1 INFO-VAX	Fri, 21 Apr 2000	Volume 2000 : Issue 222       Contents:/ Re: 24x7 Operations (was Re: Verify of Backups) / Re: 24x7 Operations (was Re: Verify of Backups) / Re: 24x7 Operations (was Re: Verify of Backups) / Re: 24x7 Operations (was Re: Verify of Backups) / Re: 24x7 Operations (was Re: Verify of Backups)  ALTPRI and OpenVMS7.2  Re: ALTPRI and OpenVMS7.2  Re: ALTPRI and OpenVMS7.2  Re: ALTPRI and OpenVMS7.2 $ RE: Backup Strategies for Enterprise$ Re: Backup Strategies for Enterprise$ Re: Backup Strategies for Enterprise? CD-R and VMS (Was:Re: Infoserver CD-R (was: Verify of Backups))  Re: DCL Hackery  Re: DCL Hackery  RE: DCL Hackery  RE: DCL Hackery  Re: DCL Hackery  Re: DCL Hackery  Re: EMC Disk Storage Re: EMC Disk Storage
 Re: excursion 
 Re: excursion , Re: Infoserver CD-R (was: Verify of Backups)1 Interesting problem with AS 255/233 and JAZ drive & Re: LK46W-A2 & "DO" key in Reflections Re: Mozilla M15's out... Re: Mozilla M15's out... Re: Mozilla M15's out... Need BLISS advice  Re: Need BLISS advice  Re: Need BLISS advice * Re: OpenVMS mail question: Attached files?  OUTCOME: DECC 6.n BUG? /OPTIMIZE# Re: Problem reading files from tape  queue status values  RE: queue status values  Re: queue status values  Re: queue status values  RE: queue status values / Re: Upgraded (hardware) 8400 System not booting / Re: Upgraded (hardware) 8400 System not booting * Re: VAX CI storage vs served fibre channel* Re: VAX CI storage vs served fibre channel Re: Verify of Backups  RE: Verify of Backups  Re: Verify of Backups  Re: Verify of Backups  Re: Verify of Backups  Re: Verify of Backups  RE: Verify of Backups  Re: Verify of Backups  Re: Web server for VAX/VMS  F ----------------------------------------------------------------------   Date: 20 Apr 2000 19:34:38 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)8 Subject: Re: 24x7 Operations (was Re: Verify of Backups)6 Message-ID: <8dnm4e$pm8$1@mailint03.im.hou.compaq.com>  T In article <01JOGLE19BRAQVNO4E@CCTR.UMKC.EDU>, Ron <RROCKWELL@CCTR.UMKC.EDU> writes:  G :  Just curious, anyone, but taking this point seriously and attempting K :to apply it to a system with multiple users receiving mail at all hours... F :How can data be compared when the snapshot is obsolete by the time it7 :has finally finished?   ...Users are online 24 by 7...   I   Congratulations, welcome to the world of continous system availability, I   extended uptime and (as I prefer to call it :-) lights-always-on system K   operations.  This includes the world of on-line system backup operations.   L   No, the problems here are not easy to solve.  Yes, there is work going on I   in OpenVMS to continue to improve the performance of the system backup  K   operations through various means, including a shadowing project known as  G   mini-copy that is expected to be part of the next OpenVMS release...    L   Shadowing and other approaches become interesting here, as do things like L   RMS journaling and associated (similar) on-line database backup tools...  K   Related work is also going on out in the area of the storage controllers, $   on-disk structures, and similar...  J   And just to keep things really interesting, the volume of data involved 9   here is also (greatly and regularly) increasing, too...   N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  + Date: Thu, 20 Apr 2000 15:59:15 -0500 (CDT) # From: Ron <RROCKWELL@CCTR.UMKC.EDU> 8 Subject: Re: 24x7 Operations (was Re: Verify of Backups). Message-ID: <01JOGPCIGS8WQVNO4E@CCTR.UMKC.EDU>  #  But Hoff! I already knew that! 8^)   B  No, no, no... shadowing is no longer a possibility, the increased>  data is already taking up the spare drives. We cannot justify?  purchasing more drives for a system which is being phased out. B  Why is it being phased out?  Not because of backups, that is only>  a straw on the camel's back. Its because there is no easy andF  affordable way to connect those new PC (pretty GUI VT320's) terminalsC  in a GUI fashion to those disks that we want to back up. But thats G  a good thing for our research people. They only want CPU, not storage. A  And the e-mail crowd has its own never-ending battles to decide. ?  But, the MS world that gets a student a job in the next couple C  years is full of Xcel, Word (is that MS-WordStar or MS-WordPerfect C  or what?) and Access. None of those run under VMS and if they did, C  then they wouldn't have any PC connections, and so, the  academic  9  world is splitting.... (pun not so funny but intended).     C  And we already went down the never ending road of trying to figure G  out which products actually perform the backup task without being too  K  much overkill or simply not-enough, or obsolete, or part of a package that A  was discontinued. And they all end up using BACKUP.EXE anyhow...   A  Perhaps the best way to do this would be to "checkpoint" the job B  in the manner that XMODEM avoided line-noise an age or so back...B  Maybe i should backup each user into his own personal and privateA  save-set, if he is online, wait him until later, if he got mail, >  just overwrite him. I've got several thousand users, but the A  majority only have small account areas.... for the others, maybe A  i should just backup their areas one directory at a time.... and @  for those that have those huge massive directorys full of mail,B  well, let me think about that some more.... perhaps forward theirF  mail temporarily, back them up, then move the mail from the forwarded?  account into their mail. Speed is not a priority for this job,   its only a preference.   A  In anycase, we only now have a "new" priority for backups (since F  all of our disks are RAID5) and that spec only calls for catastrophic?  failure (F5 tornadoes, Earthquakes, etc..) The backup does not C  need to know if it is verified exactly, only if what it did backup @  was recorded without errors. If it missed a new file, it wasn'tA  in time for the "snapshot". So, is there a way to tell if a file <  was recorded incorrectly or if was simply "not there yet"?   D  Question: I was not using the /record qualifier when i used verify.F  If i had, would /verify only bother to check files that were actuallyB  recorded?  Would it then skip over verifying a mail.mai file that6  had changed during a single backup command operation?  B =>From:	IN%"hoffman@xdelta.zko.dec.nospam" 20-APR-2000 14:56:58.61  =>To:	IN%"Info-VAX@Mvb.Saic.Com" =>CC:	7 =>Subj:	RE: 24x7 Operations (was Re: Verify of Backups)  =>V =>In article <01JOGLE19BRAQVNO4E@CCTR.UMKC.EDU>, Ron <RROCKWELL@CCTR.UMKC.EDU> writes: =>I =>:  Just curious, anyone, but taking this point seriously and attempting M =>:to apply it to a system with multiple users receiving mail at all hours... H =>:How can data be compared when the snapshot is obsolete by the time it9 =>:has finally finished?   ...Users are online 24 by 7...  =>K =>  Congratulations, welcome to the world of continous system availability, K =>  extended uptime and (as I prefer to call it :-) lights-always-on system M =>  operations.  This includes the world of on-line system backup operations.  =>N =>  No, the problems here are not easy to solve.  Yes, there is work going on K =>  in OpenVMS to continue to improve the performance of the system backup  M =>  operations through various means, including a shadowing project known as  I =>  mini-copy that is expected to be part of the next OpenVMS release...   =>N =>  Shadowing and other approaches become interesting here, as do things like N =>  RMS journaling and associated (similar) on-line database backup tools...  M =>  Related work is also going on out in the area of the storage controllers, & =>  on-disk structures, and similar... =>L =>  And just to keep things really interesting, the volume of data involved ; =>  here is also (greatly and regularly) increasing, too...  =>P => --------------------------- pure personal opinion ---------------------------N =>   Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com =>     Ron Rockwell   Academic Computing Services '   University of Missouri at Kansas City    ------------------------------   Date: 20 Apr 2000 21:24:01 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)8 Subject: Re: 24x7 Operations (was Re: Verify of Backups)6 Message-ID: <8dnshh$s90$1@mailint03.im.hou.compaq.com>  T In article <01JOGPCIGS8WQVNO4E@CCTR.UMKC.EDU>, Ron <RROCKWELL@CCTR.UMKC.EDU> writes:     [discussion expurgated]   J   I am aware of no particular way to deal with these problems that is all 0   of the following: cheap, fast, and reliable...  I   I know there are various engineering projects working in this area, and 4   there are various current and planned approaches.   H   As for the BACKUP/VERIFY/RECORD, I don't believe that BACKUP does any L   sort of special-casing of the verification based on the /RECORD setting.  J   I don't know this for certain, but I cannot recall ever having heard of H   any similar processing at any time in the past...  (Does that "answer"   sound too political? :-)  G   As an aside: one approach -- with checkpointing -- is RMS journaling.   F   Approaches for remote disk support include the OpenVMS Disk ServicesD   for Windows NT stuff, PATHWORKS (Advanced Server), SAMBA, and NFS.  G   Synchronization of any of these -- or of a live application that does F   not have an in-built synchronization mechanism for use as part of an+   on-line BACKUP -- is clearly a problem...   N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 21 Apr 2000 12:21:01 +1200 * From: "Ryan C. Price" <pricerc@ihug.co.nz>8 Subject: Re: 24x7 Operations (was Re: Verify of Backups)* Message-ID: <8do6tr$rks$1@news.ihug.co.nz>  ; Would it help to immediately follow a 'full' backup with an 
