each other.  E Since DSSI is so close to SCSI you wonder why SCS didn't run on SCSI  # clusters.  Maybe FDDI will be next.    The future is 10 gb ethernet.    >  >    ------------------------------  % Date: Sun, 10 Jul 2005 20:37:31 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net> Subject: Re: DECnet over DSSI + Message-ID: <42D1CD5B.FE9082EF@comcast.net>    H Vlems wrote: > K > Years ago it was possible to run DECnet over the CI bus. IIRC it involved + > loading the CYDRIVER to make that happen. 4 > The question is: can DECnet run over the DSSI bus? > If so, with what driver?  C I don't recall ever reading about a driver that would allow passing  DECnet over the DSSI bus.   @ Even on CI, it's meant as a high-cost fall-back only in case theB ethernet proved unusable, for whatever reason. Early CI had enoughG congestion problems without adding DECnet into the mix. Even 140Mbit CI A still is less than ideal as a DECnet medium shared by the storage  interconnect.   D The ethernet LAN is almost always preferable, but I can see where itD might be helpful in special cases to pass DECnet over the CI from an? end-node to a Level-I routing node, and thence to the ethernet.    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Sun, 10 Jul 2005 23:56:00 +0200 0 From: Keith Cayemberg <keith.cayemberg@arcor.de>B Subject: Re: interchangeability of memory between various machinesB Message-ID: <42d19972$0$20129$9b4e6d93@newsread2.arcor-online.net>   Main, Kerry wrote: >  >>-----Original Message-----1 >>From: Phillip Helbig---remove CLOTHES to reply  + >>[mailto:helbig@astro.multiCLOTHESvax.de]   >>Sent: July 10, 2005 11:24 AM >>To: Info-VAX@Mvb.Saic.Com @ >>Subject: interchangeability of memory between various machines >>= >>I'm doing an inventory of all my hardware, making notes of   >>what doesn't> >>work (graphics cards and power supplies are the main things 
 >>which don't @ >>work) and consolidating things.  If I have two machines which  >>don't workA >>perfectly, I might combine them into one, for example.  So, I'm G >>wondering if memory is exchangeable between any of the machines below 9 >>(of course, I don't expect anything to be exchangeable   >>between VAX and 	 >>ALPHA):  >> >>VAXstation   3100/30 >>VAXstation   3100/38 >>VAXstation   3100/48 >>VAXstation   4000/60 >>VAXstation   4000/90 >>VAXstation   4000/90A  >>VAX          4000/100A >>VAX          4000/105A >>   >>DEC          3000/300LX (EV4)  >>DEC          3000/600   (EV4)  >>ALPHAserver  2000       (EV4)   >>ALPHAserver  2100       (EV45)  >>ALPHAstation  255/233   (EV45)  >>ALPHAserver  1200       (EV56) >>E >>When it comes to actually swapping SIMMs, I'll dig out the hardware D >>documentation.  However, I don't know where it is documented which8 >>memory actually works in more than 1 type of machine.  >>H >>In particular, can I put 4MB SIMMs for a VAXstation 4000/60 (I presumeE >>that's what it is; part number is 54-19145-AU or 54-19103) in the 4 D >>empty slots of a VAXstation 4000/90 which has 4 16-MB SIMMs at the? >>moment?  Can I put 16-MB SIMMs from a VAXstation 4000/90 (or  
 >>90A) into a E >>VAX 4000/105A?  Can I swap SIMMs among the various VAXstation 3100  G >>models listed above?  Can I swap SIMMs among the various EV4 systems   >>listed above?  >> >> >  >  > Phillip  - > 7 > Some pointers to assist you: (url's will likely wrap)  > J > http://h18002.www1.hp.com/alphaserver/options/mvx/mvx_1_options.html#Opt
 > ListHead > J > http://h18002.www1.hp.com/alphaserver/options/cat/cat_1_options.html#Opt
