1 INFO-VAX	Tue, 25 Mar 2003	Volume 2003 : Issue 166       Contents:3 Re: An opportunity for HP? (wes: Micromedex on VMS) 3 Re: An opportunity for HP? (wes: Micromedex on VMS) 3 Re: An opportunity for HP? (wes: Micromedex on VMS) 3 Re: An opportunity for HP? (wes: Micromedex on VMS) . Announcing: System Healthcheck V2.2 for OpenVM: Re: asymmetric cluster? question on cluster communications: Re: asymmetric cluster? question on cluster communications Re: HP doesn't know OpenVMS # Re: I Have Seen The IA64 Future !!! - Re: Microvax 3100 dead - need data from drive - Re: Microvax 3100 dead - need data from drive - Re: Microvax 3100 dead - need data from drive  Minimum version  Re: Minimum version  Re: MSIE ftp hangs FTPD $ Re: Netscape gags on VMS FTP Servers$ Re: Netscape gags on VMS FTP Servers$ Re: Netscape gags on VMS FTP Servers$ Re: Netscape gags on VMS FTP Servers$ Re: new releases of MySQL and Python$ Re: new releases of MySQL and Python$ Re: new releases of MySQL and Python0 Re: next installment - alpha 3000 console now...  Re: OpenVMS SMP License QuestionP Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha RetaiP Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai Restore with /select Re: Restore with /select Re: Restore with /select Re: Restore with /select Re: Restore with /select Re: Restore with /select Standalone backup on 6510  Re: Standalone backup on 6510 & Re: Sun unix/linux new nasty bug cert!% Re: System reqirements of OpenVMS 7.3 % Re: System reqirements of OpenVMS 7.3 % Re: System reqirements of OpenVMS 7.3 $ Re: System Special for VMS NewsGroup$ Re: System Special for VMS NewsGroup1 Re: Terminal driver for VAX synchronous interface 1 Re: Terminal driver for VAX synchronous interface > Testing HP's attitude to VMS, was: Re: HP doesn't know OpenVMS Re: X-window and Xlib questionH Re: [Q] Files not found during image backup, w/ and w/o directory-specs.H Re: [Q] Files not found during image backup, w/ and w/o directory-specs.  F ----------------------------------------------------------------------  # Date: Tue, 25 Mar 2003 13:28:09 GMT ( From: David Harrold <DHarrold@wi.rr.com>< Subject: Re: An opportunity for HP? (wes: Micromedex on VMS)@ Message-ID: <a4065c030c93852e802aa0fdf17bdb77@news.teranews.com>  7 On Mon, 24 Mar 2003 21:50:55 -0600, "David J. Dachtera"  <djesys.nospam@fsi.net> wrote:   > H >Well, I'd be careful there. Cerner with the back-end on AIX is at least+ >as viable as having the back-end on VMS...   I Except that none of the AIX sites I've talked to can scale to as large an < installation as VMS sites can and as large as ours requires.  K Also, Rumor has it that SCO is threatening to pull the license for AIX from 6 IBM because IBM is moving so much technology to Linux.  2 http://www.eweek.com/article2/0,3959,922913,00.asp   Dave        N ..............................................................................N David Harrold                              E-Mail: David_Harrold at aurora.orgI Sr. Software Systems Engineer              Phone:          (414) 647-6204 I                                            Pager:          (414) 941-4634 G Aurora Health Care                         Fax:          (414) 647-4999  3031 W. Montana Street Milwaukee, WI 53215    ------------------------------    Date: 25 Mar 2003 08:11:36 -08001 From: keithparris_NOSPAM@yahoo.com (Keith Parris) < Subject: Re: An opportunity for HP? (wes: Micromedex on VMS)= Message-ID: <cf15391e.0303250811.34bba00e@posting.google.com>   ` "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message news:<3E7FD21F.68CEAD0A@fsi.net>...I > Well, I'd be careful there. Cerner with the back-end on AIX is at least , > as viable as having the back-end on VMS...  F IBM has been waffling lately as to whether AIX will survive long-term,? given the popularity of Linux.  In contrast, VMS has the strong 9 support of HP, and a long-term path laid out ahead of it.   C And since 80% of Cerner customers use VMS compared with 20% at most C for AIX, I think it is the AIX back-end which is arguably much less  viable.   E "Today Cerner serves some 1,500 clients, 80% of whom run their Cerner 6 applications on HP OpenVMS AlphaServer technology." --= http://h71000.www7.hp.com/openvms/brochures/cerner/cerner.pdf    ------------------------------  % Date: Tue, 25 Mar 2003 15:58:23 +0000 ' From: Andrew Harrison SUNUK Consultancy < Subject: Re: An opportunity for HP? (wes: Micromedex on VMS). Message-ID: <3E807C9F.2080706@nospamn.sun.com>   Brian Tillman wrote: >>Brian, >>? >>Re: Healthcare ... IDX and Cerner are growing with OpenVMS ..  >>I >>Btw - remember, Cerner only runs on AIX or OpenVMS i.e.. not Windows or  >>HP-UX or Solaris.  >  > N > This is certainly a non-sequitor.  What does this have to do with Micromedex > disappearing on VMS?    @ Since this question is about Micromedex and not Cerner its worth? pointing out that Thomson Micromedex are not cutting Solaris or  Windows Intranet support.   G The HealthCare Series for Intranets is Supported on Solaris and Windows B so this doesn't appear to be a wholesale shift from Intranet based@ delivery to a Internet based delivery though they have dropped a number of platforms.  G EOL announcments have been made for Alpha OpenVMS, OS400, CICS, AIX and  SCO.   regards  Andrew Harrison    ------------------------------  % Date: Tue, 25 Mar 2003 16:46:54 +0000 ' From: Andrew Harrison SUNUK Consultancy < Subject: Re: An opportunity for HP? (wes: Micromedex on VMS). Message-ID: <3E8087FE.6030101@nospamn.sun.com>   Keith Parris wrote: b > "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message news:<3E7FD21F.68CEAD0A@fsi.net>... > I >>Well, I'd be careful there. Cerner with the back-end on AIX is at least , >>as viable as having the back-end on VMS... >  > H > IBM has been waffling lately as to whether AIX will survive long-term,A > given the popularity of Linux.  In contrast, VMS has the strong ; > support of HP, and a long-term path laid out ahead of it.  > E > And since 80% of Cerner customers use VMS compared with 20% at most E > for AIX, I think it is the AIX back-end which is arguably much less 	 > viable.  > G > "Today Cerner serves some 1,500 clients, 80% of whom run their Cerner 8 > applications on HP OpenVMS AlphaServer technology." --? > http://h71000.www7.hp.com/openvms/brochures/cerner/cerner.pdf   5 However IBM are supplying Cerner with the development 8 and deployment software platform for their new products.  7 WebSphere and DB2, and this is a huge threat to OpenVMS 3 if it is as people suggest the largest platform for  Cerner currently.    Regards  Andrew Harrison    ------------------------------    Date: 25 Mar 2003 10:41:29 -08001 From: susan_skonetski@hotmail.com (Sue Skonetski) 7 Subject: Announcing: System Healthcheck V2.2 for OpenVM = Message-ID: <857e9e41.0303251041.47da263b@posting.google.com>    Dear Newsgroup,   E Please be aware that this only went out this morning and may not have  reached all of hp yet.  
 Warm Regards,  sue / _______________________________________________    From:  Tom Kilgarriff  	 HPCS R&D  & Mission Critical & Proactive Services  Hewlett Packard Company  Galway, Ireland.   (   (353) 91-754128  *   tom.kilgarriff@hp.com      Sent: 25 March 2003 10:30 8 Subject: Announcing: System Healthcheck V2.2 for OpenVMS    1 Announcing:  System Healthcheck V2.2 for OpenVMS  ? System Healthcheck (SHC) V2.2 for OpenVMS...  the fastest, most 9 cost-effective way to identify security, performance, and C configuration problems before they affect your critical operations.    What is System Healthcheck? E System Healthcheck (SHC) is a tool produced by HP Services to analyze ? a system's configuration, performance, and security status. SHC C analysis results are provided in a series of reports that highlight @ any problems found on a system and provide suggestions on how toD resolve these problems. For Tru64 UNIX, Windows 2000, Windows NT andE OpenVMS systems, SHC provides performance, availability, security and > configuration assessments against sets of best practice system management rules.   A System Healthcheck can be used as a standalone System Healthcheck A assessment service or as part of HP Mission Critical Proactive 24  Service or Crirical Services.      7 For more information on System Healthcheck please visit . www.compaq.com/support/svctools/shc/index.html   @ Contact Tom Kilgarriff, HPCS R&D - mailto:tom.kilgarriff@hp.com,- 353-91-754-128  or mailto:shc.feedback@hp.com    ------------------------------    Date: 25 Mar 2003 03:53:10 -0800% From: Bart.Zorn@xs4all.nl (Bart Zorn) C Subject: Re: asymmetric cluster? question on cluster communications = Message-ID: <a98cd882.0303250353.65eec24f@posting.google.com>   v keithparris_NOSPAM@yahoo.com (Keith Parris) wrote in message news:<cf15391e.0303242008.6b77a8df@posting.google.com>...~ > Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> wrote in message news:<01KTX41QHE0W9H3FKQ@sysdev.deutsche-boerse.com>...I > > On machine 1 I had VOTES=1 and EXPECTED_VOTES=1.  On machine 2, I had K > > VOTES=1 and EXPECTED_VOTES=2.  Usually, machine 1 is in the cluster and K > > machine 2 occasionally is.  I wanted to do some tests leaving machine 2 G > > in the cluster unattended, but wanted to avoid the possibility of a B > > hang, so I set VOTES=0 on machine 2.  However, I forgot to set$ > > EXPECTED_VOTES to a lower value. > > J > > Machine 2 wasn't able to join the cluster (I will qualify this below). > > H > > I realise my mistake (I think and hope): Since the highest value of K > > EXPECTED_VOTES in the cluster was 2, quorum was 2, and since machine 2  D > > didn't contribute any votes itself, this could never be reached. > D > Right.  If machine 2 had been admitted to the cluster, the cluster > would have lost quorum.   D I don't get it. When a non-voting member would join the cluster, you would lose quorum?  G > > The system didn't join the cluster, but just continued the messages  > D > Right.  If another system with 1 vote had joined the cluster, thenF > machine 2 would have been able to join also.  So it persisted, being
