INFO-VAX Tue, 24 Jun 2008 Volume 2008 : Issue 350 Contents: Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Re: Book "OpenVMS for Unix Users" Book: "Writing for the Reader" by John O'Rourke Re: Book: "Writing for the Reader" by John O'Rourke Re: Does anyone know if and how VMS figured in this? Re: How things change for VMS Re: KZPEA ID question Re: KZPEA ID question Re: LMF and abandonned products Re: LMF and abandonned products Re: LMF and abandonned products Re: LMF and abandonned products Re: PerfectCache on Integrity - anyone else using it? Re: Very cool Ethernet speedup on OpenVMS Re: Virtualized VMS in clusters (general questions) RE: Virtualized VMS in clusters (general questions) Re: Virtualized VMS in clusters (general questions) Re: Virtualized VMS in clusters (general questions) Re: Virtualized VMS in clusters (general questions) Re: Virtualized VMS in clusters (general questions) Re: Virtualized VMS in clusters (general questions) Re: Virtualized VMS in clusters (general questions) Re: Virtualized VMS in clusters (general questions) ---------------------------------------------------------------------- Date: Mon, 23 Jun 2008 20:09:14 +0200 From: Michael Unger Subject: Re: Book "OpenVMS for Unix Users" Message-ID: <6ca7ifF3feejoU1@mid.individual.net> On 2008-06-23 17:51, "Richard B. Gilbert" wrote: > [...] > > If you have access to a VMS system, "HELP" is your friend! Also the VMS > documentation is available on-line at HP.COM. Google for it! (HP's > search engine couldn't find its ass with both hands!) Well, starting from it was a really small step to ... Michael -- Real names enhance the probability of getting real answers. My e-mail account at DECUS Munich is no longer valid. ------------------------------ Date: Mon, 23 Jun 2008 20:16:29 +0200 From: Ralf Folkerts Subject: Re: Book "OpenVMS for Unix Users" Message-ID: Richard B. Gilbert schrieb: [...] > I just checked Amazon.com and the book appears to be out of print! > > If you have access to a VMS system, "HELP" is your friend! Also the VMS > documentation is available on-line at HP.COM. Google for it! (HP's > search engine couldn't find its ass with both hands!) > Hi Richard, thanks for the hint! Well, I do use "help" for Reference-Purposes (i.e. I know the Command but not it's Arguments). However, my Problem is mostly one of the different Concepts / Design / Ideas / Semantics between Unix and OpenVMS. So I'm looking for a book / material that helps me get into them. Thought that a Book that puts Unix and OpenVMS face to face might be an interesting way to learn... Cheers, _ralf_ ------------------------------ Date: Mon, 23 Jun 2008 20:29:46 +0200 From: Ralf Folkerts Subject: Re: Book "OpenVMS for Unix Users" Message-ID: yyyc186 schrieb: > You might also want to visit http://www.theminimumyouneedtoknow.com Hi, thanks for that Hint! Just bookmarked the Site :-) The Book "The Minimum you need to know to be an OpenVMS Application Developer" might be helpful! Cheers, _ralf_ ------------------------------ Date: Mon, 23 Jun 2008 20:27:17 +0200 From: Ralf Folkerts Subject: Re: Book "OpenVMS for Unix Users" Message-ID: <6h65j5-183.ln1@boss.home.folkerts-net.de> Anton Shterenlikht schrieb: [...] > I recommend "unix for openvms users", isbn 1-55558-276-1 Hi Anton, thanks for the hint! However, I'm afraid such a book describes things from the "false" (for me) perspective. I have almost no plan of OpenVMS but know Unix for about 10 Years. I'm afraid a "Unix for OpenVMS Users" would describe the Unix-things I already know somewhat in depth but at best mention their OpenVMS Counterpart. That said I might be completely wrong here... I just checked with Amazons "Search Inside"; well, it really seems to be the Opposite of what I am looking for :-( On the Pages I saw it does to much description of the Unix-Commands (which I already know). Cheers, _ralf_ ------------------------------ Date: Mon, 23 Jun 2008 15:29:19 -0400 From: "Richard B. Gilbert" Subject: Re: Book "OpenVMS for Unix Users" Message-ID: Ralf Folkerts wrote: > Richard B. Gilbert schrieb: > > [...] >> I just checked Amazon.com and the book appears to be out of print! >> >> If you have access to a VMS system, "HELP" is your friend! Also the >> VMS documentation is available on-line at HP.COM. Google for it! >> (HP's search engine couldn't find its ass with both hands!) >> > > Hi Richard, > > > thanks for the hint! Well, I do use "help" for Reference-Purposes (i.e. > I know the Command but not it's Arguments). > Try HELP HELP for an explanation of HELP! > However, my Problem is mostly one of the different Concepts / Design / > Ideas / Semantics between Unix and OpenVMS. So I'm looking for a book / > material that helps me get into them. Thought that a Book that puts Unix > and OpenVMS face to face might be an interesting way to learn... > The differences between Unix and VMS are largely philosophical. One of the biggest differences is that in Unix EVERYTHING may be treated as a file! That's easy because a file is simply an ordered sequence of bytes and any semblance of "structure" is a figment of the user's imagination. VMS was designed from the ground up whereas Unix just grew. Unix has some very nice tools, some of which have been ported to VMS; e.g. grep and gawk. Brevity was of great importance to the Unix developers. Anyone who has ever tried to type anything on a Model 33 Teletype will not need to ask why! For those who aren't old farts like me and never had the "pleasure", operating the keyboard required a degree of force greater than any other keyboard device I've ever used, including a mechanical typewriter! ------------------------------ Date: Mon, 23 Jun 2008 16:08:34 -0400 From: JF Mezei Subject: Re: Book "OpenVMS for Unix Users" Message-ID: <486002c5$0$23832$c3e8da3@news.astraweb.com> Ralf Folkerts wrote: > > However, my Problem is mostly one of the different Concepts / Design / > Ideas / Semantics between Unix and OpenVMS. So I'm looking for a book / > material that helps me get into them. Thought that a Book that puts Unix > and OpenVMS face to face might be an interesting way to learn... Prior to VMS 7.3, the HELP topics that were in lowercase were very informative general introduction to concepts. There was Introduction (now empty in 8.3), DCL_Tips, Hints etc. They covered a lot of topics, such as Queue management which gave you a short list of varous commands needed to manage queues. At 7.3 they started to mess things up, and with 8.3, the help is now structured in a non obvious way for those topics ("Introduction" is now empty for instance). ------------------------------ Date: Mon, 23 Jun 2008 16:10:53 -0400 From: JF Mezei Subject: Re: Book "OpenVMS for Unix Users" Message-ID: <4860034f$0$23832$c3e8da3@news.astraweb.com> Another thing: On VMS, the DEC keyboards had a "HELP" button instead of F15, and a DO button instead of F16. Most screen oriented