 incremental ? C The window for the incremental would be much smaller than the full,  and 8 so would likely have fewer changes to affect the verify.  E This doesn't remove the need for a verify, and does required /record,  but ? it might help improve a little the amount of useful data in the  backup.   B RMS journalling would be good too, but doesn't that cost extra ? I once> set up a whole disaster recovery site using RMS journalling to maintain@ a backup copy of a file-based treasury system (mostly written in Fortran  IIRC).   /Ryan       ? "Hoff Hoffman" <hoffman@xdelta.zko.dec.nospam> wrote in message 0 news:8dnshh$s90$1@mailint03.im.hou.compaq.com... > 4 > In article <01JOGPCIGS8WQVNO4E@CCTR.UMKC.EDU>, Ron! <RROCKWELL@CCTR.UMKC.EDU> writes:  >  >   [discussion expurgated]  > D >   I am aware of no particular way to deal with these problems that is all2 >   of the following: cheap, fast, and reliable... > A >   I know there are various engineering projects working in this 	 area, and 5 >   there are various current and planned approaches.  > E >   As for the BACKUP/VERIFY/RECORD, I don't believe that BACKUP does  any C >   sort of special-casing of the verification based on the /RECORD  setting.B >   I don't know this for certain, but I cannot recall ever having heard ofA >   any similar processing at any time in the past...  (Does that  "answer" >   sound too political? :-) > = >   As an aside: one approach -- with checkpointing -- is RMS  journaling.  > ? >   Approaches for remote disk support include the OpenVMS Disk  ServicesF >   for Windows NT stuff, PATHWORKS (Advanced Server), SAMBA, and NFS. > D >   Synchronization of any of these -- or of a live application that doesE >   not have an in-built synchronization mechanism for use as part of  an- >   on-line BACKUP -- is clearly a problem...  > , >  --------------------------- pure personal# opinion --------------------------- 1 >    Hoff (Stephen) Hoffman   OpenVMS Engineering  hoffman#xdelta.zko.dec.com >  >    ------------------------------  % Date: Thu, 20 Apr 2000 20:47:16 -0400 . From: "Andrew Robert" <robert_a@ix.netcom.com>8 Subject: Re: 24x7 Operations (was Re: Verify of Backups)3 Message-ID: <8do8g2$lkc$1@nntp9.atl.mindspring.net>   * A solution I implemented at my shop was to  H Prior to the production cycle starting, do an image copy of the critical" production disks to  spare drives.  8 Once the image completes, we start the production cycle.  B While production executes, the spare drive is backed down to tape.  H In the event that there is a problem, we can easily roll the spare drive0 over and restart as if production never started.  J The disk to disk image backup runs for approximately 20 minutes and allows7 for greater recovery options in the event of a failure.    Regards,
 Andrew Robert  Principal Systems Analyst   Massachusetts Financial Services  3 Ryan C. Price <pricerc@ihug.co.nz> wrote in message $ news:8do6tr$rks$1@news.ihug.co.nz...= > Would it help to immediately follow a 'full' backup with an  > incremental ? E > The window for the incremental would be much smaller than the full,  > and : > so would likely have fewer changes to affect the verify. > G > This doesn't remove the need for a verify, and does required /record,  > but A > it might help improve a little the amount of useful data in the 	 > backup.  > D > RMS journalling would be good too, but doesn't that cost extra ? I > once@ > set up a whole disaster recovery site using RMS journalling to
 > maintainB > a backup copy of a file-based treasury system (mostly written in	 > Fortran  > IIRC). >  > /Ryan  >  >  > A > "Hoff Hoffman" <hoffman@xdelta.zko.dec.nospam> wrote in message 2 > news:8dnshh$s90$1@mailint03.im.hou.compaq.com... > > 6 > > In article <01JOGPCIGS8WQVNO4E@CCTR.UMKC.EDU>, Ron# > <RROCKWELL@CCTR.UMKC.EDU> writes:  > >  > >   [discussion expurgated]  > > F > >   I am aware of no particular way to deal with these problems that > is all4 > >   of the following: cheap, fast, and reliable... > > C > >   I know there are various engineering projects working in this  > area, and 7 > >   there are various current and planned approaches.  > > G > >   As for the BACKUP/VERIFY/RECORD, I don't believe that BACKUP does  > any E > >   sort of special-casing of the verification based on the /RECORD 
 > setting.D > >   I don't know this for certain, but I cannot recall ever having
 > heard ofC > >   any similar processing at any time in the past...  (Does that 
 > "answer" > >   sound too political? :-) > > ? > >   As an aside: one approach -- with checkpointing -- is RMS 
 > journaling.  > > A > >   Approaches for remote disk support include the OpenVMS Disk 
 > ServicesH > >   for Windows NT stuff, PATHWORKS (Advanced Server), SAMBA, and NFS. > > F > >   Synchronization of any of these -- or of a live application that > doesG > >   not have an in-built synchronization mechanism for use as part of  > an/ > >   on-line BACKUP -- is clearly a problem...  > > . > >  --------------------------- pure personal% > opinion --------------------------- 3 > >    Hoff (Stephen) Hoffman   OpenVMS Engineering  > hoffman#xdelta.zko.dec.com > >  > >  >  >    ------------------------------    Date: 20 Apr 2000 21:09:20 -0500/ From: jlauret@?.chem.sunysb.edu (Jerome LAURET)  Subject: ALTPRI and OpenVMS7.2. Message-ID: <38ffaa40_1@dilbert.ic.sunysb.edu>  E 	This has been buggig me since a while but I forgot to mention it ... I Imagine that a process was created at priority 3 (let's say a batch or a  " detached process) and one tries to $ SET PROC/ID=xxxxx/PRIO=4  & 	with ALTPRI enabled, I currently get   C %SET-E-NOTSET, need ALTPRI privilege to elevate above base priority   ? 	i.e. it completly ignore the privilege. This is also happening > with the SYSTEM account (not only from a priviledged account).  . 	Any patch one knows about which fixes this ?? 	 D 	Correlatively, from an OpenVMS6.2 (mixed with 7.2) I currently have the following behavior :   $ set proc/id=2800013C/prio=4 & SET-E-NOTSET, error modifying 2800013CG -SYSTEM-F-NOPRIV, insufficient privilege or object protection violation  $ show proc/id=2800013C G %SYSTEM-F-NOPRIV, insufficient privilege or object protection violation0  - 	all of this running from the SYSTEM account.n   	Well ? Comments ? Reactions ?     -- e6                   Jerome LAURET S.U.N.Y. @ Stony Brook$        ,,,,,      Dept. of Chemistry+       ( o o )     Stony Brook NY 11794-3400e;   ---m---U---m---------------------------------------------e&   E-mail: jlauret@mail.chem.sunysb.edu<   URL   : http://nucwww.chem.sunysb.edu/jlauret/jlauret.html   ------------------------------  # Date: Fri, 21 Apr 2000 02:00:14 GMTa* From: young_r@eisner.decus.org (Rob Young)" Subject: Re: ALTPRI and OpenVMS7.2' Message-ID: <2000Apr20.220014.1@eisner>C  ` In article <38ffaa40_1@dilbert.ic.sunysb.edu>, jlauret@?.chem.sunysb.edu (Jerome LAURET) writes:   >  > $ set proc/id=2800013C/prio=4o( > SET-E-NOTSET, error modifying 2800013CI > -SYSTEM-F-NOPRIV, insufficient privilege or object protection violation  > $ show proc/id=2800013CRI > %SYSTEM-F-NOPRIV, insufficient privilege or object protection violation  > / > 	all of this running from the SYSTEM account.c >   > 	Well ? Comments ? Reactions ? >    	Yeah... do this:    	$ show proc/priv  ! Prior toP 	$ show proc/cont/id=2800013ce  > 	Somebody been fooling with your system account?  Normally all 	privs are turned on.e   				Robt   ------------------------------  # Date: Fri, 21 Apr 2000 02:05:18 GMT / From: "John Nixon" <jorlnixon@worldnet.att.net> " Subject: Re: ALTPRI and OpenVMS7.2G Message-ID: <yHOL4.15227$PV.1045717@bgtnsc06-news.ops.worldnet.att.net>i  I You say you have the ALTPRI priv enabled, but are you sure it is not just G set as an authorized privilege, but not a default priv?  Have you trieda doing SET PROC/PRIV=ALTPRI?   < "Jerome LAURET" <jlauret@?.chem.sunysb.edu> wrote in message( news:38ffaa40_1@dilbert.ic.sunysb.edu... >eF > This has been buggig me since a while but I forgot to mention it ...J > Imagine that a process was created at priority 3 (let's say a batch or a$ > detached process) and one tries to > $ SET PROC/ID=xxxxx/PRIO=4 >A& > with ALTPRI enabled, I currently get >UE > %SET-E-NOTSET, need ALTPRI privilege to elevate above base priorityo >y@ > i.e. it completly ignore the privilege. This is also happening@ > with the SYSTEM account (not only from a priviledged account). >s/ > Any patch one knows about which fixes this ??s >?E > Correlatively, from an OpenVMS6.2 (mixed with 7.2) I currently have  > the following behavior : >a > $ set proc/id=2800013C/prio=4a( > SET-E-NOTSET, error modifying 2800013CI > -SYSTEM-F-NOPRIV, insufficient privilege or object protection violationp > $ show proc/id=2800013CbI > %SYSTEM-F-NOPRIV, insufficient privilege or object protection violationS > . > all of this running from the SYSTEM account. >a > Well ? Comments ? Reactions ?o >m >s > --8 >                   Jerome LAURET S.U.N.Y. @ Stony Brook& >        ,,,,,      Dept. of Chemistry- >       ( o o )     Stony Brook NY 11794-3400s= >   ---m---U---m---------------------------------------------a( >   E-mail: jlauret@mail.chem.sunysb.edu> >   URL   : http://nucwww.chem.sunysb.edu/jlauret/jlauret.html   ------------------------------   Date: 21 Apr 2000 00:20 CST.' From: carl@gerg.tamu.edu (Carl Perkins)t" Subject: Re: ALTPRI and OpenVMS7.2- Message-ID: <21APR200000204207@gerg.tamu.edu>e  & jlauret@mail.chem.sunysb.edu writes...E }	Correlatively, from an OpenVMS6.2 (mixed with 7.2) I currently haven }the following behavior :r }  }$ set proc/id=2800013C/prio=4' }SET-E-NOTSET, error modifying 2800013ChH }-SYSTEM-F-NOPRIV, insufficient privilege or object protection violation }$ show proc/id=2800013CH }%SYSTEM-F-NOPRIV, insufficient privilege or object protection violation } . }	all of this running from the SYSTEM account. }  }	Well ? Comments ? Reactions ?  }-- 7 }                  Jerome LAURET S.U.N.Y. @ Stony Brooke  F Just because the account is named SYSTEM does not mean that it has all) privileges, or that they are all enabled.   7 I suspect that you do not have WORLD privilege enabled.   9 What does SHOW PROC/PRIV show when it gives these errors?a   --- Carl   ------------------------------  % Date: Thu, 20 Apr 2000 20:27:26 +0100s& From: Alan Fay <alanfay@webcode.co.uk>- Subject: RE: Backup Strategies for Enterprisey( Message-ID: <8dnls6$24c$1@supernews.com>   >Don Rogstad wrote:- > V > Our current environment has lots of Sun boxes, Digital Unix boxes, Windows NT boxes ] > and of course, OpenVMS VAX and Alpha machines.  We are now looking for ways to consolidate N8 > the backup process from multiple machines / platforms. >dX > Does anyone consolidate backups from various platforms?   What products do you use?   ] > Given the reliability of OpenVMS backup, I would like to use OpenVMS as the backup server,  , > but this is not necessarily a requirement. > ` >    It looks like Compaq has two solutions.   One, ABS and the other Legato.  Has anyone used ic > Legato Networker?  Wasn't ABS sold to CA?   Anyone used either one for multiple platform backups?d >h7 >   Your thoughts and ideas on backups are appreciated.  >D	 > Thanks,l