 > hopeful. > M > > (I could have done a SET CLUSTER/EXPECTED_VOTES to allow it to, though I  1 > > didn't try this since I had run out of time).  > C > Actually, no.  What you needed to do is to either raise the total H > number of votes contributed, or else lower the EXPECTED_VOTES value onH > machine 2, and you couldn't do that without a reboot of one machine or > the other.  7 Of course there is always AMDS or Availability Manager!   K > > Machine 1 and machine 2 both had some locally connected SCSI disks (no  0 > > dual-ported stuff) and all were MSCP served. >  ... > > However, I could mount  * > > the disks on machine 2 from machine 1. > > J > > It seems to me that if the system is obviously not yet in the cluster L > > (i.e. it wants to join but can't), then I shouldn't be able to mount an 0 > > MSCP-served disk on it from another machine! > G > But what if there were a quorum disk and it was served to the cluster H > by machine 2?  If you enable MSCP serving on a system, VMS starts MSCPH > serving quite early in the boot so it can properly handle such a case.G >  Machine 2 had not yet achieved quorum, so it was not going to modify H > its disks itself, so there was really no risk of uncoordinated access. > E > > A (perhaps) related question: when the cluster was up and running J > > normally, SYSMAN commands executed on machine 1 would not work on nodeL > > 2, but vice versa was OK.  (The error message was something like remote ; > > authorization invalid.)  There is a common SYSUAF file.  > H > I'd double-check the SYSUAF and RIGHTSLIST logical names on machine 2., >  Are they $DEFINEd /SYSTEM and /EXECUTIVE?  3 Or use /TABLE=LNM$SYSCLUSTER (and /EXEC) for those!    Regards,  	 Bart Zorn    ------------------------------  + Date: Tue, 25 Mar 2003 15:06:26 +0100 (MET) 9 From: Phillip Helbig <HELBPHI@sysdev.deutsche-boerse.com> C Subject: Re: asymmetric cluster? question on cluster communications ; Message-ID: <01KTY201XZGIA9LL1B@sysdev.deutsche-boerse.com>   J > > > I realise my mistake (I think and hope): Since the highest value of M > > > EXPECTED_VOTES in the cluster was 2, quorum was 2, and since machine 2  F > > > didn't contribute any votes itself, this could never be reached. > > F > > Right.  If machine 2 had been admitted to the cluster, the cluster > > would have lost quorum.  > F > I don't get it. When a non-voting member would join the cluster, you > would lose quorum?  H Yes, because EXPECTED_VOTES was two.  The problem was that the VOTES of G each machine and the highest value of EXPECTED_VOTES were inconsistent.   M > > (I could have done a SET CLUSTER/EXPECTED_VOTES to allow it to, though I  1 > > didn't try this since I had run out of time).  > C > Actually, no.  What you needed to do is to either raise the total H > number of votes contributed, or else lower the EXPECTED_VOTES value onH > machine 2, and you couldn't do that without a reboot of one machine or > the other.  A Why not?  What would be an example of SET CLUSTER/EXPECTED_VOTES  C actually doing something useful?  (Above, it should be set to 1 of   course.)  7 Of course there is always AMDS or Availability Manager!   K > > Machine 1 and machine 2 both had some locally connected SCSI disks (no  0 > > dual-ported stuff) and all were MSCP served. >  ... > > However, I could mount  * > > the disks on machine 2 from machine 1. > > J > > It seems to me that if the system is obviously not yet in the cluster L > > (i.e. it wants to join but can't), then I shouldn't be able to mount an 0 > > MSCP-served disk on it from another machine! > G > But what if there were a quorum disk and it was served to the cluster H > by machine 2?  If you enable MSCP serving on a system, VMS starts MSCPH > serving quite early in the boot so it can properly handle such a case.G >  Machine 2 had not yet achieved quorum, so it was not going to modify H > its disks itself, so there was really no risk of uncoordinated access.  D OK.  Of course, there is little point of MSCP serving a quorum disk.  H > I'd double-check the SYSUAF and RIGHTSLIST logical names on machine 2., >  Are they $DEFINEd /SYSTEM and /EXECUTIVE?   Ah, forgot RIGHTSLIST!  5 > Or use /TABLE=LNM$SYSCLUSTER (and /EXEC) for those!   I In general, the old system-table logical names which VMS uses, and which  B are normally defined separately on each machine---can one use the  cluster table now?   ------------------------------    Date: 25 Mar 2003 05:45:39 -0800( From: bob@instantwhip.com (Bob Ceculski)$ Subject: Re: HP doesn't know OpenVMS: Message-ID: <d7791aa1.0303250545.27240@posting.google.com>  R brandon@dalsemi.com wrote in message news:<03032414351498@dscis6-0.dalsemi.com>...Q > >>> ... the bottom of that page is the phone number 800-282-6672.  I called ...  > C > It use to be I could call DEC Direct and get excellent support.    > M > Now they take your name and number and they will have someone get back with  > you. >  > Bull sh**. > Q > I left a message over a month ago - simple little question about power supplies  > for the DS20...  > ( > Support in this area has gone to hell. >  >  >  > John Brandon > VMS Systems Administrator  > Dallas Semiconductor > john.brandon@dalsemi.com > 972.371.4172 wk  > 972.371.4003 fx   < why don't you try Dave at Island computer or one of the many< DEC/Q/HP resellers ... you get fast service with a smile ...   ------------------------------  % Date: Tue, 25 Mar 2003 18:11:04 +0100 $ From: Michael Unger <unger@decus.de>, Subject: Re: I Have Seen The IA64 Future !!!* Message-ID: <00A1D682.5CBBE333.3@decus.de>  & "Hoff Hoffman" <hoff@hp.nospam> wrote:  S > In article <00A1D5B8.55A7AF43.4@decus.de>, Michael Unger <unger@decus.de> writes: E > :"Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com> wrote:  > : B > :> "JF Mezei" <jfmezei.spamnot@vl.videotron.ca> wrote in message. > :> news:3E7C11A9.BF345AA4@vl.videotron.ca... > :> > question: > :>  
 > :> [...] > :>  O > :> > Will EFI (the vanilla version stored in ROM) be able to hide drives from ' > :> > NT, or will NT see all devices ?  > :>  H > :> It will see any drive that has a valid partition record, and an EFIQ > :> partition with OS Loaders.  So it will see all the drives, including NT ones  > :> that might be attached. > : H > :So there is definitely a requirement for an (additional) EFI console C > :(ROM!) command to make a drive unavailable for access by any OS:  > :  > :	SET <drive> OFFLINE  > : @ > :and of course the inverse command to make it available again. > F >   If a system operator or system manager really wants to corrupt theF >   disks, there is nothing that we can do and nothing we should do toE >   prevent it.  This is, has been, and will remain the design center C >   -- we cannot protect against a privileged user, and a user with F >   unfettered access to the console hardware is -- by definition -- a >   fully privileged user.  E Correct, but that wasn't my point. I was discussing OSs (or pieces of @ software claiming to be an OS :-) scanning any device physicallyE available and "fixing" its (file) structure if it is "unknown" to the ! OS. (See JF Mezei's posting too.)   J >   As for taking a drive off-line, consider using a StorageWorks "brick".I >   Remove (or uncable) the disk ("brick"), and you have effectively set  I >   the disk drive entirely and quite reliably off-line.  I've found this G >   physical partitioning scheme a most effective approach, personally.   F It may become impractical for large(r) systems and/or "virtual disks"  built by RAID controllers.   > [...]  > E >   If you wish to utilize dual-boot systems with any combinations of J >   operating systems -- I've been reading the Linux and Windows dual-bootF >   materials in recent times -- you will have to know details of the H >   low-level system environment, and you will have to exercise a degreeJ >   of caution.  The existing repartitioning documents are seemingly a seaI >   of warnings around requirements for data backups when repartitioning.   F This "degree of caution" was exactly the reason for my question/demandG -- I'd prefer a concept with a designated (boot) disk for each OS which G can be set offline if another OS should be booted. The very same may be  valid for data disks.    Michael    ------------------------------  # Date: Tue, 25 Mar 2003 09:13:38 GMT A From: "Colin Butcher" <colin_DOT.butcher_AT@xdelta_DOT.co_DOT.uk> 6 Subject: Re: Microvax 3100 dead - need data from drive= Message-ID: <65Vfa.5888$cW7.43925697@news-text.cableinet.net>   J I assume that there are no backups of the dead system available to try andJ recover the data from. This scenario is why regular backups and compatibleG media with other systems you have are useful. I've seen too many people & learn the hard way - including myself!  L Taking the disc drive(s) out of the uVAX3100 seems like a good start. UnlessG you have a modified uVAX3100 (unlikely) then they will be narrow (8bit) J single-ended SCSI drives, probably RZ25s/26s. Don't forget the rules about using anti-static straps etc.   J If possible you should write-protect the disc(s) by adding the appropriateL jumper to the drive configuration jumper block. You may also need to set theL SCSI ID to a different value to avoid conflicts with any SCSI devices in the Alpha.  K I assume that the Alpha is a fairly early machine since it's running V1.5 - @ so it's probably got an 8bit single-ended SCSI controller on theL motherboard. Before you add any devices to the Alpha make sure that you knowD its current configuration (SHOW DEV and SHOW CONF at the >>> console prompt).  K Install the disc(s) into the Alpha and hook them up to the SCSI bus (or add D another SCSI controller to get a second 8bit single-ended SCSI bus).  K Power up the Alpha and see if you can find the added devices at the console I prompt (>>>) with the SHOW DEV command. If not then keep trying until you : have it figured out - or get someone to come and help you.  I Once you have the devices visible at the >>> SHOW DEV level then boot the ; Alpha and try to mount the target disc(s) using MOUNT <dev> I /OVERRIDE=ID/NOWRITE You should then be able to read the disc(s) - unless L the file structure is toast. First thing you do is then make an image backupI of the disc(s) - BACKUP <dev>/IMAGE/NOALIAS <saveset>/VERIFY - personally J I'd go for a disc to disc backup if there's sufficient space on the Alpha, otherwise use tape.   J If the disc(s) won't mount then you're going to have to figure it out - or! get someone to come and help you.   I The level of aggravation and time you'll spend has to be balanced against J sending for professional help - depends how much the data is worth to yourK company. If you're not thoroughly familiar with these things then please be H aware that you can make things worse rather than better and that you canJ make life more difficult (or even impossible) for someone who does come to' help you when you finally ask for help.    -- Hope this helps. Cheers, Colin. ' (colinDOT.butcherAT@xdeltaDOT.coDOT.uk)    ------------------------------    Date: 25 Mar 2003 09:08:39 -0800( From: rickaldred@yahoo.com (Rick Aldred)6 Subject: Re: Microvax 3100 dead - need data from drive< Message-ID: <4c6c3ca.0303250908.6c05cffb@posting.google.com>   "Colin Butcher" <colin_DOT.butcher_AT@xdelta_DOT.co_DOT.uk> wrote in message news:<65Vfa.5888$cW7.43925697@news-text.cableinet.net>...L > I assume that there are no backups of the dead system available to try andL > recover the data from. This scenario is why regular backups and compatibleI > media with other systems you have are useful. I've seen too many people ( > learn the hard way - including myself!  C How right you are! No backups at all. We had hoped that the VAX was 8 now totally gone as a supported platform at our company.  M > I assume that the Alpha is a fairly early machine since it's running V1.5 - B > so it's probably got an 8bit single-ended SCSI controller on theN > motherboard. Before you add any devices to the Alpha make sure that you knowF > its current configuration (SHOW DEV and SHOW CONF at the >>> console