applications would support those. So pressing F15 while in TPU would get you a help screen for instance. ------------------------------ Date: Mon, 23 Jun 2008 13:23:11 -0700 (PDT) From: johnwallace4@gmail.com Subject: Re: Book "OpenVMS for Unix Users" Message-ID: On Jun 23, 7:16 pm, Ralf Folkerts wrote: > Richard B. Gilbert schrieb: > > [...] > > > I just checked Amazon.com and the book appears to be out of print! > > > If you have access to a VMS system, "HELP" is your friend! Also the VMS > > documentation is available on-line at HP.COM. Google for it! (HP's > > search engine couldn't find its ass with both hands!) > > Hi Richard, > > thanks for the hint! Well, I do use "help" for Reference-Purposes (i.e. > I know the Command but not it's Arguments). > > However, my Problem is mostly one of the different Concepts / Design / > Ideas / Semantics between Unix and OpenVMS. So I'm looking for a book / > material that helps me get into them. Thought that a Book that puts Unix > and OpenVMS face to face might be an interesting way to learn... > > Cheers, > _ralf_ As someone who learnt VMS and Unix at the same time (more accurately, Unixes, as there was V7, Sys V, and two flavours of BSD4 in the shop), the thing that struck me most at the time is that VMS shows signs of being designed (and even reviewed) for functionality, logicalness and consistency, whereas Unix was mostly just strung together so it mostly worked (Richard already said this). E.g. back then, fd = open() took either two or three parameters depending on whose Unix it was. From my point of view this consistency shows particularly in things like resources and locking (on which the file system, RMS, proper databases, clusters, etc are built), the networking, the command language environment, the security architecture, the common language runtime and consistent APIs... the list goes on, and really there is no "compare and contrast" at this conceptual level, Unix and VMS are just too different. If you are a developer and you are willing to learn something different, once you have grasped the basics of VMS system services and mechanisms like descriptors, ASTs, and other VMS-unique approaches, then afaict the rest should follow relatively easily, much more easily than for someone who is moving the other way, to an environment where consistency of approach is largely unheard of. One of the better DEC manuals for the new VMS user/programmer would have been the MicroVMS User's Guide and/or the matching Programmers Guide. It would be very dated now (VMS V4 era), but stuff in there would largely still work, and covers the stuff you're likely to actually need, in a semi-tutorial style, where the full doc set covers everything there was to know, most of which you'd never need, and much of it was in a "reference" style which was fine if you knew what you wanted, and OK eventually if you didn't... If you do happen across a MicroVMS docset, it's worth keeping for that reason alone, imo. Hth, good luck. ------------------------------ Date: 23 Jun 2008 15:48:14 -0500 From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) Subject: Re: Book "OpenVMS for Unix Users" Message-ID: In article , "Richard B. Gilbert" writes: > That's easy because a file is simply an ordered sequence of bytes > and any semblance of "structure" is a figment of the user's imagination. Says who? ------------------------------ Date: Mon, 23 Jun 2008 21:04:56 +0000 From: ChrisQ Subject: Re: Book "OpenVMS for Unix Users" Message-ID: <9eU7k.68798$8H5.56424@newsfe10.ams2> Ralf Folkerts wrote: > Hi, > > while searching for Books that might help me getting into OpenVMS I > found the abovementioned Book "OpenVMS for Unix Users" (ISBN: > 1555583253) from Digital Press. > > However, I was unable to locate a Copy or find a Review. I searched > Alibris and the Sites mentioned in the OpenVMS-FAQ (inkl. elsevier) > amongst googling for it. > > Does anyone have a hint where that book might be available? Or could > anyone either recommend or disadvice that Book? I have been using Unix > (and FreeBSD and Linux) for many years now, so I hope that this book > might help me getting onto track faster... > > Cheers, > _ralf_ Hi, Just checked abe books and they list 42 copies, from ~17 ukp upwards. http://www.abebooks.com or http://www.abebooks.co.uk and in other countries. First place I look for books and generally end spending far too much money :-)... Chris ------------------------------ Date: Mon, 23 Jun 2008 17:14:12 -0400 From: "Richard B. Gilbert" Subject: Re: Book "OpenVMS for Unix Users" Message-ID: <5YydnYsrxcsyj_3VnZ2dnUVZ_gGdnZ2d@comcast.com> Bob Koehler wrote: > In article , "Richard B. Gilbert" writes: > >> That's easy because a file is simply an ordered sequence of bytes >> and any semblance of "structure" is a figment of the user's imagination. > > Says who? > Says me! You can create any sort of file structure you like but Unix neither knows nor cares! Indexed sequential? You can roll your own, but Unix neither knows nor cares? Direct access? You can do it but the O/S doesn't know or care. You can do any of the fancy stuff that VMS provides but you have to do it all yourself or purchase commercial software that does what you want. ------------------------------ Date: Mon, 23 Jun 2008 20:45:27 -0400 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= Subject: Re: Book "OpenVMS for Unix Users" Message-ID: <486043a1$0$90273$14726298@news.sunsite.dk> Bob Koehler wrote: > In article , "Richard B. Gilbert" writes: >> That's easy because a file is simply an ordered sequence of bytes >> and any semblance of "structure" is a figment of the user's imagination. > > Says who? What Unix/filesystem provide any other service to the apps ? Arne ------------------------------ Date: Mon, 23 Jun 2008 20:46:14 -0400 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= Subject: Re: Book "OpenVMS for Unix Users" Message-ID: <486043d0$0$90273$14726298@news.sunsite.dk> Anton Shterenlikht wrote: > On Mon, Jun 23, 2008 at 05:37:37PM +0200, Ralf Folkerts wrote: >> while searching for Books that might help me getting into OpenVMS I >> found the abovementioned Book "OpenVMS for Unix Users" (ISBN: >> 1555583253) from Digital Press. > I recommend "unix for openvms users", isbn 1-55558-276-1 Wrong way in more than one meaning ! :-) Arne ------------------------------ Date: Tue, 24 Jun 2008 01:03:26 GMT From: VMS is Virus Free Subject: Book: "Writing for the Reader" by John O'Rourke Message-ID: <0ih064l46gakk13vsg95pk8fqna0r2ku6g@4ax.com> I had run across a gem of a book about how to do technical writing when I was working at Digital back in the 80s. It was a wonderful book by John O'Rourke entitle "Writing for the Reader". It, more than any other book before or since, changed the way I viewed writing. Alas, over the years, time has managed to hide that book so well that I no longer can find it. I have