 > Don Rogstade         VERITAS NETBACKUPo; ------------------------------------------------------------ OpenVMS Client      7 The OpenVMS client is the client side implementation ofi2 NetBackup. This client will work with all UNIX and8 Windows NT NetBackup servers. Backup data flows from the4 VMS client, over a TCP/IP network to the server, and& then onto backup disk or tape devices.  - The VERITAS Netbackup OpenVMS Client runs on ,  (     VAX   - OpenVMS 5.5 to OpenVMS 7.2-1(     Alpha - OpenVMS 6.1 to OpenVMS 7.2-1  : Supported TCP/IP stacks: UCX, TCPWare, Pathway & Multinet.  : Backups can be initiated from the server or the VMS client: with native VMS file-specifications and backed-up files on? the server can be browsed with native VMS file-specifications, p. for example to backup the disk DKA500: specify  
 $ RUN NBU_AXPe' NBU> BACKUP/LOG DKA500:[000000...]*.*;*k  2 to backup just the directory DKA500:[ZENO] specify  " NBU> BACKUP/LOG DKA500:[ZENO]*.*;*  > to backup the directory ZENO and all directories below specify  % NBU> BACKUP/LOG DKA500:[ZENO...]*.*;*r  = to backup the disk DKA500: ignoring file locks by other userso( and the NOBACKUP file attribute, specify  C NBU> BACKUP/LOG DKA500:[000000...]*.*;*/IGNORE=(INTERLOCK,NOBACKUP)   9 to list files backed-up on the server from [ZENO] specifyy   NBU> DIR DKA500:[ZENO]*.*;*c  C to list files backup-up on the server from the top level directory @ of device DKA500: specifyP   NBU> DIR DKA500:[000000]*.*;*o  . to restore the files in directory ZENO specify  # NBU> RESTORE/LOG DKA500:[ZENO]*.*;*   > or all files and directories recursively below [ZENO], specify  & NBU> RESTORE/LOG DKA500:[ZENO...]*.*;*  ? to backup a rooted logical (preserving that logical name on thep server) specfy  - NBU> BACKUP/LOG ROOTED$LOGICAL:[ZENO...]*.*;*s  * and list the backed-up files on the server  & NBU> DIR ROOTED$LOGICAL:[ZENO...]*.*;*  	 NBU> EXITe $   1     Tabe 12-2  Summary of Commands and Qualifiersi=     --------------------------------------------------------->      Command           Qualifiers=     --------------------------------------------------------- 3     BACKUP            /ALTERNATE_CLIENT=client_names'                       /CLASS=class_name )                       /CLIENT=client_name -                       /CUMULATIVE_INCREMENTALf/                       /DIFFERENTIAL_INCREMENTALr+                       /EXCLUDE=exclude_filej                       /FULL 2                       /IGNORE=(INTERLOCK,NOBACKUP)"                       /INCREMENTAL-                       /LIMIT_BANDWIDTH=kbytesn                       /LOG                       /NORECORD '                       /OUTPUT=file-speco'                       /PORT=port_numbere-                       /SCHEDULE=schedule_name )                       /SERVER=server_namei  %     DCL               any-dcl-commanda       DIRECTORY         /FULLr                       /LASTn                       /NOROOT-'                       /OUTPUT=file-spec-*                       /IDENTIFIER=backupid                       /IMAGE                       /TOTAL       EXIT       HELP  3     RESTORE           /ALTERNATE_CLIENT=client_namea'                       /CLASS=class_name-)                       /CLIENT=client_namer!                       /CONTIGUOUSH*                       /IDENTIFIER=backupid$                       /KEEP_EXISTING-                       /LIMIT_BANDWIDTH=kbytes                        /LOG                       /NOROOTL'                       /OUTPUT=file-spec                         /OVERWRITE'                       /PORT=port_numberm-                       /SCHEDULE=schedule_name,)                       /SERVER=server_nameo,                       /TO=file-specification  '     SHOW CLIENT       /OUTPUT=file-spece  '     SHOW SERVER       /OUTPUT=file-specY=     ---------------------------------------------------------a      ?     The following file attributes are saved and restored by the      VMS Client:t  4     Table 8-2  Save and restored VMS file attributesB     --------------------------------------------------------------4     file allocation quantity       record attributes0     file organization              record format8     maximum record size            maximum record number9     bucket size                    block size            e<     default extention quantity     fixed length control area6     global buffer count            contiguous best try?     group number of file owner     member number of file owner AB     file protection                access control list length     *     complete access control list   (ACL's)7     creation date and time         backup date and timef9     expiration date and time       revision date and time 8     revision number                longest record lengthB     directory/file version limit   all database RMS index keys (1)B     --------------------------------------------------------------     (1) Primary and alternate.    ( Please make enquires to www.veritas.com.  
 Thank you,   alan                            ------------------------------  % Date: Thu, 20 Apr 2000 13:44:25 -0700f* From: "Jack Peacock" <peacock@simconv.com>- Subject: Re: Backup Strategies for Enterpriset> Message-ID: <YWJL4.9953$L7.1563461@news-west.usenetserver.com>  3 "Alan Fay" <alanfay@webcode.co.uk> wrote in messagea" news:8dnls6$24c$1@supernews.com... >d9 > The OpenVMS client is the client side implementation of 4 > NetBackup. This client will work with all UNIX and: > Windows NT NetBackup servers. Backup data flows from the6 > VMS client, over a TCP/IP network to the server, and( > then onto backup disk or tape devices. > H Umm, how do you restore a VMS system disk with this product?  That's oneG of the problems with the NT backup programs, you always have to rebuild H a base system then restore the backup tape.  It's much faster to restoreH a BACKUP image tape on VMS, worst case is a boot from CD-ROM on Alpha or from tape on a VAX.k  H Or is this the usual PC-centric answer...run RAIDs or mirrored drives so> you never lose a system disk, ignoring the situation where theF disgruntled/new sys admin casually erases the system root directories,F an operation which is faithfully replicated on those same RAID drives.E And don't ever worry about the weekend air-conditioning failure which & fries every single drive in the array.  G Impromptu poll, how many out there have come across inexperienced users : who believe mirrored drives or RAIDs replace backup tapes?    Jack Peacockp   ------------------------------  % Date: Fri, 21 Apr 2000 07:29:07 +0200K. From: "Gorazd Kikelj" <gorazd.kikelj@aster.si>- Subject: Re: Backup Strategies for Enterprisep/ Message-ID: <BERL4.2508$ep.85146@news.siol.net>n   Hay,  D in mixed environment like yours you can consider Time Navigator from	 Quadratec. www.quadratec-software.com  C It can do VMS image backup and it's very reliable and good product.f  L I use Time Navigator and ABS. I'm very pleased with ABS on VMS and with TiNa1 on mixed (vms/unix/nt/oracle/exchange) platforms.p   best regards, Gorazd KikeljA    8 "Don Rogstad" <Don.Rogstad@dalsemi.com> wrote in messageB news:EA5FE99F5DEAD311A6CB00805F199992BCA805@misnts1.dalsemi.com...L > Our current environment has lots of Sun boxes, Digital Unix boxes, WindowsK NT boxes and of course, OpenVMS VAX and Alpha machines.  We are now lookingdC for ways to consolidate the backup process from multiple machines /u
 platforms. > L > Does anyone consolidate backups from various platforms?   What products doG you use?   Given the reliability of OpenVMS backup, I would like to usedH OpenVMS as the backup server, but this is not necessarily a requirement. >sF >     It looks like Compaq has two solutions.   One, ABS and the otherK Legato.  Has anyone used Legato Networker?  Wasn't ABS sold to CA?   Anyonel. used either one for multiple platform backups? >r8 >    Your thoughts and ideas on backups are appreciated. >@	 > Thanks,o
 > Don Rogstad  > Dallas Semiconductor > 972.371.4570 > Don.Rogstad@dalsemi.coma   ------------------------------  % Date: Thu, 20 Apr 2000 22:56:56 +0200 . From: Marcin Szczecinski <marcin@lodz.tpsa.pl>H Subject: CD-R and VMS (Was:Re: Infoserver CD-R (was: Verify of Backups))0 Message-ID: <20000420225656.B14598@lodz.tpsa.pl>  : On Thu, 20 Apr 2000 at 19:37:27 +0000, Gord Coulman wrote: > J > One method by Bystrik Krchnavy suggests writing to a hard disk partitionE > on the Infoserver, then mounting the Infoserver partition from a PC G > equipped with a CD writer, then burning a "raw" copy of the partition F > to the CD.  Just one catch: how do you mount an Infoserver partition > from a PC?J > I hope it doesn't involve anything starting with "Path" and ending with 
 > "Works".  = What is wrong with PathWorks ? It works quite good for years.O: I have just finished burning next VMS CD (AXP VMS 6.2-1H3)> without InfoServer and "disk partitions" but with Decnet (IV),< PathWorks 32, DFY$VMSCD (by Istvan Dosa) and Adaptec EazyCD.  3 Marcin Szczecinski              marcin@lodz.tpsa.plv   ------------------------------   Date: 20 Apr 2000 18:07:12 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: DCL Hackery6 Message-ID: <8dnh0g$n89$1@mailint03.im.hou.compaq.com>   In article <C16B19ADF2A3D111882C00805FE6C65004FF365D@xch-f22bh-01.ds.boeing.com>, "Wolf, Gerald J" <Gerald.Wolf@F22.Boeing.com> writes:eF :Is there a DCL procedure out there to disuser users who have passwordK :changed is older than a specified amount of time? Is there a way to notify M :these users before they are disusered using the above method.  We would like ( :to do this on a daily or monthly basis.  K   The existing password expiration mechanism handles most of this task now,OH   though it does allow one shot at login after the password has expired.H   (I know of sites that use this mechanism, in conjunction with a local H   executable image, to regularly update the username expiration date to 5   a specified "delta" from the last password change.)-  H   The easiest (and officially supported) approach for the stated problemF   is via a fairly simple executable image and calls to sys$getuai and I   sys$setuai -- there is no supported interface for access to the system rH   authorization database from DCL, no f$getuai nor f$setuai.  There are G   tools (executable images) that allow DCL to get at the authorization dH   database by command input (for control and input) and the creation of G   symbols (for output back to DCL), and there are hacks that either a: rG   directly access SYSUAF, or b: parse output from AUTHORIZE commands.  wD   (Neither of these latter two hacks is particularly difficult, and '   neither is particularly supported...)   K :Please include the procedure if you can.  Is there a place on the internetE! :with contributed DCL procedures?g  >   There are many spots where you can pick up DCL examples, and.   contributions to the Freeware are welcome...  