 > prompt).  F I took your advice and discovered to my chagrin that the VAX drive hadF the same SCSI ID as the boot drive on the Alpha! Doh! So I reset it to
 SCSI ID 5.   M > Install the disc(s) into the Alpha and hook them up to the SCSI bus (or add F > another SCSI controller to get a second 8bit single-ended SCSI bus). > M > Power up the Alpha and see if you can find the added devices at the console K > prompt (>>>) with the SHOW DEV command. If not then keep trying until you < > have it figured out - or get someone to come and help you.   $ sho dev d   F Device                  Device           Error    Volume         Free 	 Trans Mnt F  Name                   Status           Count     Label        Blocks	 Count Cnt . ALPHA$DKA0:             Online               0F ALPHA$DKA300:           Mounted              0  ALPHAVMS015      17643	   258   1 . ALPHA$DKA500:           Online               1 $ mount dka500: /over=id" %MOUNT-F-DRVERR, fatal drive error  $ mount dka500: /foreign/over=id" %MOUNT-F-DRVERR, fatal drive error $ sho dev dka500: /full   = Disk ALPHA$DKA500:, device type Generic SCSI disk, is online, 
 file-oriented 0     device, shareable, error logging is enabled.  F     Error count                    1    Operations completed                 10F     Owner process                 ""    Owner UIC                         [0,0]3     Owner process ID        00000000    Dev Prot     S:RWED,O:RWED,G:RWED,W:RWED F     Reference count                0    Default buffer size                 512    $ mount dka500: /over=id/nowrite" %MOUNT-F-DRVERR, fatal drive error  C I keep getting the fatal drive error. It appears that this drive is E toast. Unless there are other fairly easy things that I can safely do F from my end, I'd say this rescue attempt will be aborted. Professional@ data recovery, with associated cost, will probably be judged not worthwhile.   K > Once you have the devices visible at the >>> SHOW DEV level then boot the = > Alpha and try to mount the target disc(s) using MOUNT <dev> K > /OVERRIDE=ID/NOWRITE You should then be able to read the disc(s) - unless N > the file structure is toast. First thing you do is then make an image backupK > of the disc(s) - BACKUP <dev>/IMAGE/NOALIAS <saveset>/VERIFY - personally L > I'd go for a disc to disc backup if there's sufficient space on the Alpha, > otherwise use tape.  > L > If the disc(s) won't mount then you're going to have to figure it out - or# > get someone to come and help you.  > K > The level of aggravation and time you'll spend has to be balanced against L > sending for professional help - depends how much the data is worth to yourM > company. If you're not thoroughly familiar with these things then please be J > aware that you can make things worse rather than better and that you canL > make life more difficult (or even impossible) for someone who does come to) > help you when you finally ask for help.    ------------------------------  % Date: Tue, 25 Mar 2003 12:42:30 -0600 / From: Chris Scheers <chris@applied-synergy.com> 6 Subject: Re: Microvax 3100 dead - need data from drive3 Message-ID: <3E80A316.E75FC211@applied-synergy.com>    Rick Aldred wrote: >  > "Colin Butcher" <colin_DOT.butcher_AT@xdelta_DOT.co_DOT.uk> wrote in message news:<65Vfa.5888$cW7.43925697@news-text.cableinet.net>...N > > I assume that there are no backups of the dead system available to try andN > > recover the data from. This scenario is why regular backups and compatibleK > > media with other systems you have are useful. I've seen too many people * > > learn the hard way - including myself! > E > How right you are! No backups at all. We had hoped that the VAX was : > now totally gone as a supported platform at our company. > O > > I assume that the Alpha is a fairly early machine since it's running V1.5 - D > > so it's probably got an 8bit single-ended SCSI controller on theP > > motherboard. Before you add any devices to the Alpha make sure that you knowH > > its current configuration (SHOW DEV and SHOW CONF at the >>> console > > prompt). > H > I took your advice and discovered to my chagrin that the VAX drive hadH > the same SCSI ID as the boot drive on the Alpha! Doh! So I reset it to > SCSI ID 5. > O > > Install the disc(s) into the Alpha and hook them up to the SCSI bus (or add H > > another SCSI controller to get a second 8bit single-ended SCSI bus). > > O > > Power up the Alpha and see if you can find the added devices at the console M > > prompt (>>>) with the SHOW DEV command. If not then keep trying until you > > > have it figured out - or get someone to come and help you. > 
 > $ sho dev d  > G > Device                  Device           Error    Volume         Free  > Trans Mnt H >  Name                   Status           Count     Label        Blocks > Count Cnt 0 > ALPHA$DKA0:             Online               0H > ALPHA$DKA300:           Mounted              0  ALPHAVMS015      17643 >   258   1 0 > ALPHA$DKA500:           Online               1 > $ mount dka500: /over=id$ > %MOUNT-F-DRVERR, fatal drive error" > $ mount dka500: /foreign/over=id$ > %MOUNT-F-DRVERR, fatal drive error > $ sho dev dka500: /full  > ? > Disk ALPHA$DKA500:, device type Generic SCSI disk, is online,  > file-oriented 2 >     device, shareable, error logging is enabled. > > >     Error count                    1    Operations completed
 >       103 >     Owner process                 ""    Owner UIC 
 >    [0,0]2 >     Owner process ID        00000000    Dev Prot > S:RWED,O:RWED,G:RWED,W:RWED = >     Reference count                0    Default buffer size 
 >      512 > " > $ mount dka500: /over=id/nowrite$ > %MOUNT-F-DRVERR, fatal drive error > E > I keep getting the fatal drive error. It appears that this drive is G > toast. Unless there are other fairly easy things that I can safely do H > from my end, I'd say this rescue attempt will be aborted. ProfessionalB > data recovery, with associated cost, will probably be judged not
 > worthwhile.      There is one other possibility.   H VMS 6.2 is fairly tolerant of the SCSI error handling bits on the drive.  E OTOH, I don't know about Alpha VMS 1.5.  I believe that Alpha VMS 1.5 G was based on the VAX/VMS 5.4 code base.  If it copied the SCSI handling E behavior of the VAX 5.4 driver, it is very picky about the SCSI error  handling on the drive.  C On VAX/VMS 5.4, if you try to mount a drive that has the SCSI error F correction bits set to a value that the driver does not like, you will$ get the "fatal drive error" message.   What kind of drive is this?   ; Can you try it on a VMS (VAX or Alpha) 6.2 or later system?   2 It is possible that the data is still recoverable.  G ----------------------------------------------------------------------- $ Chris Scheers, Applied Synergy, Inc.  C Voice: 817-237-3360            Internet: chris@applied-synergy.com     Fax: 817-237-3074    ------------------------------    Date: 25 Mar 2003 01:16:51 -0800* From: b.seghers@europe.com (Seghers Bruno) Subject: Minimum version= Message-ID: <1fe424ce.0303250116.105d7fa9@posting.google.com>    Hi,   A Could somebody tell me what is the minimum version requiered, the @ lasted version and the recommended one of CXX on a OpenVMS 7.3-1 system.    Thanks for your help  
 Seghers Bruno    ------------------------------  # Date: Tue, 25 Mar 2003 14:26:03 GMT , From: "James Gessling" <jgessling@yahoo.com> Subject: Re: Minimum versionA Message-ID: <%FZfa.3339$mh1.135758914@newssvr14.news.prodigy.com>   B Not completely sure.  I've been using V6.2-016, but I see there is5 6.5 available.  6.4 added 64 bit support if you care.   G As an aside, it sure would be nice if the C and C++ links were included : on the openvms doc page ( http://h71000.www7.hp.com/doc/ )@ where other languages are (COBOL and BASIC) instead of having to chase around to find them.   Jim   7 "Seghers Bruno" <b.seghers@europe.com> wrote in message 7 news:1fe424ce.0303250116.105d7fa9@posting.google.com...  > Hi,  > C > Could somebody tell me what is the minimum version requiered, the B > lasted version and the recommended one of CXX on a OpenVMS 7.3-1	 > system.  >  > Thanks for your help >  > Seghers Bruno    ------------------------------  % Date: Tue, 25 Mar 2003 11:04:30 +0100 0 From: "labadie" <en_trajectant_a_mort@127.0.0.1>  Subject: Re: MSIE ftp hangs FTPD* Message-ID: <3e802bc7@usenet01.boi.hp.com>  5 "Alder" <TFTAJLLYMXZP@spammotel.com> wrote in message 6 news:a6840bf2.0303241320.1bc9a19@posting.google.com...= > "labadie" <en_trajectant_a_mort@127.0.0.1> wrote in message , news:<Ht07a.431$fm4.339@news.cpqcorp.net>... > 2 > > Your Tqelm is too low for the ucx$ftp account.0 > > Your service limit is 50 for the TP service.L > > If you really want to allow 50 simultaneous FTP transfers at a time, you
 > > shouldF > > $ mc authorize mod ucx$ftp/tqelm=100/bytlm=500000/pgflquota=200000 > > then stop and restart Ftp. > >  > A > Where would I look to find the general rules about tuning these % > account settings for anonymous FTP?  >  > Alder    Hello   3 This is a good question, I thought I had read it on L http://h18000.www1.hp.com/support/asktima/communications/009D3461-42C5BBE0-1