 > ListHead > J > http://h18002.www1.hp.com/alphaserver/options/vax4/vax4_1_options.html#O > ptListHead > J > http://h18002.www1.hp.com/alphaserver/options/vaxs/vaxs_1_options.html#O > ptListHead > C > Note - for future reference: Retired VAX systems info:(follow the ! > options info at bottom of page) , > http://h18002.www1.hp.com/alphaserver/vax/ > 	 > Regards  >  > Kerry Main > Senior Consultant  > HP Services Canada > Voice: 613-592-4660  > Fax: 613-591-4477  > kerryDOTmainAThpDOTcom > (remove the DOT's and AT)  > 6 > OpenVMS - the secure, multi-site OS that just works.    G Here are some more documents on 3rd-Party Memory Expansion for various   VAX and Alpha Systems...    6 Kingston Technology - Memory Installation Guides - DEC6 http://www.kingston.com/support/install/memory/DEC.asp  9 Kingston Technology - Memory Installation Guides - Compaq 9 http://www.kingston.com/support/install/memory/COMPAQ.asp   9 Nemonix Engineering, Inc. - Engineering Product Manuals -  Wayback Machine H http://web.archive.org/web/20031206090905/www.nemonixinc.com/manuals.htm  : Nemonix Engineering, Inc. - VAX Memory Expansion Options - Wayback Machine G http://web.archive.org/web/20040203210728/www.nemonixinc.com/expmem.htm    Cheers!    Keith Cayemberg    ------------------------------  # Date: Sun, 10 Jul 2005 22:29:35 GMT 5 From: rdeininger@mindspringdot.com (Robert Deininger) B Subject: Re: interchangeability of memory between various machinesL Message-ID: <rdeininger-1007051829370001@user-105n8em.dialup.mindspring.com>  D In article <dareit$jh4$1@online.de>, helbig@astro.multiCLOTHESvax.de1 (Phillip Helbig---remove CLOTHES to reply) wrote:   H >I'm doing an inventory of all my hardware, making notes of what doesn'tH >work (graphics cards and power supplies are the main things which don'tI >work) and consolidating things.  If I have two machines which don't work @ >perfectly, I might combine them into one, for example.  So, I'mF >wondering if memory is exchangeable between any of the machines belowH >(of course, I don't expect anything to be exchangeable between VAX and  >ALPHA): >  >VAXstation   3100/30  >VAXstation   3100/38  >VAXstation   3100/48  >VAXstation   4000/60  >VAXstation   4000/90  >VAXstation   4000/90A >VAX          4000/100A  >VAX          4000/105A  >  >DEC          3000/300LX (EV4) >DEC          3000/600   (EV4) >ALPHAserver  2000       (EV4) >ALPHAserver  2100       (EV45)  >ALPHAstation  255/233   (EV45)  >ALPHAserver  1200       (EV56)   6 The DEC 3000/300 family uses different memory from DEC$ 3000/400/500/600/700/800/900 family.   ------------------------------  % Date: Sun, 10 Jul 2005 18:46:40 -0500 ( From: Rich Jordan <duodec@speakeasy.net>B Subject: Re: interchangeability of memory between various machines2 Message-ID: <Ac-dne-uy-qcO0zfRVn-pg@speakeasy.net>  / Phillip Helbig---remove CLOTHES to reply wrote: I > I'm doing an inventory of all my hardware, making notes of what doesn't I > work (graphics cards and power supplies are the main things which don't J > work) and consolidating things.  If I have two machines which don't workA > perfectly, I might combine them into one, for example.  So, I'm G > wondering if memory is exchangeable between any of the machines below I > (of course, I don't expect anything to be exchangeable between VAX and  	 > ALPHA):  >  > VAXstation   3100/30 > VAXstation   3100/38 > VAXstation   3100/48  G VAXstation 3100/30, /38, /40, /48 use the same memory cards, as do the  I MicroVAX 3100-10, 10e, 20, and 20e.  There used to be 4, 8, 12, and 16MB  H cards from DEC (perhaps other sizes from other vendors); the 8 and 16MB F cards had dual connectors that allowed a second single port (4, 12MB) H card to be installed, or a single port card could be installed directly D to the main logic board.  32MB maximum with 4MB embedded, 16MB dual % port, and 12MB single port installed.   = My 3100-30 at work has a 16 and a 12 for 32MB maximum memory.    Rich   ------------------------------  % Date: Sun, 10 Jul 2005 15:05:09 -0400 ( From: Bill Todd <billtodd@metrocast.net>8 Subject: Re: Less than 1 Gorham of VMS systems out there= Message-ID: <XcadnQ9b0_n77EzfRVn-iQ@metrocastcablevision.com>    Bill Gunshannon wrote:   ...   -   I aslo thought that Alpha was still holding ( > it's own against the other processors.  