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Thu, 20 Apr 2000 11:29:10 -0700O, From: Howard S Shubs <hshubs@mindspring.com> Subject: Re: DCL Hackery> Message-ID: <hshubs-199549.11291020042000@news.mindspring.com>  7 In article <8dnh0g$n89$1@mailint03.im.hou.compaq.com>, .$ hoffman@xdelta.zko.dec.nospam wrote:  E >  (Neither of these latter two hacks is particularly difficult, and -( >  neither is particularly supported...)  % Yeah, but they're kinda fun to write.b   --   Howard S Shubs, the Denim Adept.   ------------------------------  % Date: Thu, 20 Apr 2000 14:34:57 -0500 + From: Don Rogstad <Don.Rogstad@dalsemi.com>d Subject: RE: DCL HackeryH Message-ID: <EA5FE99F5DEAD311A6CB00805F199992BCA80B@misnts1.dalsemi.com>  8This reminds me of one complaint I have with the GETUAI system service.  There is no way to specify wildcard selection of users.  Has this or can this be changed?  Currently, I have procedures for reporting on users, but have to run authorize to get a list of users that I then read back in for the GETUAI calls.  ~    I think many would like to see that enhancement, which could cut out a lot of the "hacks" to get information out of SYSUAF.   Don RogstadY Dallas Semiconductor   -----Original Message-----# From: hoffman@xdelta.zko.dec.nospamm& [mailto:hoffman@xdelta.zko.dec.nospam]& Sent: Thursday, April 20, 2000 1:07 PM To: Info-VAX@Mvb.Saic.Com- Subject: Re: DCL Hackery       In article <C16B19ADF2A3D111882C00805FE6C65004FF365D@xch-f22bh-01.ds.boeing.com>, "Wolf, Gerald J" <Gerald.Wolf@F22.Boeing.com> writes:2F :Is there a DCL procedure out there to disuser users who have passwordK :changed is older than a specified amount of time? Is there a way to notifydM :these users before they are disusered using the above method.  We would likeI( :to do this on a daily or monthly basis.  K   The existing password expiration mechanism handles most of this task now,hH   though it does allow one shot at login after the password has expired.H   (I know of sites that use this mechanism, in conjunction with a local H   executable image, to regularly update the username expiration date to 5   a specified "delta" from the last password change.)   H   The easiest (and officially supported) approach for the stated problemF   is via a fairly simple executable image and calls to sys$getuai and I   sys$setuai -- there is no supported interface for access to the system  H   authorization database from DCL, no f$getuai nor f$setuai.  There are G   tools (executable images) that allow DCL to get at the authorization CH   database by command input (for control and input) and the creation of G   symbols (for output back to DCL), and there are hacks that either a: 0G   directly access SYSUAF, or b: parse output from AUTHORIZE commands.  iD   (Neither of these latter two hacks is particularly difficult, and '   neither is particularly supported...)-  K :Please include the procedure if you can.  Is there a place on the internet ! :with contributed DCL procedures?   >   There are many spots where you can pick up DCL examples, and.   contributions to the Freeware are welcome...  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 20 Apr 2000 19:53:11 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: RE: DCL Hackery6 Message-ID: <8dnn77$psu$1@mailint03.im.hou.compaq.com>  v In article <EA5FE99F5DEAD311A6CB00805F199992BCA80B@misnts1.dalsemi.com>, Don Rogstad <Don.Rogstad@dalsemi.com> writes:O :This reminds me of one complaint I have with the GETUAI system service.  Thereo4 :is no way to specify wildcard selection of users..." :Has this or can this be changed?      No and yes, respectively.t  7 :   I think many would like to see that enhancement,...   M   I would encourage a formal request via the CSC or other similar contact(s). J   (I would personally like to see this change, but I've got several other I   major projects that are completely clobbering any chances of my adding RJ   this or many other changes to my own "midnight engineering" schedule...)  K :which could cut out a lot of the "hacks" to get information out of SYSUAF.0  -   That, and f$getuai and f$setuai lexicals....  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 21 Apr 2000 11:55:49 +1200V* From: "Ryan C. Price" <pricerc@ihug.co.nz> Subject: Re: DCL Hackery* Message-ID: <8do5ek$noc$1@news.ihug.co.nz>  ( no quite DCL, but from some years ago...  E this crude but effective 'C' code on a VAX to do a report on passwordeE ages. Accessed SYSUAF directly, using a uafdef.h IIRC hacked from one F of the MACRO libraries, but it was 1994 that I last touched it. I usedB this basic code as a base for a few other UAF reports as well, nowF lost. Some bits could probably be improved for Alpha (e.g. the typedef quad shouldn't be necessary).i   /* uafreport.c */d #include <ctype.h> #include <string.h>u #include <unixio.h>  #include <file.h>o #include <lib$routines.h>  #include <libdtdef.h>- #include <stdlib.h>- #include <stdio.h> #include <starlet.h> #include "uafdef.h"l   typedef unsigned long quad[2] ;i   int main(int argc, char **argv)  {k     union UAFDEF             *uaf_rec;s  6     int     uaf_file;       /* file descriptor for uaf */;     FILE   *rep_file;       /* file descriptor for rep filei */  &     register unsigned long int status;8     unsigned long int idiff, weeks_allowed, delta_days ;=     quad today, pw_change_by, pw_changed, pw_lifetime, diff ;:  B     printf("scanning for users whose passwords have expired.\n") ;       uaf_file =C open("SYSUAF",O_RDONLY,0,"ctx=rec","shr=get","shr=put","shr=upd") ;        if ( uaf_file == -1 )h           perror("UAFopen") ;,         exit(0x10000004) ;     }D  -     rep_file = fopen("PASSWORD_REPORT","w") ;      if ( rep_file == NULL )o           perror("REPopen") ;0         exit(0x10000004);C     }   "     status = sys$gettim (&today) ;#     delta_days = LIB$K_DELTA_DAYS ;w  .     uaf_rec = calloc(1,sizeof(union UAFDEF)) ;  >     while( read(uaf_file,uaf_rec,sizeof(union UAFDEF)) !=0 ) {  	         {A             int i;             for(i=0; i<12; i++)0: if(uaf_rec->uaf$t_uafdef.uaf$t_username[i] == ' ') break ;9             uaf_rec->uaf$t_uafdef.uaf$t_username[i] = 0 ;R	         }N  E         memcpy(pw_changed,&(uaf_rec->uaf$t_uafdef.uaf$q_pwd_date),8);c         if ( pw_changed == 0) {P>             fprintf(rep_file,"** password is pre-expired\n") ;	         }O           else {  B memcpy(pw_lifetime,&(uaf_rec->uaf$t_uafdef.uaf$q_pwd_lifetime),8);             status =3 lib$add_times(pw_changed,pw_lifetime,pw_change_by);b<             status = lib$sub_times(pw_change_by,today,diff);             status =4 lib$cvt_from_internal_time(&delta_days,&idiff,diff);             if(idiff<2) { !                 fprintf(rep_file,ZE                         "%-12s:%3d days till next password change\n",d=                         uaf_rec->uaf$t_uafdef.uaf$t_username,-                         idiff); 
             } 	         }-     }-   close(uaf_file);   fclose(rep_file);  }  /* end of uafreport.c */     /Ryanl  ? "Hoff Hoffman" <hoffman@xdelta.zko.dec.nospam> wrote in message 0 news:8dnh0g$n89$1@mailint03.im.hou.compaq.com... >C > In articleF <C16B19ADF2A3D111882C00805FE6C65004FF365D@xch-f22bh-01.ds.boeing.com>,5 "Wolf, Gerald J" <Gerald.Wolf@F22.Boeing.com> writes: ? > :Is there a DCL procedure out there to disuser users who have  passwordF > :changed is older than a specified amount of time? Is there a way to notifyD > :these users before they are disusered using the above method.  We
 would like* > :to do this on a daily or monthly basis. >eC >   The existing password expiration mechanism handles most of this 	 task now,=A >   though it does allow one shot at login after the password hasT expired.C >   (I know of sites that use this mechanism, in conjunction with a- local F >   executable image, to regularly update the username expiration date to7 >   a specified "delta" from the last password change.)T >LB >   The easiest (and officially supported) approach for the stated problemmC >   is via a fairly simple executable image and calls to sys$getuaiC and=C >   sys$setuai -- there is no supported interface for access to the  systemE >   authorization database from DCL, no f$getuai nor f$setuai.  There  areL: >   tools (executable images) that allow DCL to get at the
 authorizationNF >   database by command input (for control and input) and the creation ofE >   symbols (for output back to DCL), and there are hacks that either  a:= >   directly access SYSUAF, or b: parse output from AUTHORIZEo	 commands. E >   (Neither of these latter two hacks is particularly difficult, ande) >   neither is particularly supported...)e >eD > :Please include the procedure if you can.  Is there a place on the internet# > :with contributed DCL procedures?  > @ >   There are many spots where you can pick up DCL examples, and0 >   contributions to the Freeware are welcome... >-, >  --------------------------- pure personal# opinion ---------------------------t1 >    Hoff (Stephen) Hoffman   OpenVMS Engineering  hoffman#xdelta.zko.dec.com >    ------------------------------  % Date: Fri, 21 Apr 2000 01:40:26 -0400s0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca> Subject: Re: DCL Hackery/ Message-ID: <38FFE9BB.14F3A0C3@vl.videotron.ca>    "Wolf, Gerald J" wrote:o > G > Is there a DCL procedure out there to disuser users who have password L > changed is older than a specified amount of time? Is there a way to notifyN > these users before they are disusered using the above method.  We would like) > to do this on a daily or monthly basis.   K The following would disuser any user whose UIC group is greater than 10 andi9 whose password had not been changed in the past 3 months.-     $ALLIN1/NOINIT   CAL SET_DATE TODAY -3M% DATE_CONVERT #BREAKDAY, OA$TM_DATE, 7   " FOR UAI$ WITH .UICGROUP GT 10 DO -* 	DATE_CONVERT #PWDCHANGE, .PWDCHANGE,7 \\-$ 	.IF #PWDCHANGE LTS #BREAKDAY THEN -( 		WRITE CHANGE UAI$ %KEY=.%KEY,DISUSER=1    M You could do the same type of stuff to warn users by email one month prior torJ their account being zpapped. What I had developped were monthly accountingK scripts that emailed to each user/custoemr a statement showing each sessionuI (time,date, session duration), as well as password information (date lastoT changed, date it will next expire, and warnings if it will expire inside of a week).! This is all quite easy in ALLIN1.M   ------------------------------  % Date: Thu, 20 Apr 2000 14:03:28 -0400 " From: Dan Sugalski <dan@sidhe.org> Subject: Re: EMC Disk Storageh8 Message-ID: <4.3.1.0.20000420140139.01d2b7e0@24.8.96.48>  + At 01:12 PM 4/20/00 -0400, Bill Todd wrote:aL >One might guess that EMC-style checks (which 'way back when mostly involvedM >drive 'scrubbing' that scanned its way through the entire drive sequentiallyaI >during idle periods to make sure data could still be read without heroiciM >efforts) have somewhat less value than they used to:  current drives performtJ >much more self-monitoring than earlier ones and tend to catch the kind ofL >gradual degradations (e.g., increasing re-tries) that can warn of impendingI >catastrophic failure.  They won't catch localized surface degradation iniL >areas left untouched for long periods, but that type of failure is less aptJ >to escalate suddenly to catastrophic levels and, as Dan said, that's what >redundancy is for anyway. >-E >Of course, it wouldn't kill StorageWorks to include background scansr