 C0096.html  = [TCP/IP] Tuning Hints to Improve UCX V3.0 to V4.2 Performance   8 (a little outdated for Tcpip V5, but very good document)   but it is not here.    The doc = Compaq TCP/IP Services for OpenVMS Tuning and Troubleshooting L http://h71000.www7.hp.com/doc/73final/6631/6631pro_contents.html#toc_chapter _1   is not about that very subject.   * I had done some tests with Tcpip 5.3 Eco 1- the Ftp process alone (just started) consumes  196 096 in Bytlm 8512 in Pgflquota 
 1 in Tqelm   each other user will need  about 66 000 in Bytlm  128 in pgflquota at least 1 in tqelm   ' Run you own tests, using Amds or pquota 5 http://h71000.www7.hp.com/freeware/freeware40/pquota/    Regards    Grard   ------------------------------  % Date: Tue, 25 Mar 2003 02:43:30 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>- Subject: Re: Netscape gags on VMS FTP Servers / Message-ID: <3E7FFA78.935255F5@vl.videotron.ca>    Alder wrote: > Q > Certain VMS-based FTP sites (my own among them, running TCPIP Services 5.3) are N > responding with the following message when I connect with Netscape Navigator& > 7.01 (Windows) using the FTP scheme:  . VAX VMS 7.2, TCPIP 5.3.  MAC with Netscape 4.7   No problem here.  N However, I do have the logical TCPIP$FTP_NO_VERSION defined to "1" which givesL a "clearer" listing, albeit no wait o obtain previous versions of documents,G and no ability to delete them unless to type in the fill file name witho# extension for the delete operation.    ------------------------------    Date: 25 Mar 2003 11:10:27 +0100' From: huber@mppmu.mpg.de (Joseph Huber)r- Subject: Re: Netscape gags on VMS FTP Serversr+ Message-ID: <EOp5cpAnavjS@vms.mppmu.mpg.de>i  W In article <3E7FDCA0.6020801@spammotel.com>, Alder <PGDEHMKOKIMD@spammotel.com> writes:fR > Certain VMS-based FTP sites (my own among them, running TCPIP Services 5.3) are O > responding with the following message when I connect with Netscape Navigator s& > 7.01 (Windows) using the FTP scheme: >  > AlertsR > The FTP server 200 VMS OpenVMS V7.3 on node <ftp site> is currently unsupported. > S > Does anyone get similar behaviour using VMS Mozilla clients, or is this (likely) i; > a Netscape for Windows problem as I'm tempted to believe?n >   Q I have given up to use the VMS TCPIP server because of the different assumptions e8 different clients have about (mis-)handling VMS servers.N I have excellent experience with Hunter Goatleys HGFTP server: it can be setupM to completely appear as a Unix server, and then works even with most Windows w' and macintosh clients, Mozilla as well.o  p --  N Joseph "Sepp" Huber   mailto:joseph.huber@web.de   http://www.huber-joseph.de/   ------------------------------  % Date: Tue, 25 Mar 2003 10:51:20 -0000G* From: "Richard Brodie" <R.Brodie@rl.ac.uk>- Subject: Re: Netscape gags on VMS FTP Serversr+ Message-ID: <b5pcb9$luc@newton.cc.rl.ac.uk>S  \ "Alder" <PGDEHMKOKIMD@spammotel.com> wrote in message news:3E7FDCA0.6020801@spammotel.com...  R > The FTP server 200 VMS OpenVMS V7.3 on node <ftp site> is currently unsupported.R > Does anyone get similar behaviour using VMS Mozilla clients, or is this (likely); > a Netscape for Windows problem as I'm tempted to believe?   P No. It's in the Mozilla code. http://bugzilla.mozilla.org/show_bug.cgi?id=151501Q Some FTP servers can identify as Unix and provide a Unix-like directory format. Io! don't think TCPIP is one of them.-   ------------------------------    Date: 25 Mar 2003 07:29:53 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)t- Subject: Re: Netscape gags on VMS FTP ServersS3 Message-ID: <04LalNvSWhbV@eisner.encompasserve.org>   W In article <3E7FDCA0.6020801@spammotel.com>, Alder <PGDEHMKOKIMD@spammotel.com> writes: R > Certain VMS-based FTP sites (my own among them, running TCPIP Services 5.3) are O > responding with the following message when I connect with Netscape Navigator e& > 7.01 (Windows) using the FTP scheme: >  > Alert.R > The FTP server 200 VMS OpenVMS V7.3 on node <ftp site> is currently unsupported. > S > Does anyone get similar behaviour using VMS Mozilla clients, or is this (likely) e; > a Netscape for Windows problem as I'm tempted to believe?   E    I can't recall seeing this on mozilla.  Ive seen it on Netscape intG    the past and safely ignored it (sorry, can't recall which Netscape).t    t   ------------------------------  % Date: Tue, 25 Mar 2003 09:00:07 +0100r6 From: =?iso-8859-1?Q?Jean=2DFran=E7ois=20PI=C9RONNE?= - Subject: Re: new releases of MySQL and Pythono+ Message-ID: <3E800C87.A015B586@laposte.net>c   > N > I have tried to build 4.0.12 had end up with the following error in a log of > modules that are linked. > ---->D' > %LINK-W-NUDFSYMS, 1 undefined symbol:c0 > %LINK-I-UDFSYM,         TCPIP$GET_H_ERRNO_ADDRF > %LINK-W-USEUNDEF, undefined symbol TCPIP$GET_H_ERRNO_ADDR referenced+ >         in psect $LINK$ offset %X00000040 ) >         in module MY_GETHOSTBYNAME file * > MYSQL_ROOT:[VMS.MYSQL.O_ALPHA.MYSYS]MY_G > ETHOSTBYNAME.OBJ;1 > <----oF > There are also a bunch of undifened symbols in OS0FILE and LOCK0LOCK
 > modules.) > Any idea how to resolve these warnings?n1 > Will they be a problem when running the images?C > Details follow., >     6 We have only test on VMS 7.3 and 7.3-1 using Tcpip 5.3 So Tcpip 5.1 may be to old. P I have receive mails about successful build but I don't know which Tcpip version	 was used.-   Thanks for testing.2  
 Jean-Franois2   ------------------------------  # Date: Tue, 25 Mar 2003 15:36:55 GMTt; From: "John Gemignani, Jr." <jon-nope@thiswontworkossc.net>.- Subject: Re: new releases of MySQL and Python:< Message-ID: <rI_fa.9156$TW2.1914964@news1.news.adelphia.net>  = "David Reynolds" <datre@bellnospamsouth.net> wrote in message 2 news:eRQfa.3551$hz3.828@fe06.atl2.webusenet.com...K > I have tried to build 4.0.12 had end up with the following error in a logt of > modules that are linked. > ----> ' > %LINK-W-NUDFSYMS, 1 undefined symbol:n0 > %LINK-I-UDFSYM,         TCPIP$GET_H_ERRNO_ADDRF > %LINK-W-USEUNDEF, undefined symbol TCPIP$GET_H_ERRNO_ADDR referenced+ >         in psect $LINK$ offset %X00000040d) >         in module MY_GETHOSTBYNAME filep* > MYSQL_ROOT:[VMS.MYSQL.O_ALPHA.MYSYS]MY_G > ETHOSTBYNAME.OBJ;1 > <----sF > There are also a bunch of undifened symbols in OS0FILE and LOCK0LOCK
 > modules.) > Any idea how to resolve these warnings? 1 > Will they be a problem when running the images?- > Details follow.- >- > Thanks > David Reynolds > Gallatin Steel Co. > Process Automation >o
 > ENVIRONMENTj! > DEV_MSCL2> sho sys /head/noprocsL > OpenVMS V7.2-1  on node MSCL2  24-MAR-2003 23:12:58.40  Uptime  5 00:13:25 > DEV_MSCL2> tcpip sho ver >hA >   Compaq TCP/IP Services for OpenVMS Alpha Version V5.1 - ECO 2D: >   on a AlphaServer 1200 5/400 4MB running OpenVMS V7.2-1 >h > DEV_MSCL2> cc /ver+ > Compaq C V6.4-008 on OpenVMS Alpha V7.2-1o > DEV_MSCL2> cxx /ver . > Compaq C++ V6.5-004 for OpenVMS Alpha V7.2-1 >v  > The following lines appeare in9 > MYSQL_ROOT:[VMS.MYSQ.O_ALPHA.MYSYS]MY_GETHOSTBYNAME.MMS J > I suspect this is where part of the problem is, when the *.mms file gets > generated. >s > MY_GETHOSTBYNAME.OBJ : >sL MSCL2$DKB0:[SYS0.SYSCOMMON.][SYSHLP.EXAMPLES.TCPIP]__DECC_INCLUDE_PROLOGUE.H > MY_GETHOSTBYNAME.OBJ :* SYS$COMMON:[SYSHLP.EXAMPLES.TCPIP]SOCKET.H >tL > This doesn't make sense why one line resolves SYS$COMMON but the next line	 > doesn'tw > DEV_MSCL2> SHO LOG SYS$COMMONrE >    "SYS$COMMON" = "MSCL2$DKB0:[SYS0.SYSCOMMON.]" (LNM$SYSTEM_TABLE)  > $ > References to the undefined symbol >w0 > DEV_MSCL2> sear tcpip$examples:*.* get_h_errno >h  > ******************************- > SYS$COMMON:[SYSHLP.EXAMPLES.TCPIP]NETDB.H;1r >t( >     int *TCPIP$GET_H_ERRNO_ADDR(void);. > #   define h_errno *TCPIP$GET_H_ERRNO_ADDR() >4  > ******************************. > SYS$COMMON:[SYSHLP.EXAMPLES.TCPIP]RESOLV.H;1 >S$ > int *TCPIP$GET_H_ERRNO_ADDR(void);+ > #define h_errno *TCPIP$GET_H_ERRNO_ADDR()n >o >e >m > E > "Jean-Franois PIRONNE" <jf.pieronne@laposte.net> wrote in messagei' > news:3E7991D3.DE578D9D@laposte.net...e > >  > > >g* > > > Some files have ODS5-specifications. > > >_ > > > config^.h.in > > > mysql_version^.h.inD > > > ib_config^.h.inP > >i; > > They are not necessary, I have removed all these files. E > > I have noticed that MySQL 4.0.12 was released, I have ported this  version, > a newn > > kit is online. > >c > >2 > > >e2 > > > I'm trying to compile with C 6.5 C++ 6.5 andE > > > Multinet and this gives me some unresolved references. I'm justs > producingc& > > > all listing to see what's wrong. > >nL > > Don't know if this can work with Multinet, we have used TCPIp 5.3, but I > am not+ > > aware of the use any specific routines.  > >l > > Jean-Franoisv >s >  >l  J Look in TCPIP$LIBRARY:, you should find a TCPIP$LIB.OLB file (or somethingL close).  The routine you're looking for is probably in there.  I know that IH did this original work, but don't recall the details right now and can't check as I'm not in the office.i   -Johna   ------------------------------  # Date: Tue, 25 Mar 2003 17:55:57 GMT < From: "John E. Malmberg" <Malmberg@dskwld.zko.dec.compaq.hp>- Subject: Re: new releases of MySQL and Pythonn/ Message-ID: <NK0ga.803$Rv1.16@news.cpqcorp.net>e   John Gemignani, Jr. wrote:  ? > "David Reynolds" <datre@bellnospamsouth.net> wrote in message 4 > news:eRQfa.3551$hz3.828@fe06.atl2.webusenet.com... > K >>I have tried to build 4.0.12 had end up with the following error in a logf >>of modules that are linked.i >>---->,' >>%LINK-W-NUDFSYMS, 1 undefined symbol:h0 >>%LINK-I-UDFSYM,         TCPIP$GET_H_ERRNO_ADDRF >>%LINK-W-USEUNDEF, undefined symbol TCPIP$GET_H_ERRNO_ADDR referenced+ >>        in psect $LINK$ offset %X00000040t) >>        in module MY_GETHOSTBYNAME file3* >>MYSQL_ROOT:[VMS.MYSQL.O_ALPHA.MYSYS]MY_G >>ETHOSTBYNAME.OBJ;1  G Your link command indicates that you are not using the gethostbyname() t; routine from DECC$SHR or the TCPIP$LIB.OLB for some reason.-  L > Look in TCPIP$LIBRARY:, you should find a TCPIP$LIB.OLB file (or somethingN > close).  The routine you're looking for is probably in there.  I know that IJ > did this original work, but don't recall the details right now and can't  > check as I'm not in the office  ? The current 7.3-1 DECC$SHR has definitions for two variants of u' gethostbyname(), one is a BSD44 flavor.i  E I do not know if a C RTL ECO kit will add the routines to your older o release of OpenVMS.T  E There is also a gethostbyname routine in TCPIP$LIBRARY:TCPIP$LIB.OLB.   ; As John suggested, to resolve that symbol, you can add the  G TCPIP$LIBRARY:TCPIP.OLB/LIB to your link command line as a work around.l  < Then you should not need the MY_GETHOSTBYNAME module either.  H The quoting in this thread has gotten a bit confused.  I think that the E TCPIP$LIBRARY option will only work if you have the HP TCPIP product  