E The current top-of-the-line Alpha offers only about half the SPECint  H performance of the fastest x86s and trails the faster contemporary RISC A processors there significantly as well, but is closer to x86 and  H contemporary RISCs in SPECfp and (POWER5 excepted) in some large-system C commercial benchmarks (though, curiously, HP has been reluctant to  I submit any of the latter - the early-2003 SAP SD submission being a rare  / exception, since Itanic sucked so badly at it).    - bill   ------------------------------  % Date: Sun, 10 Jul 2005 20:19:17 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>2 Subject: Re: Looking for MRU (Media Robot utility)+ Message-ID: <42D1C915.796E3802@comcast.net>    Jeff Cameron wrote:  > G > On 7/7/05 6:38 PM, in article 42CDD921.B5D3B30F@comcast.net, "David J . > Dachtera" <djesys.nospam@comcast.net> wrote: >  > > Jeff Cameron wrote:  > >>Q > >> On 7/6/05 9:23 AM, in article aATye.145$rO4.813@tor-nn1.netcom.ca, "Syltrem" & > >> <syltremzulu@videotron.ca> wrote: > >>& > >>> My colleague cannot find his CD.E > >>> It seems not to be on the layered software library either (CDs)  > >>> ) > >>> Is it available somewhere you know?  > >>> This is for an Alpha > >>> $ robot show version > >>> Media Robot Utility V1.4 > >>>  > >>> Thanks3 > >> Better than MRU is MTI's RLM, free evaluation. 9 > >> http://support.mti.com/MTICare/OpenVMS/RLM/index.htm . > >> Works in Clustered and networked systems. > >> Command line and API. > >  > > Does it support STK L700e? > G > Very good chance. If the library and it's drives are supported by VMS    Not sure what you mean there.   E The Quantum-badged SDLT320 drives work with VMS, to a point. They log C many extraneous errors and fialures such as parity errors do occur. F Restarting usingthe same tape does notr produce the same result; so, IG suspect that the drive may not be robust enough to do what VMS expects.   E As to the library, it would be just another "GK" device, at best. I'm A still looking for software for OpenVMS that claims to support STK ) L700e's robot (hence, the original post).   D At a former site, we bought an STK 9710 through MTI and used it withF Oasis/RLM and TapeControl. So, I was hoping that MTI still had a closeD enough relationship with STK that the engineering exchange was still	 going on.    >than I G > have not seen a library yet that did not work with RLM. Although some 5 > libraries the pass in/out slot may or may not work.   & Too bad - that's a "must have" for us.   > Go to the Web site andI > click on request for a 60 day evaluation license. Free 60 day telephone 
 > support. > K > 50% discount on commercial license for any comp.os.vms participants. Same J > offer goes for MTI's TapeControl, TapeConvert, Archive, and AutoStor HSMF > software licenses. One year annual maintenance required with initial@ > purchase. Send email to VMSsupport@mti.com for details or RFQ. > M > I'm trying to work out a deal with MONTGAR to provide hobbyist licenses for 0 > RLM and TapeControl media management software.  B Super! ...though I suspect the hardware that would require (a tapeG library) might be a bit pricey for hobbyists. The OpenVMS Faithful tend E to resemble church mice more than executive-types (rolling in dough).    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Sun, 10 Jul 2005 22:48:29 -0500 ! From: "ssj152" <noone@nospam.net> ; Subject: Re: Need DCPS Software for hobbyist (BHM, AL Area) + Message-ID: <_%lAe.18752$Si3.8420@fe06.lga>   K "Christopher Story" <ke6rwj@spam-eater-remove-me-msn.com> wrote in message  & news:1Nvze.31882$B_3.13322@fe05.lga...+ >> You need to specify if for Alpha or VAX.  >  > Alpha  >  >    Christopher,  L Have you located the software yet? If not, what version are you looking for?K I am pretty sure that I have it from the 7.3 distribution for both VAX and   ALPHA.   I live in the BHM, AL area.   