tried every source that I know of yet there doesn't seem to be a copy anywhere. I even asked the publisher if there was a PDF version I could buy but to no avail. If any of you has a spare copy of the gem that you'd be willing to part with, I'd be more than willing to pay for it or find some way of compensating you. If anyone can help, I'll put up a real Email address so that we can take this offline. Thx. ------------------------------ Date: 23 Jun 2008 21:32:14 -0400 From: Rich Alderson Subject: Re: Book: "Writing for the Reader" by John O'Rourke Message-ID: VMS is Virus Free writes: > I had run across a gem of a book about how to do technical writing > when I was working at Digital back in the 80s. It was a wonderful book > by John O'Rourke entitle "Writing for the Reader". It, more than any > other book before or since, changed the way I viewed writing. No such title by such an author is listed in the Library of Congress, nor does the title appear by any other writer. The closest the LoC (or ABEbooks, for that matter) comes is Mary Lynch's _Writing in the Disciplines: A Reader_. I wonder if it might have been a DEC-internal publication. -- Rich Alderson "You get what anybody gets. You get a lifetime." news@alderson.users.panix.com --Death, of the Endless ------------------------------ Date: Mon, 23 Jun 2008 20:38:19 +0000 (UTC) From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) Subject: Re: Does anyone know if and how VMS figured in this? Message-ID: In article <7c7dbeb1-5c89-4a24-99ab-0dce9c675648@b1g2000hsg.googlegroups.com>, yyyc186 writes: > How quickly they forget the lessons of Sept. 11. Those trading > companies using a distributed OpenVMS cluster continued trading > through the end of business. Those using either Windows or HP/UX went > out of business. Except the one operation which had two data centres---one in each tower of the WTC. Even with VMS, the system manager still has to think at least a bit. ------------------------------ Date: Mon, 23 Jun 2008 19:32:46 GMT From: winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing) Subject: Re: How things change for VMS Message-ID: <00A7B8B5.B4200A76@SSRL.SLAC.STANFORD.EDU> In article <3a8a0530-fdf6-452b-8096-735f62111a0e@r66g2000hsg.googlegroups.com>, Oswald writes: >On 21 jun, 06:01, Hein RMS van den Heuvel >wrote: >> On Jun 19, 3:58=A0am, Oswald wrote: >> >> >> >> > On 19 jun, 07:27, "Bart.Z...@gmail.com" wrote: >> >> > > This is NOT a list of internet connected nodes, but a list of Bitnet/ >> > > Earn connected nodes. >> >> > > Regards, >> >> > > Bart Zorn >> >> > Hoi Bart, >> >> > Remeber this one: >> >> > HDETUD5 =A0 Tech Hoogeschool Delft =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 = >=A0VMS 4 4 >> >> > :-) >> >> > Oswald >> >> Ha die Oswald! >> >> And I remember: HMARL5 =A0 =A0U Limburg >> VMS 4 >> Worked once or twice on that system while working for Digital Utrecht >> in the early 80's >> >> My brother (Bart) also worked there (still does) >> I used to send him DECNET routed Email =A0as: RHEA::DECWRL::"SYVAXVDH >> @HMARL%.BITNET" >> I also seem to remember HLSDNLxx =A0DNL Leidschendam >> >> fwiw, >> Hein. > >The software (JNET) was from Joiner Associates. Never heard of them >since. Sometime before everybody got on the Internet instead, they changed their name to WINGRA. Don't know if they survive. The JNET software worked surprisingly (to me) well. I was able to set up print queues from my 8800 to an IBM box elsewhere on the SLAC campus, automate fetching monthly report files from there (and then use VMS tools to repackage and distribute to relevant users), etc. And the mail worked, too. -- Alan ------------------------------ Date: Mon, 23 Jun 2008 14:44:09 -0400 From: "David" Subject: Re: KZPEA ID question Message-ID: <89S7k.3587$NQ5.3310@bignews6.bellsouth.net> Problem I think is either your PCI riser card is no good or you need to update the system firmware. If your firmware is older than 7.2 you will need to update it. Also - make sure the jumpers are on the internal pins next to the internal 68pin connectors. From what I remember the Adaptec 39160 (3X-KZPEA-DB) will not work with the Narrow drives. It should though, work with UW U2Wide U160/U3 and U320 drives. DT -- David B Turner ============================================= Island Computers US Corp PO Box 86 Tybee GA 31328 Toll Free: 1-877 636 4332 x201, Mobile x251 Email: dturner@islandco.com International & Local: (001)- 404-806-7749 Fax: 912 786 8505 Web: www.islandco.com ============================================= "Len Whitwer" wrote in message news:e8d1daa7-c296-4fa6-ac3a-356479648785@z24g2000prf.googlegroups.com... On Jun 23, 7:09 am, Rich Jordan wrote: > On Jun 20, 9:31 pm, s...@antinode.info (Steven M. Schweda) wrote: > > > > > > > From: Rich Jordan > > > > The console initially sees the card and the drive(s), properly > > > identifying both. > > > The console is almost always happy. VMS may or may not be. > > > > When I tried to boot one drive (narrow IBM connected via narrow/wide > > > adapter, terminated by the drive) it failed on reading the boot block > > > and the drive disappeared (but PKA was still online). After INITing > > > the drive re-appeared but was still not usable. > > > > When I tried a Fujitsu U160 drive with a U160 terminator (and good > > > cables) it reported a read error on the boot block, but the drive > > > remained visible to the console. > > > > When I booted the CD (V8.3) the drive and PKA/PKB were online, but > > > attempting the mount reported drive offline errors. > > > If VMS says that the PK device is online, then you should have a > > VMS-compatible card. (With the non-VMS-compatible card, the PK device > > is left offline.) Then, if the DK/MK devices don't work, I'd suspect > > the usual cables and termination. (A bad card is possible, but not the > > most likely thing.) > > > ------------------------------------------------------------------------ > > > Steven M. Schweda sms@antinode-info > > 382 South Warwick Street (+1) 651-699-9818 > > Saint Paul MN 55105-2547 > > Steven, > the home DS10L liked the card, and liked the drives I had at home > to connect, once I told the card which were narrow and slow. Its the > real deal. I don't know what is wrong with the box at work (or the > cables/drives there). > > Now I can speed up the home box substantially. Just need to pick > up some cheap surplus U160 drives to replace the old Ultra-40s. > > Rich- Hide quoted text - > > - Show quoted text - Hi Rich: We have had the same problem with some SCSI controllers in that they WON''T work with OVMS but WILL work on UNIX.