 installed.  F For help linking with Multinet, you will need to contact them.  Their H developers sometimes post here, and there is also a vmsnet newgroup for * Multinet.  vmsnet.networks.tcp-ip.multinet  F >>There are also a bunch of undifened symbols in OS0FILE and LOCK0LOCK
 >>modules.) >>Any idea how to resolve these warnings?.  H Search the source code and the user and system supplied header files to  determine how they are used.  I Sometimes just changing an option in CONFIG.H can make these problems go  G away, or you may not have compiled or linked in a module that you need.t  1 >>Will they be a problem when running the images?d  	 Probably.i     -Johno! malmberg@dskwld.zko.dec.compaq.hp  Personal Opinion Onlyw   ------------------------------    Date: 25 Mar 2003 07:22:18 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) 9 Subject: Re: next installment - alpha 3000 console now...u3 Message-ID: <qA0WFnTmvkrC@eisner.encompasserve.org>.  g In article <a7234bb1.0303241858.576b31e8@posting.google.com>, timasmith@hotmail.com (Tim Smith) writes:o > F > hmmm, more useless info: I can't find a 'S3 switch' anywhere and theH > one DEC 3000 document I found is of a short workstation whereas my DECF > Server 3000 is a tower shaped and has little on the back - though it@ > does have that funny shaped port perhaps MMJ which I should be > plugging in a VT Terminal? >  > Tim   H    The console will certainly be the MMJ, not a DB style connector.  DECG    made MMJ to female DB25 and MMJ to male DB25 converters.  Folks like %    Black Box used to carry them, too.n   ------------------------------  # Date: Tue, 25 Mar 2003 17:41:07 GMTo4 From: "Mark Buda" <buda@tabasco.zko.dec.no.spam.com>) Subject: Re: OpenVMS SMP License Questionn0 Message-ID: <Tw0ga.802$BC1.285@news.cpqcorp.net>  9 "Keith A. Lewis" <lewis@mazda.mitre.org> wrote in message ( news:b5ci1v$9nr$1@newslocal.mitre.org...G > If you install the CPUs before the licenses, the system will boot butU in aH > crippled mode with no clustering or network services.  But you can log in$ > from the console and check it out.   Hello Keith,  G Can you explain what you mean by saying "...crippled with no clusteringy
 services..."??   mark   ------------------------------  % Date: Tue, 25 Mar 2003 11:58:00 +0000h' From: Andrew Harrison SUNUK ConsultancyaY Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retai . Message-ID: <3E804448.6050908@nospamn.sun.com>   jlsue wrote:H > NOTE:  I know the attribution below may not be exactly correct....  inG > switching ISPs I missed some notes and had to look them up on Google.u > G > On Tue, 13 Mar 2003 09:12:05 +0000, Andrew Harrison SUNUK Consultancy,0 > <Andrew_No.Harrison_No@nospamn.sun.com> wrote: >  > F >>>I don't divulge any details about my customer.  However, there is aI >>>web page for HP wins, and one that's kinda neat is at Bank of Austria.S= >>>They have 16-P GS160 systems.  check out the report below:t >>>t< >>>http://h71000.www7.hp.com/openvms/brochures/Bank-Austria/ >>>e >> > 4 >>Great another example that proves my point thanks. >>7 >>The Turbo Laser was the codename for the 8400 so yourn8 >>customer got a 28% performance improvement moving from* >>an 8400 to a GS160 with 16 CPUs and 7.3. >>blah, blah, blah...L >  > A > You still completely ignore the point in order to continue your  > worthless rants. >       C > I never made claims about the performance in specific numbers, son0 > you're wasting your time with that argument.   >   ; No you made general claims, which you have failed miserably ! to back up with specific numbers.   : When are you going to work it out, Compaq have published a8 wealth of benchmark results that show that the GS160/320; don't perform to the levels claimed by your marketing, your'; own customers have also provided numbers for their own appse that show this.   8 You have provided no examples that refute this so in the9 circumstances unless you can I am perfectly free to claima9 with supporting collateral that GS160/320's don't perform 9 and until you come up with something to support your caset$ you are not in a possition to reply.  5 Get used to it and try being a bit more adult in youro5 responses, accusing me of lying isn't going to assist 
 your case.  " > But you're lieing in any event:  >   
 Oh really.  E > The article DOES NOT say that the GS160 gave them a 28% performanceNC > increase.  Try reading for comprehension next time.  It says thatLG > "moving to VMS 7.3 gave them a 28% performance increase.  The articlebF > actually does not give any info as to the performance gains from theC > GS160.  This is yet another attempt on your part to mis-represent-A > things to try to back up your unsupportable logical conclusion.  >   D You need to read the small print down towards the end of the article and not the headline.3  ? The article talks about two events, an upgrade to the GS and ansB upgrade to 7.3 but only refers to one 28% performance improvement.     d amount to fud. > G > You still can't prove that the GS160/320 does not perform well, ever. H > You can't even say how often it performs well.  In one of the examplesC > given, and which you conveniently ignore as a data point, anotherlG > typical tactic of yours, Oracle Financials applications on GS160 wereaA > the PREFERRED solution.  Yep, those same transaction-processingaC > applications that you swear never perform well on the GS160.  Whyt > don't you address that one?  >   H No wrong, you seem unable to accept the facts that your own benchmarkingD people have presented by publishing the banchmarks they have for the GS.y  C We know that its OLTP performance is bad, you had to do TPC-C using B a cluster in a box and even then the result wasn't competitive. InF addition it also performed very badly for Oracle Applications standard# benchmark a more complex OLTP test.   
 SAP ditto.   Kingston Ditto   Data warehousing, TPC-H ditto4  E And then we have the revelation from your own engineers that anythinggF with multiple processes sharing one global shared memory pool are alsoB likely to perform badly. Plus your customer newsgroup postings and+ now the apparent Bank of Austria reference.e  B So what you have to contribute to the argument on your side exceptH a wealth of handwavey non fact. Sorry but you have failed and failed andB failed to make you case, how long are you going to keep this up ??      B > Oh, but it's so much easier to pick-and-choose your data points,F > including repeating (for the umpteen millionth time) the TP-whateverH > benchmarks.  If your customers make money by running these benchmarks,E > good for them.  Ours typically run other apps, some of which mirror ? > the results of the benchmarks, but certainly not all of them.r >   D I havn't picked and chosen, TPC-C, Oracle Apps, SAP, TPC-C, KingstonF are practically the sum of all the public benchmarks for the GS except SPEC.     G > The FACT is that I have shown TWO examples where customers were happyr? > with their "wildfire" systems.  This flies in the face of theiC > contention that you keep repeating (ad nauseum) in here that theyr > can't peform well. >     A No you havn't you have provided one "reference" that turns out toID be just embarassing and handwavey Oracle Financials reference, whichA in the light of your Oracle Applications performance tests soundstA like it may be deal that has nothing to do with how fast the GS's. are if it exists at all.   Regardsi Andrew Harrisong   ------------------------------  # Date: Tue, 25 Mar 2003 19:09:29 GMT & From: jlsue <jefflsxxxz@sbcglobal.net>Y Subject: Re: OpenVMS.org: Marvel article and HP's press release for Marveland Alpha Retaic8 Message-ID: <q9618v8hmoa1hqs1rlb9phimbajsnl85sv@4ax.com>  E On Tue, 25 Mar 2003 11:58:00 +0000, Andrew Harrison SUNUK ConsultancyI. <Andrew_No.Harrison_No@nospamn.sun.com> wrote:   >i >e
 >jlsue wrote:eI >> NOTE:  I know the attribution below may not be exactly correct....  ineH >> switching ISPs I missed some notes and had to look them up on Google. >>  H >> On Tue, 13 Mar 2003 09:12:05 +0000, Andrew Harrison SUNUK Consultancy1 >> <Andrew_No.Harrison_No@nospamn.sun.com> wrote:  >> t >> iG >>>>I don't divulge any details about my customer.  However, there is a J >>>>web page for HP wins, and one that's kinda neat is at Bank of Austria.> >>>>They have 16-P GS160 systems.  check out the report below: >>>>= >>>>http://h71000.www7.hp.com/openvms/brochures/Bank-Austria/t >>>> >>>l >> b >eD >> I never made claims about the performance in specific numbers, so1 >> you're wasting your time with that argument.  0 >> 3 >M< >No you made general claims, which you have failed miserably" >to back up with specific numbers.  C No, I made claims that YOUR general claims are unsupportable in the-5 real world.  If I made general claims, show me where.u   > ; >When are you going to work it out, Compaq have published aT9 >wealth of benchmark results that show that the GS160/320i< >don't perform to the levels claimed by your marketing, your< >own customers have also provided numbers for their own apps >that show this.  F When in doubt, always quote these benchmark results.  That way you can ignore real-world results.   >w9 >You have provided no examples that refute this so in the': >circumstances unless you can I am perfectly free to claim: >with supporting collateral that GS160/320's don't perform: >and until you come up with something to support your case% >you are not in a possition to reply.I >o  B I have supplied examples that disprove your claims that they don't> perform well, and in particular, your generalizations that allD "transacton-oriented" applications will perform badly.  YOU made theC general claims, so all I've got to do is show counter examples thata disprove the generality.  6 >Get used to it and try being a bit more adult in your6 >responses, accusing me of lying isn't going to assist >your case.h  = That's gotta be another for the pot-and-kettle award.  But tosA demonstrate that you are talking out of the wrong end, see below:    >a >iF >> The article DOES NOT say that the GS160 gave them a 28% performanceD >> increase.  Try reading for comprehension next time.  It says thatH >> "moving to VMS 7.3 gave them a 28% performance increase.  The articleG >> actually does not give any info as to the performance gains from thecD >> GS160.  