H Reply on list and if you need what I have, we can get together off list.   ssj152 (stuart)    ------------------------------  % Date: Sun, 10 Jul 2005 14:50:09 -0400 ( From: Bill Todd <billtodd@metrocast.net>M Subject: Re: Now the UK Has it's own date in the history of al quiada attacks = Message-ID: <SsWdnbzdFN5-8EzfRVn-tA@metrocastcablevision.com>    Bill Gunshannon wrote:? > In article <FoSdnTToh5Ts503fRVn-tg@metrocastcablevision.com>, - > 	Bill Todd <billtodd@metrocast.net> writes:  > H >>Most of the rest of the world has taken a noticeably more responsible A >>attitude toward automotive energy efficiency for many decades,   >  > 	 > Really?   G Of course:  I'm almost always quite careful to state facts rather than  C garbage opinions, and when I state opinions I qualify them as such.   A    VW makes an SUV. Mercedes-Benz makes an SUV, BMW makes an SUV. J > Honda makes an SUV.  Toyota makes and SUV.  Need I go on?  And of courseG > your prime example, The Hummer is made by Daimler-Chrysler.  Not sure I > which side of the ocean owns the bigger part at this time, but it isn't  > strictly American.  D I see that you're as much of a moron as your previous comment about I hypocrisy implied (though since I wasn't positive which way you meant it  / at the time I refrained from responding to it).   I It's not about who *makes* SUVs, it's about who *buys* them.  Obviously,  G if there's a demand, someone will address that demand - and, as usual,  G superior foreign competition seems to be taking increasing portions of   that market away from the U.S.   ...   D                                                               taxingD >>gasoline (and unnecessarily gas-guzzling vehicles) at a rate that E >>encourages use of more energy-efficient products and/or collective  K >>transportation (car pools, mass transit, etc.) where it's feasible to do   >>so.    >  > 3 > No, taxation raises money for the rulling regime.   I Another idiotic observation.  The fact that taxation may raise money has  H absolutely nothing to do with the fact that it also influences behavior F - save for the fact that the more effectively it influences behavior,  the less money is raised.   F Take a look at the overall complexion of the automobile population in F countries with significantly higher gas prices than those in the U.S. H and the relationship may become clear to you (save perhaps in countries G so poor that driving is a sufficiently rare luxury that gasoline costs  @ pale in comparison to the need to take whatever vehicles may be " available at an affordable price).     That is it's only H > real purpose.  There have been numerous stories recently on the GermanI > radio station I listen to aboutt he high (and rising) cost of gasoline. G > Let's see, they are closer to the Middle East so transportation costs F > for the raw oil would be less than for the US.  Their entire countryH > is smaller than most of our states so the transportation costs for theF > finished product would be less.  So if gas cost more there than here$ > the only possible reason is taxes.  % No shit, Sherlock:  that's the point.   #    And guess what, I am not hearing E > people thanking the government the "encouragement" it is providing.   I The Great Unwashed need not appreciate the necessity of curbing usage of  A fossil fuels to be coerced into doing so - it would just be more   desirable if they did.   >  > I >>    For a while, after the 'gas crises' of the '70s (which pale beside  J >>what we're going to be in for fairly soon), we seemed to be headed in a C >>similarly more responsible direction, but Ronnie Ray-Gun and his  + >>ideological successors took care of that.  >  > I > No, what "took care of that" was that people found out the "gas crises" I > was contrived to find out just how high prices could go without a major 0 > revolt worldwide.  It was all about the money.  