 >either...  L I'm not sure I'd like the controllers to do that too often--it's extra wear / on the drives for not all that much extra gain.,   					Dan  L ----------------------------------------------------------------------------L Dan Sugalski                          General and VMS-specific perl training
 dan@sidhe.org >                                       Mail me for more details   ------------------------------  # Date: Thu, 20 Apr 2000 21:59:05 GMTt From: itjck01@my-deja.com  Subject: Re: EMC Disk Storagel) Message-ID: <8dnuj0$frk$1@nnrp1.deja.com>E  4 In article <852568C6.00688C18.00@mail2.us.hsbc.com>,1   "Paul M Boynton" <pmboynton@us.hsbc.com> wrote:l >i >oH > We are in the process of evaluating EMC disk storage to replace Compaq > StorageWorksE > controllers and disk on an OpenVMS CI cluster. Does anyone have any 
 experience$ > good or bad with EMC disk storage? >m > Paul Boynton > HSBC Bank, USA
 > Buffalo, NY  > Pmboynton@us.hsbc.com  >n  E My humor may be odd... but it is kind of funny to me that our EMC guyiG is here tonight to take care of the other end of EMC things, as in tapes drives.   D The EMC P3000 ATL we have has no tape drives online.  It has lost it- brains so to speak, even after a power cycle.   G I guess my point is you should also look at how you are going to backup E your disk space and with whom, and consider your options, at the samer- time you consider moving to new disk storage.   D The two might be highly related when considering a fully operational system.a   :) jck JKoska@bender.com.    & Sent via Deja.com http://www.deja.com/ Before you buy.:   ------------------------------  % Date: Thu, 20 Apr 2000 15:30:23 -0400n' From: Marc Lippmann <marc@lippmann.com>  Subject: Re: excursion8 Message-ID: <2fmufsskp52vinr1qd3dkj2n17sgo5gdvs@4ax.com>  D Just FYI, the latest release of Excursion is V7.1.n and is extremelyD robust. I use it to a *wide* variety of UNIX platforms and have beenE really pleased with it. There's an upgrade available if you're reallyo3 licensed.  It's part of the Pathworks kit now. See n/ http://www.digital.com/info/QA00H9/QA00H9HM.HTMt   --Marc  . On Tue, 18 Apr 2000 17:00:33 GMT, Dave Pickles! <davep@nugate.demon.co.uk> wrote:i   >Bob Ricci wrote:MN >> Does anyone have compaq's excursion 3.0 for windows nt 4.0 running and ableL >> to connect to a Sun solaris 2.6 server. I keep gettingan error adding the >> font path0 >> couldnt add font path /etc/dt/config/xfonts/C> >>                                     /usr/dt/config/xfonts/C8 >>                                    tcp/cliff.com:7100G >> it then brings up the CDE desktop to Solaris asking for username andc. >> password but doesnt recognize the password.) >> i must have an invalid path somewhere!< >> bob ricci >> ricci_r@subway.com  > H >I get these errors (four in total I think) when connecting to a SolarisJ >box, but there is no problem logging in and no obvious font problems with  >the CDE environment afterwards. >i? >What messages appear in the Solaris logs (/var/adm/authlog andoI >/var/log/syslog)? Can you connect via a telnet session from the same PC,f@ >and can you run a CDE session from any other PC or workstation? >p
 >Dave Picklesi   ------------------------------  % Date: Thu, 20 Apr 2000 16:13:23 -0400a" From: Dan Sugalski <dan@sidhe.org> Subject: Re: excursion8 Message-ID: <4.3.1.0.20000420161159.01d71c60@24.8.96.48>  / At 03:30 PM 4/20/00 -0400, Marc Lippmann wrote:pE >Just FYI, the latest release of Excursion is V7.1.n and is extremelyaE >robust. I use it to a *wide* variety of UNIX platforms and have beenp >really pleased with it.  K Have they got the thing doing anything better than 8-plane color? That was  8 the single most annoying thing about the older versions.   					Dan  I --------------------------------------"it's like this"-------------------r2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and even ;                                       teddy bears get drunk    ------------------------------  # Date: Thu, 20 Apr 2000 19:37:27 GMTl* From: Gord Coulman <gcoulman@ccinet.ab.ca>5 Subject: Re: Infoserver CD-R (was: Verify of Backups)w, Message-ID: <38FF5581.7308EB13@ccinet.ab.ca>  _ I always thought LAD/LAST was somehow tied to Decnet.  I may just dust off that Infoserver now!.  ` As suggested, I checked the VMS FAQ for CD writing techniques.  The FAQ pointed to the followingT excellent web page:  http://www.geocities.com/SiliconValley/Lakes/9999/vmscdwri.html  ` One method by Bystrik Krchnavy suggests writing to a hard disk partition on the Infoserver, then` mounting the Infoserver partition from a PC equipped with a CD writer, then burning a "raw" copy` of the partition to the CD.  Just one catch: how do you mount an Infoserver partition from a PC?P I hope it doesn't involve anything starting with "Path" and ending with "Works".   Gord.s   Larry Kilgallen wrote:  [ > In article <38FF2B85.5B3B13F1@ccinet.ab.ca>, Gord Coulman <gcoulman@ccinet.ab.ca> writes:-Z > > Regarding your suggestion of using an Infoserver to create VMS-readable CDs, I have an_ > > Infoserver 150 that I no longer use.  It had a tape drive, a CDROM, and a 1GB internal diskm >o! > An Infoserver 1000 is required.i >I` > > on it.  I don't really like the idea of dusting it off, because wouldn't I require Decnet toO > > use it?  I only run TCP/IP and really don't want to implement Decnet again.r >bC > Infoservers do not use DECnet at all.  They speak LAD/LAST across  > Ethernet to VMS.   ------------------------------  % Date: Thu, 20 Apr 2000 12:51:32 -0500 % From: Chris Scheers <asi@airmail.net>t: Subject: Interesting problem with AS 255/233 and JAZ driveO Message-ID: <5468330EAD90A82A.C704A4540C3C7D1A.CDA268DEB4770045@lp.airnews.net>@  G I have a AlphaStation 255/233 which has been running cheerfully with anX@ S3 video card, 96MB memory, and a JAZ drive for the system disk.  A The JAZ has been very convenient for being able to load different,F versions of VMS, different combinations of layered products, networks,- etc.  This machine is used as a test machine.d  D I replaced the S3 video card with a ZLXp-L1 and ran into a problem. H With VMS 7.1 and Open3D 4.5, the system would lock up with the JAZ drive9 going in and out of mount verifications with SCSI resets.   D It appears that the ZLXp-L1 pushed the memory requirements up to theG point where the system started swapping.  The SWAP I/O to the JAZ driveeG was failing for some reason, and processes were getting stuck in RWMPB.a  H The error log shows SCSI TIMEOUTs on the JAZ drive.  Since the error log3 is also on the JAZ drive, not very much got logged..  D Interestingly, if I induce a crash while this is going on, the crash' dump writes to the JAZ drive correctly.l  F Replacing the JAZ drive with a hard drive allows the processes to swapF and the system to continue execution.  (Of course this isn't nearly as( useful for testing out software combos.)  D What is fundamentally different about swap I/O that could cause this= problem?  Are there any patches for VMS 7.1 for this problem?r  - (I haven't run these tests with VMS 7.2 yet.)n   Thanx!  G ----------------------------------------------------------------------- $ Chris Scheers, Applied Synergy, Inc.  G 817-237-3360 (Voice)    817-237-3074 (Fax)    Internet: asi@airmail.netu   ------------------------------  % Date: Thu, 20 Apr 2000 15:33:16 -0400 ' From: Marc Lippmann <marc@lippmann.com>e/ Subject: Re: LK46W-A2 & "DO" key in Reflectionse8 Message-ID: <cnmufsotcfmoaulnfog5sg752e95b4d4pq@4ax.com>  E heh heh heh. Hey, Norm.  Darren, on the current version of ReflectioniE for UNIX and Digital (V7) you'll find it under the SETUP/Keyboard MapnF menu option. On older versions, the same functionality is there, but ID don't remember if the menu location was identical. Then you can save9 the profile which will make the mapping change permanent.n  E On Tue, 18 Apr 2000 13:48:46 -0400, norm.raphael@jamesbury.com wrote:B   >F >9 >C >B >Hi, >dE >     Can anyone tell me how to map the "DO" key in reflections so itdI >works under a TPU OpenVMS 7.1 session, also the F17/18/19/20 keys do notoI >appear to send the appropriate escape sequence.  I have been through theaK >Keyboard map in reflections and mapped the PC keyboard to the US DEC LK450t" >keyboard but it still won't work. >TIA	 >- DarrenrM >---------------------------------------------------------------------------- C >------------------------------------------------------------------r< >Darren James Boyle,              Telephone : (441) 299-6425 >Lead Analyst Programmer,n; >ISD VMS Technical Services,  Fax          : (441) 299-6502a >Bank of BermudaK >Par-la-Ville Branch,                EMail      : BoyleDJ@BankofBermuda.comi >Hamilton Bermuda.M >---------------------------------------------------------------------------- C >------------------------------------------------------------------  >-: >I always just use F16=DO by <Shift-F6> on my PC keyboard. >  >0 >    ------------------------------  # Date: Thu, 20 Apr 2000 19:30:51 GMTs' From: Colin Blake <colin@theblakes.com> ! Subject: Re: Mozilla M15's out...g- Message-ID: <38FF5AE7.C66AD084@theblakes.com>V   Peter LANGSTOEGER wrote:  : > GTK+ is some of kind of a [graphics] tool kit. Correct ?  8 Yes. Mozilla doesn't use MOTIF as its GUI, it uses GTK+.  , > It is a requirement for MOZILLA. Correct ?   Yes.  K > On VMS, GTK+ is not available standalone or VMS users are more uninformeddJ > than on the other platforms and so the GTK+ gets installed automatically3 > in the background for and with MOZILLA. Correct ?c  ] GTK+ is not available for VMS. Not as a separate product, anyway. As part of the Mozilla porte\ GTK+ has also been ported and work is underway to get the changes checked back into the GTK+[ source pool. Once that has happened I would expect GTK+ to be listed as a pre-requisite forf^ Mozilla/Netscape6 and for there to be a separate kit for it. Unless then it ships bundled into the Mozilla kit.  7 > Why GTK+ doesn't get a separate (PCSI) product then ?u  