(Driver Issue) We have the real 3X-KZPEA-DB's here CHEAP at $95.00 if that will help you. Len Whitwer -Len Whitwer Puget Sound Data Systems, Inc. 19501 144th Ave. NE Suite D-100 Woodinville, WA 98072 e-mail mailto:len@psds.com Internet: http://www.psds.com Toll Free: (866)857-0710 Tel: (425) 488-0710 Fax: (425) 488-6414 ------------------------------ Date: Mon, 23 Jun 2008 15:18:37 -0700 (PDT) From: Rich Jordan Subject: Re: KZPEA ID question Message-ID: On Jun 23, 1:44=A0pm, "David" wrote: > Problem I think is either your PCI riser card is no good or you need to > update the system firmware. > If your firmware is older than 7.2 you will need to update it. > Also - make sure the jumpers are on the internal pins next to the interna= l > 68pin connectors. > > From what I remember the Adaptec 39160 (3X-KZPEA-DB) will not work with t= he > Narrow drives. > It should though, work with UW U2Wide U160/U3 and U320 drives. > > DT > > -- > David B Turner > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > > Island Computers US Corp > PO Box 86 > Tybee GA 31328 > > Toll Free: 1-877 636 4332 x201, Mobile x251 > Email: dtur...@islandco.com > International & Local: (001)- 404-806-7749 > Fax: 912 786 8505 > Web:www.islandco.com > > =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D"Len Whitwer= " wrote in message > > news:e8d1daa7-c296-4fa6-ac3a-356479648785@z24g2000prf.googlegroups.com... > On Jun 23, 7:09 am, Rich Jordan wrote: > > > > > On Jun 20, 9:31 pm, s...@antinode.info (Steven M. Schweda) wrote: > > > > From: Rich Jordan > > > > > The console initially sees the card and the drive(s), properly > > > > identifying both. > > > > The console is almost always happy. VMS may or may not be. > > > > > When I tried to boot one drive (narrow IBM connected via narrow/wid= e > > > > adapter, terminated by the drive) it failed on reading the boot blo= ck > > > > and the drive disappeared (but PKA was still online). After INITing > > > > the drive re-appeared but was still not usable. > > > > > When I tried a Fujitsu U160 drive with a U160 terminator (and good > > > > cables) it reported a read error on the boot block, but the drive > > > > remained visible to the console. > > > > > When I booted the CD (V8.3) the drive and PKA/PKB were online, but > > > > attempting the mount reported drive offline errors. > > > > If VMS says that the PK device is online, then you should have a > > > VMS-compatible card. (With the non-VMS-compatible card, the PK device > > > is left offline.) Then, if the DK/MK devices don't work, I'd suspect > > > the usual cables and termination. (A bad card is possible, but not th= e > > > most likely thing.) > > > > ---------------------------------------------------------------------= --- > > > > Steven M. Schweda sms@antinode-info > > > 382 South Warwick Street (+1) 651-699-9818 > > > Saint Paul MN 55105-2547 > > > Steven, > > the home DS10L liked the card, and liked the drives I had at home > > to connect, once I told the card which were narrow and slow. Its the > > real deal. I don't know what is wrong with the box at work (or the > > cables/drives there). > > > Now I can speed up the home box substantially. Just need to pick > > up some cheap surplus U160 drives to replace the old Ultra-40s. > > > Rich- Hide quoted text - > > > - Show quoted text - > > Hi Rich: > > We have had the same problem with some SCSI controllers in that > they WON''T work with OVMS but WILL work on UNIX.(Driver Issue) We > have the > real 3X-KZPEA-DB's here CHEAP at $95.00 if that will help > you. > > Len Whitwer > > -Len Whitwer > Puget Sound Data Systems, Inc. > 19501 144th Ave. NE Suite D-100 > Woodinville, WA =A098072 > e-mail =A0 =A0mailto:l...@psds.com > Internet:http://www.psds.com > Toll Free: (866)857-0710 > Tel: (425) 488-0710 > Fax: (425) 488-6414 David entirely likely it was the drives or the cables. The card worked at home with a narrow CDROM drive in a BA364 tower (along with two Ultra-wide SBBs) on the 'B' channel, which is not selectable for narrow/wide. At work I had the IBM Ultra-narrow hard drive connected via a (Compaq/HP) 68-50 adapter but never tried setting the port termination from the bios due to lack of time; you can do that on the 'A' channel; I also never tried plugging it into the 50-pin header connector for the same reason. I doubt its the riser because the system is 100% reliable with the KZPBA SCSI card installed in it. Not surprising my home systems and equipment are more accommodating (and slightly newer/more advanced overall) than the test box at work ;) Rich ------------------------------ Date: Mon, 23 Jun 2008 20:49:15 -0400 From: =?ISO-8859-1?Q?Arne_Vajh=F8j?= Subject: Re: LMF and abandonned products Message-ID: <48604484$0$90273$14726298@news.sunsite.dk> JF Mezei wrote: > If the company that owns the rights to some software is wound down > without selling any of its remaining assets because they are considered > worthless, is there anyone left to enforce the copyright on those products ? > > In the case of a book, the writer would likely have surviving family who > would inherit the rights, or perjaps the publisher would get the upon > death of writer. > > But in the case of a small company that just stops existing, it isn't > obvious what happens to copyrights. Someone will likely still own the rights. And the fact that they are not selling it does not prevent them from suing. Arne ------------------------------ Date: 24 Jun 2008 01:02:53 GMT From: billg999@cs.uofs.edu (Bill Gunshannon) Subject: Re: LMF and abandonned products Message-ID: <6cavdtF3f84fnU3@mid.individual.net> In article <485bc1a9$0$20523$c3e8da3@news.astraweb.com>, JF Mezei writes: > Bill Gunshannon wrote: > >> No pandora's box. It is flat out illegal. Just like faking licenses >> for VMS or Windows or MS Office. > > Ok, just for the sake of discussion here. Say company X writes a piece > of software that was last updated in 1990. Company X has since gone out > of business. > > Wouldn't copyright eventually expire on said piece of software and it > would then become fair game to start to use it without paying a licence ? Sure. Current Berne Convention rules say "50 years after the death of the author" Not sure how author would be defined in the case of a corporate work for hire, but 50 years is still along time. :-) > > > OK, lets take a theoretical case of VAX-Book. It is sold to SSI > technologies in the 1990s. Since then, SSI hasn't developped it, but > remains in business due to other products. > > What does SSI have to do to keep "ownership" of VAX-Book and prevent the > copyright from lapsing ? Just fill out some form every 5 years ? Probably nothing. But, they could always just modify some piece of it with a new copyright every year or so and there is no requirement that they release it outside of the corporation. bill -- Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves billg999@cs.scranton.edu | and a sheep voting on what's for dinner. University of Scranton | Scranton, Pennsylvania | #include ------------------------------ Date: 24 Jun 2008 01:11:10 GMT From: billg999@cs.uofs.edu (Bill Gunshannon) Subject: Re: LMF and abandonned products Message-ID: <6cavteF3f84fnU4@mid.individual.net> In article , m.kraemer@gsi.de (Michael Kraemer) writes: > In article <485bc1a9$0$20523$c3e8da3@news.astraweb.com>, JF Mezei > writes: >> >> What does SSI have to do to keep "ownership" of VAX-Book and prevent the >> copyright from lapsing ? Just fill out some form every 5 years ? > > That's a rhethorical question, eh ? > We all know the probable Fatwah of our IP Mullah B.G. > If you apply common sense (i.e. the opposite of IP laws), > those rights should expire after a couple of years of > non-marketing, just as patents do. Why should my rights to the fruits of my labors ever expire? > Maybe some Ralph Nader should pursue that issue. "First, we kill all the lawyers." bill -- Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves billg999@cs.scranton.edu | and a sheep voting on what's for dinner. University of Scranton | Scranton, Pennsylvania | #include ------------------------------ Date: 24 Jun 2008 01:11:50 GMT From: billg999@cs.uofs.edu (Bill Gunshannon) Subject: Re: LMF and abandonned products Message-ID: <6cavumF3f84fnU5@mid.individual.net> In article , "Richard B. Gilbert" writes: > JF Mezei wrote: >> Bill Gunshannon wrote: >> >>> No pandora's box. It is flat out illegal. Just like faking licenses >>> for VMS or Windows or MS Office. >> >> Ok, just for the sake of discussion here. Say company X writes a piece >> of software that was last updated in 1990. Company X has since gone out >> of business. >> >> Wouldn't copyright eventually expire on said piece of software and it >> would then become fair game to start to use it without paying a licence ? >> >> >> OK, lets take a theoretical case of VAX-Book. It is sold to SSI >> technologies in the 1990s. Since then, SSI hasn't developped it, but >> remains in business due to other products. >> >> What does SSI have to do to keep "ownership" of VAX-Book and prevent the >> copyright from lapsing ? Just fill out some form every 5 years ? > > ISTR that copyright is good for fifty years or the life of the author, > whichever is greater. Berne Convention says 50 years after the death of the author. bill -- Bill Gunshannon | de-moc-ra-cy (di mok' ra see) n. Three wolves billg999@cs.scranton.edu | and a sheep voting on what's for dinner. University of Scranton | Scranton, Pennsylvania | #include ------------------------------ Date: Tue, 24 Jun 2008 06:39:57 +0800 From: "Richard Maher" Subject: Re: PerfectCache on Integrity - anyone else using it? Message-ID: Hi, FWIW, I'd just like to throw Orrible Oracle's Cache-Fusion into the discussion. (Sadly something that Oracle/Rdb has never had 'cos they're not interested in clusters since they rule out the deployment of Norm's blue-eyed Row-Cache. It's all about them really :-( ) Regards Richard Maher wrote in message news:67fc8bc1-3a04-4a35-a893-1222b44f0504@e39g2000hsf.googlegroups.com... On 21 Jun, 20:36, Marty Kuhrt wrote: > etmsr...@yahoo.co.uk wrote: > > On 17 Jun, 15:21, Jan-Erik Söderholm > > wrote: > >> etmsr...@yahoo.co.uk wrote: > >>> Hiya, > >>> We've had a couple of sites needing to use PerfectCache on Integrity, > >>> VMS versions 8.3 and 8.3-1H1. The problems that we've experienced > >>> make me suspect that few sites are now using PerfectCache so I'm > >>> wondering if my suspicions are correct? > >>> Is anyone out there using PerfectCache on Integrity? Experiences? > >>> Thanks in advance > >>> Steve > >> Doesn't the current cache features in later VMS version > >> more or less make these 3'rd party cache products obsolete ? > > >> Jan-Erik. > > > No! > > XFC is fine on a single node, but if you have a cluster that's opening > > files for write as well as read on more than one node, XFC backs off > > and won't cache it. Not helpful when you want to cache it! > > Wow! I just assumed XFC would at least try to match the capabilities of > the third party caching products. > > When I was working tech lead with I/O Express (Executive Software) and > later CacheManager (SyMark) we had block-level write invalidates to > notify other nodes of a write. This would allow the other nodes in the > cluster to do whatever without having to consider the whole file > suspect. (We worked at the block level, anyway) That was over a decade > ago. > > Of course, since I don't do product support anymore, I don't research > product features. Maybe XFC was updated at some point.- Hide quoted text - > > - Show quoted text - Thanks for you message Marty. As they say, never assume anything. I had assumed that XFC would be cluster compatible and cluster coherent, only to have it explained in terms that meant it certainly wasn't if a file is open for both read and write. Are Cache Manager or I/O Express capable of working cluster coherently on Integrity? Last I looked I/O Express wasn't. Steve ------------------------------ Date: Mon, 23 Jun 2008 15:52:22 -0400 From: JF Mezei Subject: Re: Very cool Ethernet speedup on OpenVMS Message-ID: <485ffef8$0$22500$c3e8da3@news.astraweb.com> Rich Jordan wrote: > Alphaserver DS15 systems have embedded Intel controllers instead of > the DEC ones used on DS10/10L systems. Would they be native Intel controllers, or the DEC controllers that ended up at Intel when DEC sold its ethernet controller business to Intel ? ------------------------------ Date: Mon, 23 Jun 2008 14:01:34 -0400 From: "Stanley F. Quayle" Subject: Re: Virtualized VMS in clusters (general questions) Message-ID: <485FACBE.22618.84D8DB9@infovax.stanq.com> > I think the relatively trivial answer is to run your virtual nodes as > satellite nodes over Ethernet, and then everything just works. This does work in the current CHARON-* (VAX, Alpha) products. > Is there some way in which I can dedicate a real host Fibre Channel > connection to my virtual machines? The new versions of the CHARON products will support FC device names. That will allow you to interoperate on a SAN with other CHARON or physical systems running FC > What about single-system-disk clusters? This works now, if both systems can access the same disk directly, just as with physical systems. You could use a multi-port disk box, iSCSI, SAN, etc. > Who arbitrates access to it? Can you do shadowed system disk on it? The Lock Manager. You could shadow the disk if you were running a multi-site CHARON cluster, no problem. > On the Itanium-emulated-on-Itanium approach (which isn't supported until > VMS 8.4) can you successfully run a multi-node cluster all on one physical > box with a single system disk? Good question. Have to ask HP on that one... > But I'm not clear whether the hp/ux virtual switch supports, eg, DECnet, > MOP, SCS packets There's really no point to it otherwise. > > Can you do host-based volume-shadowing on SAN disks that are actually being > > presented by non-VMS hosts (if that's how that works at all)? Presented by other clients connected to the SAN? > How about on container files presented by the