This is yet another attempt on your part to mis-representB >> things to try to back up your unsupportable logical conclusion. >> u > E >You need to read the small print down towards the end of the articleA >and not the headline. >a@ >The article talks about two events, an upgrade to the GS and anC >upgrade to 7.3 but only refers to one 28% performance improvement.e  E Yes, you are correct, it does refer to both activities.  However, thep+ EXACT quote from the article is as follows:o  ; 	"Since moving to version 7.3 of OpenVMS, we have seen a 28l  	 percent performance increase."  F Now, how do you get the 28% applied to the GS upgrade without a lot ofD blind faith, jumping to conclusions, hope-against-hope, and fud?  ItF specifically says that the 28% reference is to the VMS upgrade, and no' GS160 performance numbers are provided.e  E This is just another rathole you've started to deflect the discussionlD from the real topic:  disproving that these systems perform well and that we have happy customers.-   > blah... blah... blah >cH >> The FACT is that I have shown TWO examples where customers were happy@ >> with their "wildfire" systems.  This flies in the face of theD >> contention that you keep repeating (ad nauseum) in here that they >> can't peform well.W >> 2 >7 >wB >No you havn't you have provided one "reference" that turns out toE >be just embarassing and handwavey Oracle Financials reference, whichrB >in the light of your Oracle Applications performance tests soundsB >like it may be deal that has nothing to do with how fast the GS's >are if it exists at all.o >i  E Hmm... I take what's said in the article and report it.  You take alllC kinds of info from completely unrelated sources, including your own D hallucinations, and attempt to discredit the stated fact that the GSC series performs very well for their Oracle Financials apps and theyr! are standardizing on that config.e  E If you have evidence that they are NOT happy with the systems, please < provide that.  If not, then you are only engaging in wishful conjecture.e   ------------------------------  % Date: Tue, 25 Mar 2003 09:30:10 -0000o% From: "Tobe" <tarkless@wilco-int.com>I Subject: Restore with /selects/ Message-ID: <ImVfa.98$1u.1453@news.uk.colt.net>    Hi,   I Sorry if this has come up before I have done many searches but can't find  and answer to my scenario.  I I am trying to do a selective restore from an image backup. However everyH= time I have tried it seems to want to restore the whole disk.i  K I have taken to trying out various options with backup/list but seem to get  the same result.   Here is an example:u  C backup/list mkb600:disk5.bac/save_set/select=[cmpdev.exane.0dat]*.*C   Listing of save set(s)  * %MOUNT-I-WRITELOCK, volume is write locked3 %MOUNT-I-MOUNTED, A-C-TH mounted on _CSLA02$MKB600:  Save set:          DISK5.BAC Written by:        SYSTEM5" UIC:               [000001,000004]* Date:              20-MAR-2003 22:25:02.14 Command:= BACKUP/IMAGE/IGNORE=INTERLOCK/NOASSIST/LABEL=A-C-TH/EXCLUDE=Rt2 ESERVE.SYS* CSLA02$DKA500: CSLA02$MKB600:DISK5.BAC+ Operating system:  OpenVMS AXP version V6.2  BACKUP version:    V6.2  CPU ID register:   80000000l Node name:         _CSLA02::" Written on:        _CSLA02$MKB600: Block size:        8192  Group size:        10r Buffer count:      217   Image save of volume set Number of volumes: 1   Volume attributes  Structure level:   2 Label:             A02DATA_4 Owner:" Owner UIC:         [000001,000004]* Creation date:     30-MAY-1998 10:16:37.30 Total blocks:      17773524w Access count:      3 Cluster size:      18 $ Data check:        No Read, No Write Extension size:    5< File protection:   System:RWED, Owner:RWED, Group:RE, World: Maximum files:     467724nB Volume protection: System:RWCD, Owner:RWCD, Group:RWCD, World:RWCD Windows:           7  J [000000]000000.DIR;1                                        2  30-MAY-1998 10:16iJ [000000]AIB.DIR;1                                           1   1-MAR-2002 10:31iJ [AIB]DELIVERY_ALPHA.DIR;1                                   1   1-MAR-2002 10:31eJ [000000]BACKUP.SYS;1                                        0  30-MAY-1998 10:16aJ [000000]BADBLK.SYS;1                                        0  30-MAY-1998 10:16rJ [000000]BADLOG.SYS;1                                        0  30-MAY-1998 10:16hJ [000000]BGL_TEST.DIR;1                                      8   3-DEC-1998 11:26aJ [BGL_TEST]0DAT.DIR;1                                       16   3-DEC-1998 14:33aJ [BGL_TEST.0DAT]018RISK.RULESUM7A;1                          4  26-MAY-2000 10:14iJ [BGL_TEST.0DAT]BG5TAB.;1                                   18  15-SEP-1999 13:31oJ [BGL_TEST.0DAT]BGLIMT.;1                                   54  15-MAR-1991 16:33sJ [BGL_TEST.0DAT]BGTEMP.;1                                  153  15-FEB-1996 17:40oJ [BGL_TEST.0DAT]BGTEMP2.;1                                   9  28-APR-1998 21:51 J [BGL_TEST.0DAT]BGTRAD.;1                                   72   4-SEP-1998 21:59iJ [BGL_TEST.0DAT]BGTRAN.;1                                  324   6-OCT-2000 00:17 J [BGL_TEST.0DAT]BGVOL1.;1                                  540  26-JUL-1999 19:02tJ [BGL_TEST.0DAT]BGVOL6.;1                                 2043  13-MAY-2000 10:00aJ [BGL_TEST.0DAT]BONGO.BMV;1                                 13   5-OCT-2000 23:41rJ [BGL_TEST.0DAT]CED001.;1                                    1  11-APR-1991 11:21.J [BGL_TEST.0DAT]CM030A.CON;1                                18   8-MAY-1990 18:30yJ [BGL_TEST.0DAT]CM100A.EXT;1                                 1  30-AUG-2001 16:13mJ [BGL_TEST.0DAT]CM100A.SRT;1                                 1  30-AUG-2001 16:13 J [BGL_TEST.0DAT]CM100D.EXT;1                                 4  30-AUG-2001 18:41hJ [BGL_TEST.0DAT]CM100D.SRT;1                                 4  30-AUG-2001 18:41gJ [BGL_TEST.0DAT]CM510A.;1                                  261  25-MAY-1999 19:16 J [BGL_TEST.0DAT]CMACRE.CMD;1                                 1   3-JUL-1990 11:59EJ [BGL_TEST.0DAT]CMBNRE.CMD;1                                 1  27-NOV-1991 09:25oJ [BGL_TEST.0DAT]CMBNRX.CMD;1                                 1  20-SEP-1989 12:36 J [BGL_TEST.0DAT]CMCASH.;7                                80562  27-MAY-2002 13:40e .p .a .h  ! Any ideas would be most grateful.    Toby Arkless	 ADP Wilco    ------------------------------  % Date: Tue, 25 Mar 2003 07:27:36 -0600a From: brandon@dalsemi.coma! Subject: Re: Restore with /selectr1 Message-ID: <03032507273679@dscis6-0.dalsemi.com>e   >  BACKUP version:    V6.2  L I see that the backup was perfromed on VMS V6.2, however what version of VMS are you currently using?  N I know that there were some problems with compatibility with V7.2 and V6.2 (orN there abouts) - that dealt with journal files.  If memory serves me correctly.   That said...  K I do get the same results that you get with V7.2 - so my assumption is thatS- this is a limitation of the design of backup.    John Brandon VMS Systems Administratork Dallas Semiconductor john.brandon@dalsemi.com 972.371.4172 wkk 972.371.4003 fxe   ------------------------------  % Date: Tue, 25 Mar 2003 08:34:45 -0500 & From: Ken Robinson <kenrbnsn1@rcn.com>! Subject: Re: Restore with /selecti< Message-ID: <5.1.0.14.2.20030325083245.058e64d8@pop.rcn.com>  1 At 09:30 AM 3/25/2003 +0000, you wrote (in part):sJ >I am trying to do a selective restore from an image backup. However every> >time I have tried it seems to want to restore the whole disk. > L >I have taken to trying out various options with backup/list but seem to get >the same result.  >5 >Here is an example: > D >backup/list mkb600:disk5.bac/save_set/select=[cmpdev.exane.0dat]*.*  H AFAIK, backup/list does not use the "/select" qualifier. It will always I list the entire backup save set. The "/select" qualifier is only used on I the restore operation.   Ken    ------------------------------    Date: 25 Mar 2003 07:32:15 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler)r! Subject: Re: Restore with /select 3 Message-ID: <2+QjZpX1YMcY@eisner.encompasserve.org>.  W In article <ImVfa.98$1u.1453@news.uk.colt.net>, "Tobe" <tarkless@wilco-int.com> writes:e > Hi,S > E > backup/list mkb600:disk5.bac/save_set/select=[cmpdev.exane.0dat]*.*r >   H    You want /log, which tells you what BACKUP is doing, not /list, which.    l;ists the entire contents of the save set.   ------------------------------    Date: 25 Mar 2003 06:14:11 -0800. From: spamsink2001@yahoo.com (Alan E. Feldman)! Subject: Re: Restore with /select = Message-ID: <b096a4ee.0303250614.3f803fa2@posting.google.com>   \ "Tobe" <tarkless@wilco-int.com> wrote in message news:<ImVfa.98$1u.1453@news.uk.colt.net>... > Hi,- > K > Sorry if this has come up before I have done many searches but can't find7 > and answer to my scenario. > K > I am trying to do a selective restore from an image backup. However every ? > time I have tried it seems to want to restore the whole disk.e > M > I have taken to trying out various options with backup/list but seem to geto > the same result. >  > Here is an example:l > E > backup/list mkb600:disk5.bac/save_set/select=[cmpdev.exane.0dat]*.*c >  > Listing of save set(s) > , > %MOUNT-I-WRITELOCK, volume is write locked5 > %MOUNT-I-MOUNTED, A-C-TH mounted on _CSLA02$MKB600:u > Save set:          DISK5.BAC > Written by:        SYSTEMn$ > UIC:               [000001,000004], > Date:              20-MAR-2003 22:25:02.14