 See above.  I > >The other kits are in zip format. If you zip the OpenVMS PCSI file its B > >only 17 Mbytes instead of 30Mb for the self-extracting version. >eG > This raises again the question, why DCXEXE ? Use ZIP or [UN]ZIPSFX...0  [ Because some people don't like being given kits when some other tool that isn't part of theA? standard OpenVMS distribution is required to "unpack" that kit.f   ------------------------------    Date: 20 Apr 2000 21:45:30 +0200* From: eplan@kapsch.net (Peter LANGSTOEGER)! Subject: Re: Mozilla M15's out...g( Message-ID: <38ff5e5a@news.kapsch.co.at>  W In article <38FF5AE7.C66AD084@theblakes.com>, Colin Blake <colin@theblakes.com> writes:u >Peter LANGSTOEGER wrote:) >0; >> GTK+ is some of kind of a [graphics] tool kit. Correct ?i >;9 >Yes. Mozilla doesn't use MOTIF as its GUI, it uses GTK+.b >k- >> It is a requirement for MOZILLA. Correct ?r >e >Yes.  > L >> On VMS, GTK+ is not available standalone or VMS users are more uninformedK >> than on the other platforms and so the GTK+ gets installed automaticallys4 >> in the background for and with MOZILLA. Correct ? > ^ >GTK+ is not available for VMS. Not as a separate product, anyway. As part of the Mozilla port] >GTK+ has also been ported and work is underway to get the changes checked back into the GTK+b\ >source pool. Once that has happened I would expect GTK+ to be listed as a pre-requisite for_ >Mozilla/Netscape6 and for there to be a separate kit for it. Unless then it ships bundled intod >the Mozilla kit.o >c8 >> Why GTK+ doesn't get a separate (PCSI) product then ? >$ >See above.    Sounds good. Thanks.  J >> >The other kits are in zip format. If you zip the OpenVMS PCSI file itsC >> >only 17 Mbytes instead of 30Mb for the self-extracting version.< >>H >> This raises again the question, why DCXEXE ? Use ZIP or [UN]ZIPSFX... >h\ >Because some people don't like being given kits when some other tool that isn't part of the@ >standard OpenVMS distribution is required to "unpack" that kit.  & Did you see "...SFX" in my statement ?  M UNZIPSFX is an .EXE to be copied into the ZIP file to get a selfunpacking kit J just like many .EXEs on the PC (which brave users don't run, but trying toH UNZIP the .EXE before, just to avoid virii, or if it won't work do otherJ security checks instead of "clicking .EXEs to death"). And therefore, thisN people get what they want (I like .ZIP much more than .EXE - as I wrote above)   -- T< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888y< FBFV/Information Services           E-mail  eplan@kapsch.netF <<< KAPSCH AG  Wagenseilgasse 1     PSImail PSI%(0232)281001141::EPLANH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"N "VMS is today what Microsoft wants Windows NT V8.0 to be!" Compaq, 22-Sep-1998   ------------------------------    Date: 20 Apr 2000 20:41:19 -0500/ From: jlauret@?.chem.sunysb.edu (Jerome LAURET) ! Subject: Re: Mozilla M15's out... . Message-ID: <38ffa3af_3@dilbert.ic.sunysb.edu>  F 	I confirm : Mozilla M15 is not working with Multinet TCPIP ... I haveN sent an Email to the Multinet group but basically, web pages are not displayed) at all ... as repoted by Dan O'Reilly ...f 	 D 	So, if you have Multinet TCPIP, beware of the upgrade (it will look more like a downgrade :{ ).  	t     -- r6                   Jerome LAURET S.U.N.Y. @ Stony Brook$        ,,,,,      Dept. of Chemistry+       ( o o )     Stony Brook NY 11794-3400d;   ---m---U---m---------------------------------------------L&   E-mail: jlauret@mail.chem.sunysb.edu<   URL   : http://nucwww.chem.sunysb.edu/jlauret/jlauret.html   ------------------------------  % Date: Thu, 20 Apr 2000 15:51:25 -0400f1 From: Marc Lippmann <marc@lippmann.spamsucks.com>c Subject: Need BLISS advice8 Message-ID: <2qnufssusv7l2kdn2q5g0hem2df8ojg2b0@4ax.com>  D I'm trying to compile legacy VAX BLISS code on an Alpha. A number ofC L32 files were missing and I've located almost all of them, but I'm"A hitting a stone wall on TPAMAC.L32. Anyone know what this is, and - where I might find a copy for Alpha?  Thanks!p   ------------------------------  % Date: Thu, 20 Apr 2000 16:16:12 -0400 , From: Steve Lionel <Steve.Lionel@compaq.com> Subject: Re: Need BLISS advice8 Message-ID: <9bpufsgnf8hqrvv6mrbkmpdcsf8heh8i3s@4ax.com>  1 On Thu, 20 Apr 2000 15:51:25 -0400, Marc Lippmanno$ <marc@lippmann.spamsucks.com> wrote:  E >I'm trying to compile legacy VAX BLISS code on an Alpha. A number ofAD >L32 files were missing and I've located almost all of them, but I'mB >hitting a stone wall on TPAMAC.L32. Anyone know what this is, and. >where I might find a copy for Alpha?  Thanks!  A It's definitions for LIB$TPARSE.  There should be a TPAMAC.REQ inh2 SYS$LIBRARY: - just compile it with BLISS/LIBRARY.    - Steve Lionel (mailto:Steve.Lionel@compaq.com)a Fortran Engineeringt& Compaq Computer Corporation, Nashua NH  6 Compaq Fortran web site: http://www.compaq.com/fortran   ------------------------------   Date: 20 Apr 2000 20:51:43 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: Need BLISS advice6 Message-ID: <8dnqkv$rnt$1@mailint03.im.hou.compaq.com>  l In article <2qnufssusv7l2kdn2q5g0hem2df8ojg2b0@4ax.com>, Marc Lippmann <marc@lippmann.spamsucks.com> writes:E :I'm trying to compile legacy VAX BLISS code on an Alpha. A number ofOD :L32 files were missing and I've located almost all of them, but I'mB :hitting a stone wall on TPAMAC.L32. Anyone know what this is, and. :where I might find a copy for Alpha?  Thanks!  B   lib$table_parse is the OpenVMS Alpha interface for lib$tparse...<   You may/will end up with a little recoding in this area...  @   As was suggested elsewhere, you can build it yourself from the!   (provided) TPAMAC.REQ module...e  ?   The same basic steps that are used to rebuild STARLET.REQ and 9   LIB.REQ for use by Bliss will work here for TPAMAC.REQ:l  D $ BLISS SYS$LIBRARY:STARLET.REQ /LIBR=SYS$COMMON:[SYSLIB]STARLET.L32H $ BLISS SYS$LIBRARY:STARLET.REQ+LIB.REC /LIBR=SYS$COMMON:[SYSLIB]LIB.L32    N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------   Date: 20 Apr 2000 16:16:54 PDTT From: Fairfield@SLAC.Stanford.EDU (Ken Fairfield; SLAC: 650-926-2924; FAX: 926-3515)3 Subject: Re: OpenVMS mail question: Attached files? 3 Message-ID: <DxLLQg1ujUZH@mccdev.slac.stanford.edu>h  , In article <38ff1cd9.8146323@news.wku.edu>, 4     	goathunter@PROCESS.COM (Hunter Goatley) writes: [...]aB > The funtionality is limited to one file, but with MultiNet's and; > TCPware's SMTP implementation, you can send one file as aeF > base64-encoded MIME attachment.  Even with the limitation, it can be  > very handy.  It's invoked via: > / >     MAIL> send/noedit/foreign/type=1 file.extt  H         I (we) have known about the undocumented send/foreign within VMSH     Mail for quite a long  time.   However,  what is the "/type=1"?  AndF     does "file.ext" need to be base64 encoded _first_, or does /type=14     request base64 encoding?  More details please?             Thanks, Kenn -- fM  Kenneth H. Fairfield            |  Internet: Fairfield@SLC.Slac.Stanford.EduE:  SLAC, 2575 Sand Hill Rd, MS 46  |  Voice:    650-926-2924:  Menlo Park, CA  94025           |  FAX:      650-926-3515N  -----------------------------------------------------------------------------B  These opinions are mine, not SLAC's, Stanford's, nor the DOE's...   ------------------------------  % Date: Fri, 21 Apr 2000 10:25:11 +0930a/ From: Mark Daniel <mark.daniel@wasd.vsm.com.au>i) Subject: OUTCOME: DECC 6.n BUG? /OPTIMIZEj/ Message-ID: <38FFA6EF.564B6BEB@wasd.vsm.com.au>p  J The mainboard was replaced midweek (system began crashing :^) - problem is no longer present.  H Subtle h/w problem that (in hindsight) probably began manifesting itselfG around about the time the VMS SETI v2 was released (that's when I first $ started to notice subtle anomalies).  = Thanks to all those that took the time to try my snippet out.X   Mark Daniel wrote:  E > /OPTIMIZE seems to BREAK SIMPLE INTEGER ARITHMETIC in the followingo > environment: >a- >   DEC 3000 Model 500 running OpenVMS V7.2-1d- >   Compaq C V6.2-003 on OpenVMS Alpha V7.2-1  >n@ > The included demonstration of the problem produces the resultsE > following the code, vastly different with /OPT (and incorrect) thanrC > /NOOPT (correct).  Essentially 1,100,000 divided by 1,000,000 cansD > produce a result of 0!!  It is not loop unrolling that seems to beE > doing it either.  The same problem (and others seemingly) occurs in C > "straight" sections of code (one of which originally caused me to C > pursue this).  Upgrading from DECC 6.0 to 6.2 did not resolve theo > issue. >fB > Bringing an .EXE compiled using DECC 5.6 under VMS 7.1 onto thisE > system produces correct results so it can't be the CRTL; points theaH > finger at 6.n.  Makes me a bit wary of using /OPT if it can break code  > in such a fundamental fashion. >bG > Apologies for the (linear) length of this post.  Any comments please.t >a > TIA.  Mark Daniel. >a+ > -----------------------------------------n > #include <stdio.h> >n > main() > {a+ >    unsigned long  Count, Whole, Fraction;n >u@ >    for (Count = 0; Count < (unsigned long)0xffffffff; Count++) >    {  >       Whole = Count / 1000000;+ >       Fraction = (Count % 1000000) / 100;m" >       if ((Count % 100000) == 0)B >          fprintf (stdout, "%d %d %d\a\n", Count,Whole,Fraction); >    } > }t+ > -----------------------------------------p >J > $ cc testl
 > $ link testf > $ run test > 0 0 0d > 100000 0 999 > 200000 0 1999  > 300000 0 2999t > 400000 0 3999m > 500000 0 4999  > 600000 0 5999i > 700000 0 6999e > 800000 0 7999o > 900000 0 8999e > 1000000 0 9999    <--- > 1100000 0 10999   <--- > 1200000 0 11999   <--- > 1300000 0 12999   <--- > 1400000 1 3999 > 1500000 1 4999 > 1600000 1 5999 > 1700000 1 6999 > 1800000 1 7999 > 1900000 1 8999 > 2000000 1 9999    <--- > 2100000 1 10999   <--- > 2200000 1 11999   <--- > 2300000 1 12999   <--- > 2400000 1 13999   <--- > 2500000 1 14999   <--- > 2600000 1 15999   <--- > 2700000 1 16999   <--- > 2800000 2 7999 > 2900000 2 8999 > 3000000 2 9999 > 3100000 2 10999I > blah blahR >a > $ cc test /noopt
 > $ link test  > $ run test > 0 0 0e > 100000 0 1000a > 200000 0 2000S > 300000 0 3000d > 400000 0 4000y > 500000 0 5000a > 600000 0 6000. > 700000 0 7000  > 800000 0 8000  > 900000 0 9000g