host? That's in CHARON today, whether SAN-connected or local disk. --Stan Quayle Quayle Consulting Inc. ---------- Stanley F. Quayle, P.E. N8SQ Toll free: 1-888-I-LUV-VAX 8572 North Spring Ct., Pickerington, OH 43147 USA stan-at-stanq-dot-com http://www.stanq.com/charon-vax.html "OpenVMS, when downtime is not an option" ------------------------------ Date: Mon, 23 Jun 2008 19:47:59 GMT From: winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing) Subject: RE: Virtualized VMS in clusters (general questions) Message-ID: <00A7B8B7.D49DB8D7@SSRL.SLAC.STANFORD.EDU> In article <003c01c8d557$967b4070$2802a8c0@CHARONLAP>, "Peter Weaver" writes: >> I'm trying to wrap my head around how virtualized VMS systems participate >> in >> certain aspects of clustering and volume shadowing. There may be >> something >> I'm just not getting. So this is kind of general. >>... > >You are right that a user (or even a system manager) that logs into >CHARON-VAX or CHARON-AXP will just see a faster VAX or Alpha. Nothing from >VMS will let them know that it is running on virtual hardware. > >The network driver will direct everything the network card receives to the >virtual machine. So DECNet, SCS, IP traffic will all get to the VMS system. > >Disk drives can either be physical disks or container files. If the disk is >a container file then the Windows host will just see one big file (i.e. >D:\DISKS\DISK1.VDSK). > >If the disk is a physical disk then it can be a LUN on a Fibre SAN, an IDE >disk, a SCSI disk, or an iSCSI container (VMS does not yet support iSCSI, >but you can use iSCSI with CHARON-VAX or CHARON-AXP). If you use a physical >disk then Windows sees it as an unformatted disk. For the physical disk you >put a line in your CHARON Configuration file like; > set DUA container[1]="\\.\PhysicalDrive2" >then VMS will see DUA1 and DUA1 will be what Windows calls PhysicalDrive2. > And can you also configure a host-id, eg set $9$DUA container[1]="\\.\PhysicalDrive2" ? If I'm getting this, for the SAN device you still have to present the VMS disk to the Windows box, and just be careful that the Windows box never writes a "harmless signature" on the disk, right? The Windows box has to have ownership access because it's going to send out the format/init whatever operations on behalf of the virtualized VMS box. (But for corresponding coolness, this means you can run your virtualized VAX off a fast SAN disk and be in the 20th century without having to get an FC adapter for your BI bus, if those things even exist.) >VMS can mirror these disks, share these disks in a cluster environment or do >anything else you want to do that a hardware VAX or hardware Alpha would do. >You can shadow a system disk the same way you shadow a system disk on a >hardware VAX. Also, Charon knows what the underlying device is, but VMS doesn't. VMS can't keep Windows from messing up your underlying devices because it doesn't own them. Am I also correct that you could really mess up VMS by presenting the same real physical drive under different names to the virtualized boxes? All the Distributed Lock Manager has to go on is the device name from the config file. (It's okay if the answer is "Yes, you just have to be careful"; I'm just failing to understand how there even could be any other answer than that based on what I've learned so far about the virtualized environment. There are certainly non-virtual circumstances where you have to be careful (dual-ported RA81s with write access from both sides, for example) too, but not the same circumstances. SAN disks present with actual device names / unit numbers, and it's on the SAN administrator to present only to members of the same cluster. Everybody in the cluster who gets the unit presented by the SAN will see the same unit, so everybody thinks it's the same device. > >I have installed CHARON-VAX at a few customers where management have told >the VMS people that they are buying CHARON but the VMS people were not yet >convinced that CHARON would work. We get CHARON installed and the VMS person >starts asking questions like, "On a VAX I do a SHOW DEVICE at the dead >sergeant prompt, how do I do it on the emulator?" I answer "The same way." >We go through this a half dozen times or so and the answer is always "The >same way." Each time I have been in this situation I always smile when the >person I am working with announces to his co-workers that it is a VAX. > This stuff is very cool. -- Alan ------------------------------ Date: Mon, 23 Jun 2008 20:00:11 GMT From: winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing) Subject: Re: Virtualized VMS in clusters (general questions) Message-ID: <00A7B8B9.8920EFCE@SSRL.SLAC.STANFORD.EDU> In article <485FACBE.22618.84D8DB9@infovax.stanq.com>, "Stanley F. Quayle" writes: >> I think the relatively trivial answer is to run your virtual nodes as >> satellite nodes over Ethernet, and then everything just works. > >This does work in the current CHARON-* (VAX, Alpha) products. > >> Is there some way in which I can dedicate a real host Fibre Channel >> connection to my virtual machines? > >The new versions of the CHARON products will support FC device names. That will allow >you to interoperate on a SAN with other CHARON or physical systems running FC Ooh. Okay, that seems to answer a bunch of my questions. (The emulated VMS box sees the same device name / unit-number that the real host sees? If so, good deal.) > >> What about single-system-disk clusters? > >This works now, if both systems can access the same disk directly, just as with physical >systems. You could use a multi-port disk box, iSCSI, SAN, etc. > >> Who arbitrates access to it? Can you do shadowed system disk on it? > >The Lock Manager. You could shadow the disk if you were running a multi-site CHARON >cluster, no problem. > But the Lock Manager can't prevent the non-VMS host on which the virtualized VAX/Alpha is running from doing whatever it wants to the disk. (Of course, it can't prevent a malfunctioning disk controller from doing a wild write on a disk even in a fully-non-virtual environment; this would have the same effect.) >> On the Itanium-emulated-on-Itanium approach (which isn't supported until >> VMS 8.4) can you successfully run a multi-node cluster all on one physical >> box with a single system disk? > >Good question. Have to ask HP on that one... > >> But I'm not clear whether the hp/ux virtual switch supports, eg, DECnet, >> MOP, SCS packets > >There's really no point to it otherwise. I'd think so, but (a) the hpvm project is on hp/ux and they'll have a bunch more hp/ux and windows clients than VMS clients. If the switch is a proper switch and just passes packets, great, it'll all work for everybody. If they implemented per-protocol support - well, at one point Cisco charged extra to support DECnet packets, so who knows. I'd expect all IP and trad Unix-related services (bootp, tftp, etc) to be supported, but thought it was worth asking whether the legacy-Digital protocols were supported, especially in light of TCP/IP clustering coming in the same OS release on which virtualized VMS is supported. What if you needed TCP/IP clustering because SCS traffic wouldn't pass the virtual switch? [I'm just explaining what I was wondering about, really not trying to start a rumor.] > >> > Can you do host-based volume-shadowing on SAN disks that are actually being >> > presented by non-VMS hosts (if that's how that works at all)? > >Presented by other clients connected to the SAN? Didn't know another word for "presented" here. I meant, EVA presents to non-VMS host, non-VMS host runs virtualized VAX/Alpha/Integrity (which sees disk/file/whatever as a device name specified in local config or in startup command for virtual machine); can HBVS work. I think the answer is "yes", from what you say above and other people tell me. > >> How about on container files presented by the host? > >That's in CHARON today, whether SAN-connected or local disk. So I could have cluster-visible DS devices in which each "disk" was actually a container file on Windows disk. It makes me nervous, but it's pretty cool. Thanks! -- Alan ------------------------------ Date: Mon, 23 Jun 2008 16:00:20 -0400 From: JF Mezei Subject: Re: Virtualized VMS in clusters (general questions) Message-ID: <486000d7$0$23864$c3e8da3@news.astraweb.com> >> On the Itanium-emulated-on-Itanium approach (which isn't supported until >> VMS 8.4) can you successfully run a multi-node cluster all on one physical >> box with a single system disk? From last years presentations I saw, the HP-UX attempt at copying the 1970s IBM "VM" product will support multiple instances of an OS running, and instances of different OSs. So you could run an instance of HP-UX on HP-UX, as well as an instance of VMS and one of Linux (I assume there is some version of linux for those IA64 contraptions) But this is not a Galaxy system, so there is no interconnection between instances within the memory/machine (unless the VMS folks worked on that since last year). So give them each a ethernet card and they can talks SCS over it, give them each their onw card to access disk storage and they can talk over that, just like separate machines would. ------------------------------ Date: Mon, 23 Jun 2008 16:04:03 -0400 From: JF Mezei Subject: Re: Virtualized VMS in clusters (general questions) Message-ID: <486001b6$0$23872$c3e8da3@news.astraweb.com> One more thing: With the virtualised HP-UX thing, you can have some interesting fun with CPUs. If the hardware host has 3 CPUs, you can configure the VMS instance with any number of CPUs you want. Bassed on the presentation I saw, you could configure the VMS systsem as a single CPU machine and then make it run on all 3 host CPUs (so you get processing power of 3 CPUs that look like 1 CPU to VMS (can be interesting in terms of licences). Again, this was from a marketing presentation, so I am not sure how this maps to actual reality once the product is delivered. Remember that VMS runs as an HP-UX process. ------------------------------ Date: Mon, 23 Jun 2008 20:26:31 GMT From: winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing) Subject: Re: Virtualized VMS in clusters (general questions) Message-ID: <00A7B8BD.3683FB74@SSRL.SLAC.STANFORD.EDU> In article <486001b6$0$23872$c3e8da3@news.astraweb.com>, JF Mezei writes: >One more thing: With the virtualised HP-UX thing, you can have some >interesting fun with CPUs. > >If the hardware host has 3 CPUs, you can configure the VMS instance with >any number of CPUs you want. Bassed on the presentation I saw, you could >configure the VMS systsem as a single CPU machine and then make it run >on all 3 host CPUs (so you get processing power of 3 CPUs that look like >1 CPU to VMS (can be interesting in terms of licences). Again, this was >from a marketing presentation, so I am not sure how this maps to actual >reality once the product is delivered. I attended the "configure and boot a virtual VMS system on an HP/UX host" session - that wasnt it's name - at the HP Tech Forum just completed. All kinds of caveats about how the actual released version may be different, this is a technology demo. In specifying the virtual machine you can indicate what percentage of a CPU your virtual machine can have. I _think_ it only sees one CPU even if the execution is spread over multiple ones. I am now speaking from assumptions: My impression is that the virtualized-Itanium-on-Itanium uses the on-chip support for virtualization. This leads me to think that a virtualized machine only gets to use the processors that are on the chip that it's *on*, which means either all the cores in the socket (2) or just the core you're running on (1). Could be wrong, again, but don't see how you get to use on-chip virtualization support for a virtual machine that goes off the chip. -- Alan ------------------------------ Date: Mon, 23 Jun 2008 22:28:23 +0200 From: Wilm Boerhout Subject: Re: Virtualized VMS in clusters (general questions) Message-ID: <4860077d$0$6017$ba620dc5@text.nova.planet.nl> on 23-6-2008 21:47 Alan Winston - SSRL Central Computing wrote... > And can you also configure a host-id, eg > set $9$DUA container[1]="\\.\PhysicalDrive2" ? Yes, you can, but the $9$ part has to come from VMS, not from CHARON. I'm sure that the CHARON software could benefit from something like a SCSI port allocation class, but I don't see how this would fit in the current architecture. BTW, your controller designation to VMS can be DU, DG, DI, etc. independent of the actual hardware. I.e. the SRM console sees DGA0 disks even if it's a container file. > If I'm getting this, for the SAN device you still have to present the VMS > disk to the Windows box, and just be careful that the Windows box never > writes a "harmless signature" on the disk, right? The Windows box has to > have ownership access because it's going to send out the format/init > whatever operations on behalf of the virtualized VMS box. Please define "ownership". Windows cannot mess with the disk assigned to CHARON because the CHARON "process" owns the device. All that's done by the Windows driver is low-level block I/O. VMS can, no *must* synchronize using the DLM. > (But for corresponding coolness, this means you can run your virtualized VAX > off a fast SAN disk and be in the 20th century without having to get an FC > adapter for your BI bus, if those things even exist.) This is what we CHARON buffs do routinely :-) > Also, Charon knows what the underlying device is, but VMS doesn't. VMS can't > keep Windows from messing up your underlying devices because it doesn't own > them. Am I also correct that you could really mess up VMS by presenting > the same real physical drive under different names to the virtualized > boxes? All the Distributed Lock Manager has to go on is the device name > from the config file. > (It's okay if the answer is "Yes, you just have to be careful"; I'm just > failing to understand how there even could be any other answer than that > based on what I've learned so far about the virtualized environment. There > are certainly non-virtual circumstances where you have to be careful > (dual-ported RA81s with write access from both sides, for example) too, but > not the same circumstances. SAN disks present with actual device names / > unit numbers, and it's on the SAN administrator to present only to members > of the same cluster. Everybody in the cluster who gets the unit presented > by the SAN will see the same unit, so everybody thinks it's the same > device. On Windows, you can always identify a SAN "disk" (LUN) by is virtual SCSI ID. Instead of specifying "\\.