 > Command:? > BACKUP/IMAGE/IGNORE=INTERLOCK/NOASSIST/LABEL=A-C-TH/EXCLUDE=R 4 > ESERVE.SYS* CSLA02$DKA500: CSLA02$MKB600:DISK5.BAC- > Operating system:  OpenVMS AXP version V6.2u > BACKUP version:    V6.2  [...]eL > [BGL_TEST.0DAT]CM100D.SRT;1                                 4  30-AUG-2001 > 18:41aL > [BGL_TEST.0DAT]CM510A.;1                                  261  25-MAY-1999 > 19:16 L > [BGL_TEST.0DAT]CMACRE.CMD;1                                 1   3-JUL-1990 > 11:595L > [BGL_TEST.0DAT]CMBNRE.CMD;1                                 1  27-NOV-1991 > 09:25vL > [BGL_TEST.0DAT]CMBNRX.CMD;1                                 1  20-SEP-1989 > 12:363L > [BGL_TEST.0DAT]CMCASH.;7                                80562  27-MAY-2002 > 13:40- > .- > .  > .e > # > Any ideas would be most grateful.  >  > Toby Arkless > ADP Wilco2     /LIST ignores /SELECT. t  @ Try it with /LOG instead of /LIST and give it a place to put theF files. Or use /CONFIRM. You'll probably see that it works fine. Try it on a scracth disk if you can.    Disclaimer: JMHO Alan E. Feldman.   ------------------------------  % Date: Tue, 25 Mar 2003 16:13:45 -0000 % From: "Tobe" <tarkless@wilco-int.com>t! Subject: Re: Restore with /selecti0 Message-ID: <4h%fa.106$1u.1519@news.uk.colt.net>  4 Thanks to all who responded, It is now working fine.   Toby   ------------------------------  # Date: Tue, 25 Mar 2003 14:01:59 GMT ) From: "Tom Triske" <tom884@insightbb.com> " Subject: Standalone backup on 6510/ Message-ID: <rjZfa.231264$qi4.119191@rwcrnsc54>f  G I am trying to do a Standalone backup on a 6510 that has the standaloneoK backup kit built on $7$dua1:.  Does anyone  know the command to do this?  Io3 can't find the manual to tell me the proper format.    Thanks,p   TomX   ------------------------------  # Date: Tue, 25 Mar 2003 14:09:43 GMTo+ From: Jeff Cameron <JCam90502@jcameron.com>o& Subject: Re: Standalone backup on 65102 Message-ID: <BAA5A326.65BF%JCam90502@jcameron.com>  F On 3/25/03 6:01 AM, in article rjZfa.231264$qi4.119191@rwcrnsc54, "Tom% Triske" <tom884@insightbb.com> wrote:e  I > I am trying to do a Standalone backup on a 6510 that has the standalone M > backup kit built on $7$dua1:.  Does anyone  know the command to do this?  Iu5 > can't find the manual to tell me the proper format.l > 	 > Thanks,w >  > Tomi >  > = I believe it is done using the following at the boot prompt :a   >>>BOOT/R5:E0000000 DUA1  * This tells it to boot from root E on DUA1.   Jeff   ------------------------------  % Date: Tue, 25 Mar 2003 11:25:22 +0000c' From: Andrew Harrison SUNUK Consultancy / Subject: Re: Sun unix/linux new nasty bug cert! . Message-ID: <3E803CA2.4060407@nospamn.sun.com>   Bob Ceculski wrote:u > Andrew Harrison SUNUK Consultancy <Andrew_No.Harrison_No@nospamn.sun.com> wrote in message news:<3E7EF2D1.8060209@nospamn.sun.com>...e >  >>Rudolf Wingert wrote:T >>	 >>>Hello,w >>>e >>>Andrew Harrison wrotes: >>>  >>>e >>>Well to be more clear.P- >>>Solaris, HP-UX, Tru64, OpenVMS, AIX, Linux # >>>IRIX are all probably vunerable.b >>><<< >>> L >>>Where did you get the info about OpenVMS. Years ago we did use the XDR toL >>>exchange data from Sun platform to OpenVMS. The Sun user did report a bugI >>>within OpenVMS (ACCVIO). What was the problem? H e did not specify any(I >>>memory address (0 for get automatic memory). So the memory address was O >>>outsite of his space. On Sun this is not an error. In case of this, there isAM >>>no possibility to get root priviledges under OpenVMS. All what you do (XDRwK >>>or other tools with the buffer overflow problem), can you do only withinnK >>>your context, or yue will be stopped via the OS. Look also the intrutioneO >>>report of DECUS Muenchen. XDR under OpenVMS may be buggy, but you do not gett- >>>any right over the normal user's right!!!!  >>>i >>>Best regards Rudolf Wingert >>>e >>5 >>I didn't there is currently no response for OpenVMSm2 >>on the CERT web page along with HP-UX and Tru64. >>3 >>However it is a generic XDR issue so Bob's claimsm8 >>for OpenVMS being invunerable are as always premature. >>	 >>Regardse >>Andrew Harrisonc >  > B > wrong, as always they are right on the money ... VMS can squelchC > the nastiest unix kernel bugs ... just admit that VMS is superiors? > to slowaris, unix, linux, windoze or any other garbage os outl > there ...e  2 Bob, Bob why do you always end up getting yourself4 into trouble by forgetting that there are documented4 examples of OpenVMS not squelching nasty UNIX kernel2 bugs. Or had you forgotten the CERTS and non CERTS for POD, DOS etc ?????  4 We have had this discussion on a number of occasions4 in the past and it is a great shame for you that you3 continue to repeat arguments that are not supportedp by HPQ's own patch reports.e   Regardsa Andrew Harrison    ------------------------------    Date: 25 Mar 2003 07:28:38 -0600B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley). Subject: Re: System reqirements of OpenVMS 7.33 Message-ID: <5rzNN6odbOAb@eisner.encompasserve.org>i  x In article <b6BuMjauRxdB@eisner.encompasserve.org>, clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley) writes: > L > However, as an aside, I did once manage to install V7.2 on an AlphaStationK > 200 4/166 with only 24Mb of memory installed. I did have to do a _lot_ ofvG > manual adjustment of SYSGEN parameters (turning off the VFC, reducingl=                                                           ^^^M s/VFC/VIOC/i  D I've obviously been doing too much reading about the XFC recently...  N > non paged memory requirements, etc) and startup files (ie: stop AUTOGEN fromO > running after discovering the SYSGEN parameters were too low) in order to getn > the boot to finish.n >    Simon.   --  B Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       L VMS advocate: One who makes a Mac advocate look like a beginner at advocacy.   ------------------------------    Date: 25 Mar 2003 07:18:17 -0600B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley). Subject: Re: System reqirements of OpenVMS 7.33 Message-ID: <b6BuMjauRxdB@eisner.encompasserve.org>?  U In article <3E7ED01E.C2FE4D49@mtu-net.ru>, GluckMaker <gluckmaker@mtu-net.ru> writes:t
 > Hello, All!D > R > Now I am trying to install VMS 7.3 on my Alpha machine: DEC 3000/300LX (please, N > don't laugh!!!). It is 21064-125 with 32 Mb of RAM and 2Gb HDD. InstallationM > starts, all selected software (VMS itself, DECwindows, DECnet and TCPIP) isDO > copied, installator proposes me to restart system, but after reboot I recieve  > these messages:  >   H Others have already told you that 64Mb is the supported minimum for VMS, and they are correct.Y  J However, as an aside, I did once manage to install V7.2 on an AlphaStationI 200 4/166 with only 24Mb of memory installed. I did have to do a _lot_ ofaE manual adjustment of SYSGEN parameters (turning off the VFC, reducing L non paged memory requirements, etc) and startup files (ie: stop AUTOGEN fromM running after discovering the SYSGEN parameters were too low) in order to get' the boot to finish.r  J I don't remember all the details as this was at least 2-3 years ago, and IF went and bought more memory the day after I tried to install GNAT. :-)  F As you are a VMS newcomer, I would not recommend that you get involvedG in trying to do this, it requires VMS knowledge to know what parametersi to adjust and by how much.  : > %%%%%%%%%%%  OPCOM  23-MAR-2003 14:58:28.00  %%%%%%%%%%%# > Message from user SYSTEM on ALPHA C > %JBC-E-OPENERR, error opening SYS$COMMON:[SYSEXE]QMAN$MASTER.DAT;  > : > %%%%%%%%%%%  OPCOM  23-MAR-2003 14:58:28.00  %%%%%%%%%%%# > Message from user SYSTEM on ALPHAD > -RMS-E-FNF, file not found >   L You don't have a queue manager database yet; issue a START/QUEUE/MANAGER/NEWD to create one. Warning: If you do this again, you will overwrite the< existing database; see the VMS documentation or online help.   > K > P.S. Maybe, my questions are silly, but I have never seen OpenVMS before.I >    I don't consider them silly.  I BTW, have you read the VMS FAQ yet ? You will find lots of information inh/ there. Start at http://www.openvms.compaq.com .a   Simon.   -- aB Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       L VMS advocate: One who makes a Mac advocate look like a beginner at advocacy.   ------------------------------  % Date: Tue, 25 Mar 2003 10:08:59 -0500IA From: "Fred Kleinsorge" <kleinsorge@star-dot-zko-dot-dec-dot-com>e. Subject: Re: System reqirements of OpenVMS 7.3. Message-ID: <3e80710b$1_1@hpb10302.boi.hp.com>  K As has been said, 64mb is really the smallest memory configuration.  If youlI want to try running it, you should not install DECnet/OSI (only phase 4),A: and look to reduce any use of physical memory - like VIOC.  J I have installed V7.3-1 on a Alphabook, with I think only 32mb, and only a? 1.2 GB drive - but it is a tight squeeze, and very unsupported.T      G "Simon Clubley" <clubley@remove_me.eisner.decus.org-Earth.UFP> wrote inA5 message news:5rzNN6odbOAb@eisner.encompasserve.org... 5 > In article <b6BuMjauRxdB@eisner.encompasserve.org>,'D clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley) writes: > >.A > > However, as an aside, I did once manage to install V7.2 on an  AlphaStationJ > > 200 4/166 with only 24Mb of memory installed. I did have to do a _lot_ ofI > > manual adjustment of SYSGEN parameters (turning off the VFC, reducingG? >                                                           ^^^D
 > s/VFC/VIOC/* >*F > I've obviously been doing too much reading about the XFC recently... > K > > non paged memory requirements, etc) and startup files (ie: stop AUTOGEND fromJ > > running after discovering the SYSGEN parameters were too low) in order to get > > the boot to finish.D > >p >e > Simon. >  > --= > Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFPoD > VMS advocate: One who makes a Mac advocate look like a beginner at	 advocacy.e   ------------------------------  % Date: Tue, 25 Mar 2003 08:14:57 -0500y, From: "Island" <dbturner@nospamislandco.com>- Subject: Re: System Special for VMS NewsGroup / Message-ID: <v80ljgbtpbcg50@news.supernews.com>5  A The DEPVZ has 4MB SGRAM on board and uses the Permedia 2 Chipset.nH Shame that this board is no longer manufactured, as it was a great idea.  I We have one in stock for about $675 - list price is/was way too high - weI* just purchased one for stock just in case.  I AS for dual heading, I don't think this card has a VGA enable/disable, so G you would need a to use this card as a primary video card and your elsay$ gloria (no other card) as secondary.I FYI, the ELSA has 8MB SGRAM, but I assume that isn't too much of an issuet   Davidc   ------------------------------  % Date: Tue, 25 Mar 2003 09:51:33 -0600-( From: Michael Rice <marice@whiteice.com>- Subject: Re: System Special for VMS NewsGroup./ Message-ID: <v80uo6hv2m3m8e@corp.supernews.com>M  # On 3/25/2003 7:14 AM, Island wrote:3C > The DEPVZ has 4MB SGRAM on board and uses the Permedia 2 Chipset.RJ > Shame that this board is no longer manufactured, as it was a great idea. > K > We have one in stock for about $675 - list price is/was way too high - wei&                             ^^^^^^^^^^7 I'm not *that* keen on having a dual-headed VMS system.    ------------------------------  % Date: Tue, 25 Mar 2003 08:28:31 +0000N( From: Nic Clews <sendspamhere@127.0.0.1>: Subject: Re: Terminal driver for VAX synchronous interface) Message-ID: <3E80132F.FEAF660E@127.0.0.1>A   Michael Unger wrote: > G > > This is for the hobby system. I am using AUTODIAL UUCP (ported fromaJ > > UNIX) to manage the dial-up, wrapped in a DCL procedure to perform theK > > dial out and establish the link, then switch up a dynamic PPP link fromnG > > the IP software. The complication is that the ISDN terminal adapternJ > > talks synchronous, and no serial ports I know of are capable under VAXK > > and VMS which present themselves as terminal class (unless anyone knows 7 > > better). Only looking for single port connectivity.a  J > Strange -- I haven't seen any ISDN adapter with a synchronous interface,F > just native RS/EIA-232 ones and asynchronous communication. To avoidJ > problems the baud rate from/to the computer has to be set above the ISDND > basic channel rate of 64kbit and _hardware_ flow control has to be > activated.  D The terminal adapter I'm planning to use is a US Robotics Courier-I.C Examining the manuals, it says that the dial-up link can be made indG asynchronous at whatever baud rate you decide, then once the connection H is established, it switches to synchronous for the 64k or 128k link. TheH device supports bandwidth on demand, so arguably the device could switch  from 64k to 128k and back again.  E I'll ask another question then, based on your information, if I had ahE terminal server that did hardware flow control, and could communicateo: (say) at 115,000 bps (say a DS 700-8MC) that even when theE modem/terminal adapter (pegged at 64k) flips into synchronous, it maye work?o  & (Also thanks to Mr. SMS for the email)  G FYI: Why ISDN: I'm now living in the back of beyond where gone 10.AM ishH "early" according to the (few) locals, and broadband (other than perhaps# satellite) is not due anytime soon.b   --  ? Regards, Nic Clews a.k.a. Mr. CP Charges, CSC Computer SciencesO nclews at csc dot com-   ------------------------------  % Date: Tue, 25 Mar 2003 18:12:28 +0100s$ From: Michael Unger <unger@decus.de>: Subject: Re: Terminal driver for VAX synchronous interface* Message-ID: <00A1D682.8ED0EA69.5@decus.de>  + "Nic Clews" <sendspamhere@127.0.0.1> wrote:b   > Michael Unger wrote: > > I > > > This is for the hobby system. I am using AUTODIAL UUCP (ported fromtL > > > UNIX) to manage the dial-up, wrapped in a DCL procedure to perform theM > > > dial out and establish the link, then switch up a dynamic PPP link fromeI > > > the IP software. The complication is that the ISDN terminal adapteriL > > > talks synchronous, and no serial ports I know of are capable under VAXM > > > and VMS which present themselves as terminal class (unless anyone knowsB9 > > > better). Only looking for single port connectivity.B > L > > Strange -- I haven't seen any ISDN adapter with a synchronous interface,H > > just native RS/EIA-232 ones and asynchronous communication. To avoidL > > problems the baud rate from/to the computer has to be set above the ISDNF > > basic channel rate of 64kbit and _hardware_ flow control has to be > > activated. > F > The terminal adapter I'm planning to use is a US Robotics Courier-I.E > Examining the manuals, it says that the dial-up link can be made ingI > asynchronous at whatever baud rate you decide, then once the connection J > is established, it switches to synchronous for the 64k or 128k link. TheJ > device supports bandwidth on demand, so arguably the device could switch" > from 64k to 128k and back again.  & I dont't know that particular adapter.  A ISDN communication is synchronous, RS/EIA-232 asynchronous -- theeH terminal adapter should handle the "translation". The baud rate from theH ISDN adapter to the computer should be above the ISDN rate to avoid dataA congestion (I don't know if the ISDN adapter can deal with a slowtE connection _to_ the computer); the baud rate from the computer to then@ ISDN adapter may be below the ISDN rate (which doesn't cause anyG problem) or above that (which is handled by the hardware flow control).rE Just don't use _software_ flow control (i.e., XON/XOFF) because thesee9 "characters" may be embedded in the "normal" data stream..  G > I'll ask another question then, based on your information, if I had abG > terminal server that did hardware flow control, and could communicate < > (say) at 115,000 bps (say a DS 700-8MC) that even when theG > modem/terminal adapter (pegged at 64k) flips into synchronous, it maym > work?0  B It really should! But even a "simple" serial line interface should9 support hardware flow control, even on a 9-pin connector.y  G At home I'm using a similar setup to dial-up to the DECUS VMS system: aaG tiny PBX system (with an integrated modem adapter) connected to an ISDNnH line; the connection to the PC is made via a USB interface. The terminalF emulation running on the PC uses a "virtual COM port" established by aF special software driver which physically uses the USB interface to theE PBX system. It's running without any problems. (Well, not considering'/ the "normal" Winwoes problems of course ... :-)e   Michaele   ------------------------------    Date: 25 Mar 2003 06:52:31 -0600B From: clubley@remove_me.eisner.decus.org-Earth.UFP (Simon Clubley)G Subject: Testing HP's attitude to VMS, was: Re: HP doesn't know OpenVMSo3 Message-ID: <gmmLQQoCgkNd@eisner.encompasserve.org>r  b In article <3e7f6556$1@news.si.com>, "Brian Tillman" <Brian.Tillman@notnoone.notnohow.com> writes:? > I wanted to find out some information on Disk File Optimizer.o  6 [Another example of HP not knowing about VMS deleted.]  C It would be interesting to know if the above types of incidents aree normal or unfortunate one-offs.n  L Has anybody tried pretending to be a potential new customer and then ringingI up HP or one of it's VAR's, expressing interest in "this thing called VMS . that I've just discovered on your Web pages" ?  L If the HP/VAR's response is negative in nature, then it would be interestingG to know if it's a "I don't know what VMS is" type response or a "VMS isaL obsolete, Windows (or Linux) is a far superior and reliable technology" type	 response.   E I would try this myself, but I am already a customer of the companieseK that I can think of trying, and hence I am known to them as a VMS customer.t   Simon.   -- eB Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP       L VMS advocate: One who makes a Mac advocate look like a beginner at advocacy.   ------------------------------    Date: 25 Mar 2003 10:33:13 +0100' From: huber@mppmu.mpg.de (Joseph Huber)R' Subject: Re: X-window and Xlib questionn+ Message-ID: <VB55MZwgtmwD@vms.mppmu.mpg.de>e  W In article <3E7F6B0B.DF00B2A9@columbia.edu>, Reza Tehrani <mt2151@columbia.edu> writes:e >> a> >> Can you display X11 stuff FROM somewhere else ON the ALPHA? > H > Actually no. I have tried to telnet to another VMS and ran decterm andD > it didn't give permission. I added the IP of the remote VMS to theF > security and used the following command "set dis/cre/node=IP_Number"0 > with /tran=decnet and without it. None worked.  9 The transport for IP is TCPIP, decnet is irrelevant here.iK Did You enable TCPIP as decwindow transport ? It is not enabled by default!t4 Look into sys$manager:decw$PRIVATE_SERVER_SETUP.COM:E is the symbol decw$server_transports defined with TCPIP in the list ? = If decw$PRIVATE_SERVER_SETUP.COM doesn't exist, copy it from h# decw$PRIVATE_SERVER_SETUP.template.e  9 >> > X-SERVER    6000  TCP  not defined  0.0.0.0 Disabledd8 >> > FONT-SERVER 7100  TCP  not defined 0.0.0.0 Disabled( >> What are the "OpenVMS network ports"?( >> Note that X can use ports 6000--6063.F > When I did "ucx show services" I didn't see any port in the range ofD > 6000 and 7000. Therefore I added the 6000 and 7100 to the servicesH > using  "ucx set service /port=6000 ..." command. I am not sure if thisE > is the correct way of doing. How did you enable your X-lib ports tow > receive the connection?   t> You should not deal manually with any services on port 6000 ,  decw$startup does it.,D Do ucx or tcpip "show device": it will list local ports 6000 for theA decwindows  server, and remote ports 6000 for decwwindow clients.t   --  N Joseph "Sepp" Huber   mailto:joseph.huber@web.de   http://www.huber-joseph.de/   ------------------------------    Date: 25 Mar 2003 07:26:06 -0600; From: koehler@eisner.nospam.encompasserve.org (Bob Koehler) Q Subject: Re: [Q] Files not found during image backup, w/ and w/o directory-specs.a3 Message-ID: <TPSn8jFIlfw9@eisner.encompasserve.org>h  a In article <240320031847599514%carlos@canada.terra.sol>, Carlos <carlos@canada.terra.sol> writes:t  E > When I look at the disk after, the files indicated really are *not* G > there, so BACKUP isn't lying. But why does it try to copy them? We're0B > still working on this with Gold support and if we come up with a > solution I'll post it.  F    What does analyze/disk_structure say about the volume?  Can it findH    the files or fix the headers?  Mostly I'm guessing, but it looks likeH    the index file has pointers to files that have been deleted, possibly*    from a directory that has been deleted.   ------------------------------  # Date: Tue, 25 Mar 2003 13:32:00 GMT>9 From: Alan Adams <alan.adams@orchard-way.freeserve.co.uk> Q Subject: Re: [Q] Files not found during image backup, w/ and w/o directory-specs.u? Message-ID: <4ab1bcd84b.Alan.Adams@orchard-way.freeserve.co.uk>   & In message <3E7FD2E4.6D0C2CCE@fsi.net><           "David J. Dachtera" <djesys.nospam@fsi.net> wrote:   > Carlos wrote:n > > K > > I had the exact same errors using BACKUP/IMAGE in VMS 7.1-2. In my casepI > > I was doing a BACKUP/IMAGE from one disk drive to another in order torJ > > defragment the disk. Both drives were not mounted system wide, I did aH > > purge/keep=1, then an ANAL/DISK/REPAIR on the source drive, then the > > BACKUP/IMAGE.a > > C > > I got the same error of NOSUCHFILE, with the empty [] directoryiH > > indication. Absolutely there was no activity on the drive except for > > the BACKUP.r > > H > > Being a GOLD Support client, I escalated this to HP, showing them myJ > > DCL code and the error log. Basically, the guy I talked to was stumpedH > > and his only suggestion was to do explicit ALLOCATE before doing theE > > BACKUP/IMAGE. Doubtfully I did do that but with the same results.t > I > What you're likely seeing is the result of "orphaned" directory entriesyF > as well as "lost" files (files with no directory entry or an invalid > directory backlink). >   I The output of ANALYZE/DISK/NOREPAIR would be interesting. I would tend toiO suggest regular use of ANALYZE/DISK/REPAIR - we used to run it over the weekendnO when the exclusive lock on the disk for several minutes didn't cause a problem.l   Alan   -- a
 Alan Adams& alan.adams@orchard-way.freeserve.co.uk http://www.nckc.org.uk/    ------------------------------   End of INFO-VAX 2003.166 ************************