 > 1000000 1 0h > 1100000 1 1000 > 1200000 1 2000 > 1300000 1 3000 > 1400000 1 4000 > 1500000 1 5000 > 1600000 1 6000 > 1700000 1 7000 > 1800000 1 8000 > 1900000 1 9000
 > 2000000 2 0  > 2100000 2 1000 > 2200000 2 2000 > 2300000 2 3000 > 2400000 2 4000 > 2500000 2 5000 > 2600000 2 6000 > 2700000 2 7000 > 2800000 2 8000 > 2900000 2 9000
 > 3000000 3 0a > 3100000 3 1000 > 3200000 3 2000   ------------------------------  % Date: Thu, 20 Apr 2000 15:36:18 -0400h' From: Marc Lippmann <marc@lippmann.com>a, Subject: Re: Problem reading files from tape8 Message-ID: <eumufs4sj8u1t2qv0tpid4g9ttofkkg5rv@4ax.com>  C Gerke, if Hoff's advice, which I completely agree with, doesn't panrB out, there are data recovery services which often can recover dataE from damaged tapes. They tend to be EXTREMELY expensive, so one wouldnD only consider it if the data were awfully valuable. If you choose to7 go that route, email me; I'll send you my contact info.e  @ On 19 Apr 2000 13:04:38 GMT, hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) wrote:r   > g >In article <01bfa9f8$5abccf10$8d4c15ac@HKTGN9911301604>, "Gerke Grashuis" <g.grashuis@kpn.com> writes:-F >:I'm using a TZ89 drive and trying to restore a great number of filesL >:(112.000 files) with backup/select from a saveset on a tape in that drive.K >:Doing this, the tape always stops responding at exactly the same file (ata >:approx. 85.000) files.I >:The error rate on the drive starts increasing and the status changes tor >:MountVerify. >:What can I do to tackle this?s > J >  Given your comments, I will assume this is OpenVMS VAX or OpenVMS Alpha >  environment.e >cK >  I would first look for BACKUP and mandatory ECO kits for the particular eK >  OpenVMS version in use, and I would also check the system error log for oH >  messages, and I would then contact your hardware support organizationH >  as this initially looks like a tape media or tape drive or tape driveK >  firmware or SCSI error (length, termination, cabling, etc), and (if the  J >  hardware is correct and at current revision) I would next contact your 0 >  software support organization for assistance. >MO > --------------------------- pure personal opinion --------------------------- M >   Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.comu   ------------------------------  % Date: Thu, 20 Apr 2000 20:44:18 +0200c> From: "Jean-Franois Marchal" <jean-francois.marchal@x9000.fr> Subject: queue status values+ Message-ID: <8dnjok$81s$1@minus.oleane.net>s  	 Bonjour !   F I'm going to use the queue status of a print queue in a DCL procedure.H The information displayed by a simple show queue is sufficient (stopped,
 idle, ...)7  I got the following values for the QUEUE_STATUS item :M $.& $ ALIGNING                = %X00000001& $ IDLE                    = %X00000002& $ PAUSED                  = %X00000010& $ PAUSING                 = %X00000020& $ REMOTE                  = %X00000040& $ RESETTING               = %X00000080& $ RESUMING                = %X00000100& $ STALLED                 = %X00000400& $ STARTING                = %X00000800& $ STOPPED                 = %X00001000& $ STOPPING                = %X00002000& $ UNAVAILABLE             = %X00004000& $ CLOSED                  = %X00008000& $ BUSY                    = %X00010000& $ AVAILABLE               = %X00040000& $ STOP_PENDING            = %X00200000 $rF What of theses values can be set together (as AVAILABLE and STOPPED) ?J What mask could be applied to the status value to get the status name in a simple manner ?4H (using separate QUEUE_IDLE, QUEUE_STOPPED, ... need to issue many getqui/ and consolidate the TRUE/FALSE values obtained)y   Cordialement Jean-Franois MarchalT X9000 - LYON   ------------------------------  % Date: Thu, 20 Apr 2000 14:40:30 -0500t+ From: Don Rogstad <Don.Rogstad@dalsemi.com>   Subject: RE: queue status valuesH Message-ID: <EA5FE99F5DEAD311A6CB00805F199992BCA80C@misnts1.dalsemi.com>  I Below is a DCL subroutine, GET_QUE_STATUS, to return the queue status.  =-C Set the numeric queue status in the variable TEMP and it sets the =-E variable QUESTAT containing a comma separated ASCII string of states.    Don Rogstad  Dallas Semiconductor   [DCL Begins]I $!----------------------------------------------------------------------=l -------r $GET_QUE_STATUS:3 $   QStatus =3D "Aligning,Idle,Lowercase,Operator =  Request,Paused,Pausing,Remote" $   QStatus =3D QStatus + =-5 ",Resetting,Resuming,Server,Stalled,Starting,Stopped"- $   QStatus =3D QStatus + =-7 ",Stopping,Unavailable,Closed,Busy,Undefined,Available"nE $   QStatus =3D QStatus + ",Disabled,AutoStart Inactive,Stop Pending"  $   QueStat =3D "" $   Count   =3D 0o $STAT_LOOP:l+ $   If Temp .eq. 0 Then Goto Done_Stat_Loops $   Shift =3D Temp / 2 $   Bit =3D Temp - (Shift * 2) $   Temp =3D Temp / 2r. $   If Bit Then QueStat =3D QueStat + ", " + = F$Element(Count,",",QStatus) $   Count =3D Count + 1  $   Goto Stat_Loop $! $DONE_STAT_LOOP: $   QueStat =3D QueStat - ", "
 $   ReturnI $!----------------------------------------------------------------------=  -------h  
 [DCL ENDS]   -----Original Message-----C From: Jean-Frangois Marchal [mailto:jean-francois.marchal@x9000.fr] & Sent: Thursday, April 20, 2000 1:44 PM To: Info-VAX@Mvb.Saic.Comw Subject: queue status values    	 Bonjour !s  F I'm going to use the queue status of a print queue in a DCL procedure.@ The information displayed by a simple show queue is sufficient =	 (stopped,a
 idle, ...)7  I got the following values for the QUEUE_STATUS item :a $k( $ ALIGNING                =3D %X00000001( $ IDLE                    =3D %X00000002( $ PAUSED                  =3D %X00000010( $ PAUSING                 =3D %X00000020( $ REMOTE                  =3D %X00000040( $ RESETTING               =3D %X00000080( $ RESUMING                =3D %X00000100( $ STALLED                 =3D %X00000400( $ STARTING                =3D %X00000800( $ STOPPED                 =3D %X00001000( $ STOPPING                =3D %X00002000( $ UNAVAILABLE             =3D %X00004000( $ CLOSED                  =3D %X00008000( $ BUSY                    =3D %X00010000( $ AVAILABLE               =3D %X00040000( $ STOP_PENDING            =3D %X00200000 $eF What of theses values can be set together (as AVAILABLE and STOPPED) ?G What mask could be applied to the status value to get the status name =O in a simple manner ?KC (using separate QUEUE_IDLE, QUEUE_STOPPED, ... need to issue many =i getqui/ and consolidate the TRUE/FALSE values obtained)I   Cordialement Jean-Fran=E7ois Marchalr X9000 - LYON   ------------------------------  % Date: Thu, 20 Apr 2000 21:54:38 +0200o> From: "Jean-Franois Marchal" <jean-francois.marchal@x9000.fr>  Subject: Re: queue status values+ Message-ID: <8dnn83$a1h$1@minus.oleane.net>G   Merci !a  0 I will try to integrate this in my procedure ...   Cordialement Jean-Franois Marchal  X9000 - LYON    8 "Don Rogstad" <Don.Rogstad@dalsemi.com> wrote in messageB news:EA5FE99F5DEAD311A6CB00805F199992BCA80C@misnts1.dalsemi.com...K Below is a DCL subroutine, GET_QUE_STATUS, to return the queue status.  SetsF the numeric queue status in the variable TEMP and it sets the variable< QUESTAT containing a comma separated ASCII string of states.   Don Rogstado Dallas Semiconductor   [DCL Begins]L $!-------------------------------------------------------------------------- ---b $GET_QUE_STATUS:/ $   QStatus = "Aligning,Idle,Lowercase,Operator  Request,Paused,Pausing,Remote" $   QStatus = QStatus +P5 ",Resetting,Resuming,Server,Stalled,Starting,Stopped"  $   QStatus = QStatus + 7 ",Stopping,Unavailable,Closed,Busy,Undefined,Available" C $   QStatus = QStatus + ",Disabled,AutoStart Inactive,Stop Pending"t $   QueStat = "" $   Count   = 0i $STAT_LOOP:I+ $   If Temp .eq. 0 Then Goto Done_Stat_Loop  $   Shift = Temp / 2 $   Bit = Temp - (Shift * 2) $   Temp = Temp / 2hG $   If Bit Then QueStat = QueStat + ", " + F$Element(Count,",",QStatus)y $   Count = Count + 1c $   Goto Stat_Loop $! $DONE_STAT_LOOP: $   QueStat = QueStat - ", "
 $   ReturnL $!-------------------------------------------------------------------------- ---   
 [DCL ENDS]   -----Original Message-----C From: Jean-Frangois Marchal [mailto:jean-francois.marchal@x9000.fr] & Sent: Thursday, April 20, 2000 1:44 PM To: Info-VAX@Mvb.Saic.Com  Subject: queue status values    	 Bonjour !1  F I'm going to use the queue status of a print queue in a DCL procedure.H The information displayed by a simple show queue is sufficient (stopped,
 idle, ...)7  I got the following values for the QUEUE_STATUS item :  $m& $ ALIGNING                = %X00000001& $ IDLE                    = %X00000002& $ PAUSED                  = %X00000010& $ PAUSING                 = %X00000020& $ REMOTE                  = %X00000040& $ RESETTING               = %X00000080& $ RESUMING                = %X00000100& $ STALLED                 = %X00000400& $ STARTING                = %X00000800& $ STOPPED                 = %X00001000& $ STOPPING                = %X00002000& $ UNAVAILABLE             = %X00004000& $ CLOSED                  = %X00008000& $ BUSY                    = %X00010000& $ AVAILABLE               = %X00040000& $ STOP_PENDING            = %X00200000 $-F What of theses values can be set together (as AVAILABLE and STOPPED) ?J What mask could be applied to the status value to get the status name in a simple manner ?1H (using separate QUEUE_IDLE, QUEUE_STOPPED, ... need to issue many getqui/ and consolidate the TRUE/FALSE values obtained)d   Cordialement Jean-Franois Marchali X9000 - LYON   ------------------------------  % Date: Thu, 20 Apr 2000 22:10:20 +0200I> From: "Jean-Franois Marchal" <jean-francois.marchal@x9000.fr>  Subject: Re: queue status values+ Message-ID: <8dno5g$b7q$1@minus.oleane.net>h   Just a comment ...  B I'm testing on a telnet queue for which the port doesn't exist ...) The queue should report "unavailable" ... + display_queue queue_available returns false91 display_queue queue_unavailable returns false toouG your procedure doesnt display anything about availability, neither showm
 queue/full4 Maybe available is not the contrary of unavailable ?A If so, it seems to be a difference between english and french ...l   Cordialement   Jean-Franois Marchal  X9000 - LYON  : > "Don Rogstad" <Don.Rogstad@dalsemi.com> wrote in messageD > news:EA5FE99F5DEAD311A6CB00805F199992BCA80C@misnts1.dalsemi.com...H > Below is a DCL subroutine, GET_QUE_STATUS, to return the queue status. Set H > the numeric queue status in the variable TEMP and it sets the variable> > QUESTAT containing a comma separated ASCII string of states. >.