H Y'know, people like you who think they have some 'inalienable right' to I rape the planet at whatever cost to future generations (including by now  H the very next one, of which my daughter is a member) disgust me a great G deal more than any 'terrorists' trying to get us the hell out of their   part of the world.  D One of democracy's great principles is that people will rise to the F challenge of being responsible citizens.  When they instead decide to H stand obstinately on their 'inalienable right' to prejudice, ignorance, @ and short-sighted interests, democracy is not likely to survive.   ...   "    Kyoto was designed to take awayF > some of the competitive advantage the US has built up over the years+ > by being better than the other countries.   H It seems that you like a dash of paranoia along with your ignorance and G jingoism.  Ever heard of the concept "from those to whom much has been  F given, much is expected", or do you simply not subscribe to it (being A that as one of the recipients of said good fortune it might seem   inconvenient)?   ...   	    If all I > those other countries are so much better than we are, why are you still  > here?   I I've answered that question before:  this is *my* country, and I want it  E back - by 'whatever means necessary', to use one of Dubya's favorite   phrases.   - bill   ------------------------------  % Date: Sun, 10 Jul 2005 23:29:35 +0800  From: prep@prep.synonet.com Y Subject: Re: Now the UK Has it's own date in the history of al quiada attacks attacksatta - Message-ID: <87zmsubrxc.fsf@prep.synonet.com>   / William Webb <william.w.webb@gmail.com> writes:   F > It takes a remarkable amount of logical gymnastics to posit that theA > liberation of 50 million souls from despotic regimes which were B > brutally oppressive to women and ethnic minorities was the wrong > thing to do.  H And what regime would that be? And pray tell use when women got the vote in Iraq as well.     --  < Paul Repacholi                               1 Crescent Rd.,7 +61 (08) 9257-1001                           Kalamunda. @                                              West Australia 6076* comp.os.vms,- The Older, Grumpier Slashdot. Raw, Cooked or Well-done, it's all half baked.F EPIC, The Architecture of the future, always has been, always will be.   ------------------------------  % Date: Sun, 10 Jul 2005 20:24:19 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net>2 Subject: Re: Performance issue after 7.3-2 upgrade+ Message-ID: <42D1CA43.1724FDD0@comcast.net>    Keith Parris wrote:  >  > Larry Fahnoe wrote: @ > > we observed nodes slow down significantly for 10-15 minutes. > I > 8.2 (and 7.3-2 with FIBRE_SCSI V0400 and above) does back off much more G > aggressively than 7.3-1 (and vanilla 7.3-2) in the face of Queue Full B > events. Brian Allison's Bootcamp presentation i222 covered this: > H > Prior to 7.3-1, VMS had a variable maximum I/O queue depth per LUN, inE > the range of 3 to 16, based on recent I/O sizes. This caused severe I > performance problems, particularly on RAID LUNs (which can have lots of I > independent disk-head actuators, and we really want to keep all of them G > busy for best throughput) and for large I/O sizes (which could reduce I > the allowable queue depth down to 3), because of the very-small maximum H > queue depths allowed. (Aside: On MSCP controllers, I've observed queueG > depths in the 100s, even 1000s, without problems. But the marketplace 5 > said Proprietary is Bad, Industry Standard is Good.   > Which standards body(-ies) publish(es) the standard(s) you are referencing here?   A ...or did you mean "de facto standard", not an accepted/published  industry standard?   > And we're doing our G > best to provide all the advantages of the old Proprietary soluions on & > the less-expensive Industry Standard   Same question ...?   > foundation, by applying our + > engineering skills on the software side.)  > [snip]   --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------  % Date: Sun, 10 Jul 2005 14:17:28 -0400 3 From: "Richard B. Gilbert" <rgilbert88@comcast.net>  Subject: Re: Upgrading VMS0 Message-ID: <2omdneOILbKk-0zfRVn-sA@comcast.com>   Main, Kerry wrote:   >>-----Original Message-----; >>From: Richard B. Gilbert [mailto:rgilbert88@comcast.net]   >>Sent: July 10, 2005 1:50 AM  >>To: Info-VAX@Mvb.Saic.Com  >>Subject: Re: Upgrading VMS >> >>David D Miller wrote:  >> >>     >>	 >>>G'day:  >>> < >>>How do you make a decision to upgrade?  Is your decision 	 >>>        >>>  >>ruled by Policy? >>     >>> >>>By convenience (nothing else to do today)?  Or do you have 	 >>>        >>>  >>an objective >>     >> >>>criteria that you apply?  >>> @ >>>Specifically, is 7.3 good enough (in some sense) or is 7.3-2 	 >>>        >>>  >>a better bet?  >>     >>> >>>I suppose the same question could be applied to patches -- 	 >>>        >>>  >>but then again,  >>     >>$ >>>maybe you apply the same process. >>>  >>>Thanks, dave. >>>  >>>  >>> 	 >>>        >>> * >>The answer to that one is, "It depends!" >>@ >>In some environments; e.g. academia, you can do a VMS upgrade  >>whenever  H >>you feel like it.  In others, e.g. a production environment where the E >>business depends on the system and applications being available an  6 >>upgrade can be a very complex and expensive project. >>? >>I recall one upgrade I took part in a few years ago in which   >>we upgraded D >>VMS/VAX V5.5-2 to  6.2 while simultaneously upgrading to the then > >>current version of RDB, ACMS, and thirty-nine other layered  >>products.   7 >>We had to build a test cluster with cloned disks and   >>benchmark it.  Then H >>we upgraded the test cluster and bench marked it again.   Every batch E >>job was run in both the original and upgraded environments and the  @ >>results compared for accuracy and performance.   A lot of RDB 
 >>tuning was  F >>done and benchmarks were repeated.  Then we upgraded the production B >>cluster on a long weekend.   We invested 3,250 man hours in the G >>project.  It was done this way because, if the system had not worked  8 >>properly on Monday morning, we would have been out of  >>business until we $ >>could restore the status quo ante. >>H >>You don't invest that amount of resources unless you bloody well have H >>to.   If we had not needed some of the features in the new version of   >>RDB we would not have done it. >>= >>I did another one a couple of years later for a subsequent   >>employer.  We H >>needed to upgrade to faster hardware.   The faster hardware would not 9 >>run the older version of VMS that we were using so the   >>hardware upgrade  H >>required that we also upgrade VMS.   The VMS upgrade required that we = >>also upgrade Oracle.   The Oracle upgrade required that we   >>upgrade the = >>ERP package.  I was never told just how much that one cost   >>but $500,000  < >>seems like a reasonable estimate.  About $100,000 of that  >>went for two  ) >>Alpha server ES40s, the rest was labor.  >> >> >>     >> > I >Yep, while some folks are sometimes amazed at how many VAX's and Alpha's H >running older versions of OpenVMS are out there, he reality is that theC >same can also be said for Solaris, AIX and likely even to a larger  >extent - Windows. > F >While there are certainly good reasons for keeping relatively currentI >(within 1 or 2 point releases), one always needs to balance the costs of  >the upgrade to the business.  > I >However, one of the biggest issues associated with poor IT Management is H >that IT gets ignored for to long, then the costs of being to far behind >become exponentially higher.  > F >What typically happens - all of a sudden there is a critical piece ofI >new business functionality required that requires new systems and/or new F >OS versions. Cust single billing using Web App and/or J2EE/.Net basedI >requirement as an example. Because the upgrade to this new functionality I >has so many things that need to be changed on so many different systems, I >the cost in terms of what needs to be tested is much, much higher. Then, E >when the IT shop gets pressured to upgrade in timeframes that do not G >allow proper testing, something goes wrong and the business only views $ >the issue as "IT screwed up again". > C >I recently was involved in the multi-platform inventory of a large H >financial institution. They had approx 6,800 servers in their 4 main NAE >datacenters, but they had a much larger number if you included their H >regional sites as well. Solaris 2/7/8, Windows NT3.51 (one system), NT4F >(some W2K), AIX 4, HP-UX v10.x, Tru64 V4 was what they were primarilyH >running. Numerous mergers with different IT environments, lack of focusF >on IT (cost vs. competitive advantage), internal politics etc had all; >combined to result in a real mess from an IT perspective.   > G >And of course the App's folks thought they had good App inventory, but H >ask them for a list of servers where each App was installed and all youI >got was the deer in the lights look. "Go talks to the OPS folks" was the = >std answer who of course, only knew the OS and HW inventory.  > E >This may be a case of a company with poor IT management focus, but I H >would suggest that there are a huge number of companies out there today6 >that are very much in the same boat as this Customer. > B >On a related note, a recent IT magazine was reviewing why so manyD >Companies were still using W2K and not W2K3. One of the interestingG >comments was "fear of the unknown" i.e. Techies were afraid of the new D >W2K3 version and after spending so much time getting W2K relativelyF >stable, were reluctant to start over again with the new W2K3 version. > H >Yes, there needs to be a balance between always upgrading to the latestE >version and the needs of the business. However, I would also suggest I >that the future needs of the business *requires* that the IT environment G >*not* be ignored to the point where it becomes the equivalent of a run E >down slum where future support costs rise exponentially and IT is no 9 >longer viewed as a competitive advantage to the company.  >  >Regards >  >Kerry Main  >Senior Consultant >HP Services Canada  >Voice: 613-592-4660 >Fax: 613-591-4477 >kerryDOTmainAThpDOTcom  >(remove the DOT's and AT)   > 5 >OpenVMS - the secure, multi-site OS that just works.  >  >  >  >  >  >  >    > 
 Gee Kerry,  H It sounds as if you have worked at some of the same places I have!!! :-)   ------------------------------  % Date: Sun, 10 Jul 2005 20:31:11 -0500 2 From: David J Dachtera <djesys.nospam@comcast.net> Subject: Re: Upgrading VMS+ Message-ID: <42D1CBDF.2F2E2C95@comcast.net>    David D Miller wrote:  >  > G'day: > K > How do you make a decision to upgrade?  Is your decision ruled by Policy? I > By convenience (nothing else to do today)?  Or do you have an objective  > criteria that you apply? > L > Specifically, is 7.3 good enough (in some sense) or is 7.3-2 a better bet? > L > I suppose the same question could be applied to patches -- but then again,# > maybe you apply the same process.   E At work, the decision is based largely on what the application vendor ; has "blessed". At home, it's based purely on personal need.   H That said, V7.3-2 is published as the official "landing zone" for V7.3-xE support. Some features that work o.k. in V7.3 work rather smoother in @ V7.3-2. V7.3-2 also has support for longer strings than previous versions. V8.2 continues that.    ; My advice: Go to V7.3-2, unless software issues prevent it.    --   David J Dachtera dba DJE Systems  http://www.djesys.com/  ) Unofficial OpenVMS Hobbyist Support Page: " http://www.djesys.com/vms/support/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/   " Unofficial OpenVMS-IA32 Home Page: http://www.djesys.com/vms/ia32/    Coming soon:& Unofficial OpenVMS Marketing Home Page   ------------------------------   End of INFO-VAX 2005.383 ************************