\PhysicalDrive0" in the CHARON config file, you specify "\\.\scsi2: 0 2 1". So, the answer is indeed yes, you have to be careful, but a SAN administrator always is. > This stuff is very cool. Amen! -- Wilm Boerhout Zwolle, NL remove OLD PAINT from return address to reply ------------------------------ Date: 23 Jun 2008 16:49:59 -0400 From: brooks@cuebid.zko.hp.nospam (Rob Brooks) Subject: Re: Virtualized VMS in clusters (general questions) Message-ID: (Alan Winston - SSRL Central Computing) writes: > "Stanley F. Quayle" writes: >>> But I'm not clear whether the hp/ux virtual switch supports, eg, DECnet, >>> MOP, SCS packets >> >>There's really no point to it otherwise. > > I'd think so, but (a) the hpvm project is on hp/ux and they'll have a bunch > more hp/ux and windows clients than VMS clients. If the switch is a proper > switch and just passes packets, great, it'll all work for everybody. If they > implemented per-protocol support - well, at one point Cisco charged extra to > support DECnet packets, so who knows. I'd expect all IP and trad Unix-related > services (bootp, tftp, etc) to be supported, but thought it was worth asking > whether the legacy-Digital protocols were supported, especially in light of > TCP/IP clustering coming in the same OS release on which virtualized VMS > is supported. What if you needed TCP/IP clustering because SCS traffic > wouldn't pass the virtual switch? [I'm just explaining what I was wondering > about, really not trying to start a rumor.] At the boot camp two years ago, I demonstrated a very early implementation of VMS as a guest under HP-UX. At that time, I had a cluster of a VAX, several Alphas, several "real" I64 systems, and two virtual I64 systems. The virtual systems had no problem using MSCP in either direction, nor were there any problems using DECnet. Host-based volume shadowing worked in all its glory -- things like HBMM worked perfectly. Note that VMS under the HP-UX hypervisor *does* know that it's a guest. There are some specific code paths within the kernel that are only taken when it's a guest. Things like mapping registers for I/O don't exist for a virtual machine, so changes needed to be made to the ethernet and SCSI drivers to accomodate that. Two years ago, it was a simple programmed I/O model, used simply to get stuff working. The performance was dismal, as expected. A more robust solution is being worked on now. -- Rob Brooks MSL -- Nashua brooks!cuebid.zko.hp.com ------------------------------ Date: Tue, 24 Jun 2008 00:30:04 GMT From: winston@SSRL.SLAC.STANFORD.EDU (Alan Winston - SSRL Central Computing) Subject: Re: Virtualized VMS in clusters (general questions) Message-ID: <00A7B8DF.3CD7C9DD@SSRL.SLAC.STANFORD.EDU> In article <4860077d$0$6017$ba620dc5@text.nova.planet.nl>, Wilm Boerhout writes: >on 23-6-2008 21:47 Alan Winston - SSRL Central Computing wrote... > >> And can you also configure a host-id, eg >> set $9$DUA container[1]="\\.\PhysicalDrive2" ? > >Yes, you can, but the $9$ part has to come from VMS, not from CHARON. >I'm sure that the CHARON software could benefit from something like a >SCSI port allocation class, but I don't see how this would fit in the >current architecture. BTW, your controller designation to VMS can be DU, >DG, DI, etc. independent of the actual hardware. I.e. the SRM console >sees DGA0 disks even if it's a container file. > >> If I'm getting this, for the SAN device you still have to present the VMS >> disk to the Windows box, and just be careful that the Windows box never >> writes a "harmless signature" on the disk, right? The Windows box has to >> have ownership access because it's going to send out the format/init >> whatever operations on behalf of the virtualized VMS box. > >Please define "ownership". Windows cannot mess with the disk assigned to >CHARON because the CHARON "process" owns the device. All that's done by >the Windows driver is low-level block I/O. VMS can, no *must* >synchronize using the DLM. So of course you're running nothing but Charon on the Windows box, and nothing but Charon can do anything to the device. OK. > >> (But for corresponding coolness, this means you can run your virtualized VAX >> off a fast SAN disk and be in the 20th century without having to get an FC >> adapter for your BI bus, if those things even exist.) > >This is what we CHARON buffs do routinely :-) > >> Also, Charon knows what the underlying device is, but VMS doesn't. VMS can't >> keep Windows from messing up your underlying devices because it doesn't own >> them. Am I also correct that you could really mess up VMS by presenting >> the same real physical drive under different names to the virtualized >> boxes? All the Distributed Lock Manager has to go on is the device name >> from the config file. > >> (It's okay if the answer is "Yes, you just have to be careful"; I'm just >> failing to understand how there even could be any other answer than that >> based on what I've learned so far about the virtualized environment. There >> are certainly non-virtual circumstances where you have to be careful >> (dual-ported RA81s with write access from both sides, for example) too, but >> not the same circumstances. SAN disks present with actual device names / >> unit numbers, and it's on the SAN administrator to present only to members >> of the same cluster. Everybody in the cluster who gets the unit presented >> by the SAN will see the same unit, so everybody thinks it's the same >> device. > >On Windows, you can always identify a SAN "disk" (LUN) by is virtual >SCSI ID. Instead of specifying "\\.\PhysicalDrive0" in the CHARON config >file, you specify "\\.\scsi2: 0 2 1". So, the answer is indeed yes, you >have to be careful, but a SAN administrator always is. > >> This stuff is very cool. > >Amen! Wilm, this is the stuff that I was incoherently trying to ask about in the session at the Tech Forum. I don't think I made myself clear there, but I appreciate your patience. Thanks! --Alan ------------------------------ End of INFO-VAX 2008.350 ************************