 > Don Rogstadr > Dallas Semiconductor >- > [DCL Begins] >0L $!-------------------------------------------------------------------------- > ---R > $GET_QUE_STATUS:1 > $   QStatus = "Aligning,Idle,Lowercase,Operator.  > Request,Paused,Pausing,Remote" > $   QStatus = QStatus +m7 > ",Resetting,Resuming,Server,Stalled,Starting,Stopped"I > $   QStatus = QStatus + 9 > ",Stopping,Unavailable,Closed,Busy,Undefined,Available"iE > $   QStatus = QStatus + ",Disabled,AutoStart Inactive,Stop Pending"l > $   QueStat = "" > $   Count   = 0i
 > $STAT_LOOP:I- > $   If Temp .eq. 0 Then Goto Done_Stat_Loop$ > $   Shift = Temp / 2 > $   Bit = Temp - (Shift * 2) > $   Temp = Temp / 2dI > $   If Bit Then QueStat = QueStat + ", " + F$Element(Count,",",QStatus)  > $   Count = Count + 1  > $   Goto Stat_Loop > $! > $DONE_STAT_LOOP: > $   QueStat = QueStat - ", " > $   Return >TL $!-------------------------------------------------------------------------- > ---  >$ > [DCL ENDS] >E > -----Original Message-----E > From: Jean-Frangois Marchal [mailto:jean-francois.marchal@x9000.fr]L( > Sent: Thursday, April 20, 2000 1:44 PM > To: Info-VAX@Mvb.Saic.Com- > Subject: queue status values >e >n > Bonjour !p >MH > I'm going to use the queue status of a print queue in a DCL procedure.J > The information displayed by a simple show queue is sufficient (stopped, > idle, ...)9 >  I got the following values for the QUEUE_STATUS item :  > $i( > $ ALIGNING                = %X00000001( > $ IDLE                    = %X00000002( > $ PAUSED                  = %X00000010( > $ PAUSING                 = %X00000020( > $ REMOTE                  = %X00000040( > $ RESETTING               = %X00000080( > $ RESUMING                = %X00000100( > $ STALLED                 = %X00000400( > $ STARTING                = %X00000800( > $ STOPPED                 = %X00001000( > $ STOPPING                = %X00002000( > $ UNAVAILABLE             = %X00004000( > $ CLOSED                  = %X00008000( > $ BUSY                    = %X00010000( > $ AVAILABLE               = %X00040000( > $ STOP_PENDING            = %X00200000 > $tH > What of theses values can be set together (as AVAILABLE and STOPPED) ?L > What mask could be applied to the status value to get the status name in a > simple manner ?aJ > (using separate QUEUE_IDLE, QUEUE_STOPPED, ... need to issue many getqui1 > and consolidate the TRUE/FALSE values obtained)5 >  > Cordialement > Jean-Franois Marchal- > X9000 - LYON >- >- >-   ------------------------------  % Date: Thu, 20 Apr 2000 15:42:43 -0500.+ From: Don Rogstad <Don.Rogstad@dalsemi.com>r  Subject: RE: queue status valuesH Message-ID: <EA5FE99F5DEAD311A6CB00805F199992BCA80D@misnts1.dalsemi.com>  G    According to the documentation, only one of the following "Current =a" Queue State" can be set at a time:G 	Available, Busy, Disabled, Idle, Paused, Pausing, Resuming, Stalled, =m Starting, Stopped, Stopping.  B  http://www.openvms.digital.com:8000/72final/4527/4527pro_049.html  E Available is defined as "Queue is processing work but is capable of =n processing additional work"   I Unavailable is NOT a Queue State and is defined as  "Physical device to =E) which queue is assigned is not available"n  ? So theoretically, I guess you could have both "Available" and =-E "Unavailable" set.  What string was returned?  Are any of the above =d "states" in it? =20r  H You might have to use the DEVICE_NAME item and then use F$GETDVI calls = to check for existence.    Don Rogstad) Dallas Semiconductor     -----Original Message-----C From: Jean-Frangois Marchal [mailto:jean-francois.marchal@x9000.fr] & Sent: Thursday, April 20, 2000 3:10 PM To: Info-VAX@Mvb.Saic.Com   Subject: Re: queue status values     Just a comment ...  B I'm testing on a telnet queue for which the port doesn't exist ...) The queue should report "unavailable" ...E+ display_queue queue_available returns false 1 display_queue queue_unavailable returns false too G your procedure doesnt display anything about availability, neither show/
 queue/full4 Maybe available is not the contrary of unavailable ?A If so, it seems to be a difference between english and french ...e   Cordialement   Jean-Fran=E7ois Marchal- X9000 - LYON  : > "Don Rogstad" <Don.Rogstad@dalsemi.com> wrote in messageD > news:EA5FE99F5DEAD311A6CB00805F199992BCA80C@misnts1.dalsemi.com...B > Below is a DCL subroutine, GET_QUE_STATUS, to return the queue = status.  Set0A > the numeric queue status in the variable TEMP and it sets the =u variable> > QUESTAT containing a comma separated ASCII string of states. >r
 > Don Rogstad} > Dallas Semiconductor >- > [DCL Begins] >-I $!----------------------------------------------------------------------=9 ---- > ---0 > $GET_QUE_STATUS:3 > $   QStatus =3D "Aligning,Idle,Lowercase,Operator5  > Request,Paused,Pausing,Remote" > $   QStatus =3D QStatus +07 > ",Resetting,Resuming,Server,Stalled,Starting,Stopped"  > $   QStatus =3D QStatus +29 > ",Stopping,Unavailable,Closed,Busy,Undefined,Available" G > $   QStatus =3D QStatus + ",Disabled,AutoStart Inactive,Stop Pending"9 > $   QueStat =3D "" > $   Count   =3D 00
 > $STAT_LOOP:-- > $   If Temp .eq. 0 Then Goto Done_Stat_Loop  > $   Shift =3D Temp / 2  > $   Bit =3D Temp - (Shift * 2) > $   Temp =3D Temp / 2 0 > $   If Bit Then QueStat =3D QueStat + ", " + = F$Element(Count,",",QStatus) > $   Count =3D Count + 1o > $   Goto Stat_Loop > $! > $DONE_STAT_LOOP:  > $   QueStat =3D QueStat - ", " > $   Return > I $!----------------------------------------------------------------------=0 ---- > ---  >0 > [DCL ENDS] >0 > -----Original Message-----E > From: Jean-Frangois Marchal [mailto:jean-francois.marchal@x9000.fr] ( > Sent: Thursday, April 20, 2000 1:44 PM > To: Info-VAX@Mvb.Saic.Com  > Subject: queue status values >2 >0 > Bonjour !3 >0? > I'm going to use the queue status of a print queue in a DCL = 
 procedure.B > The information displayed by a simple show queue is sufficient =	 (stopped,0 > idle, ...)9 >  I got the following values for the QUEUE_STATUS item :: > $ * > $ ALIGNING                =3D %X00000001* > $ IDLE                    =3D %X00000002* > $ PAUSED                  =3D %X00000010* > $ PAUSING                 =3D %X00000020* > $ REMOTE                  =3D %X00000040* > $ RESETTING               =3D %X00000080* > $ RESUMING                =3D %X00000100* > $ STALLED                 =3D %X00000400* > $ STARTING                =3D %X00000800* > $ STOPPED                 =3D %X00001000* > $ STOPPING                =3D %X00002000* > $ UNAVAILABLE             =3D %X00004000* > $ CLOSED                  =3D %X00008000* > $ BUSY                    =3D %X00010000* > $ AVAILABLE               =3D %X00040000* > $ STOP_PENDING            =3D %X00200000 > $9H > What of theses values can be set together (as AVAILABLE and STOPPED) = ?bI > What mask could be applied to the status value to get the status name =e in a > simple manner ?aE > (using separate QUEUE_IDLE, QUEUE_STOPPED, ... need to issue many =e getqui1 > and consolidate the TRUE/FALSE values obtained)  >o > Cordialement > Jean-Fran=E7ois Marchals > X9000 - LYON >  >m >s   ------------------------------  % Date: Thu, 20 Apr 2000 18:37:24 -0700 - From: David R Barnes <dbproductions@juno.com>E8 Subject: Re: Upgraded (hardware) 8400 System not booting: Message-ID: <20000420.184804.816.0.dbproductions@juno.com>  H Make sure your CPU firmware is at the latest, and more importantly, thatB both CPU's match in their firmware revs!  There is a 'hole' in theF 8400/GS140 firmware that will not give you a specific error indicating5 that is the problem.. I got bit by it with my 8400.      							David Barness    I On Thu, 20 Apr 2000 14:55:20 +0000 Steve.Spires@YELLOWPAGES.CO.UK writes:  > cc:r > bcc:? > Contact:   Tel: 3063  -  VSSG, 1st Floor, Bridge Street PlazaO > - > Upgraded (hardware) 8400 System not booting  >  > Dear All,- > G > We have just upgraded an 8400 replacing the existing single CPU with M	 > two newaC > ones, and adding another 2 GB of memory. When trying to boot the D
 > system, the@ > following occurs;  >  > jumping to bootstrap codee >  > CPU 0 halted >   halt code = 2  >   kernel stack not valid halta >   PC = fffffffffffffffci > F > I suspect this may be a memory related error, but haven't seen this  > particular > message before.  > E > Can anyone help to resolve this? I will supply more information if D	 > needed. A > System is running VMS 7.1-2 and has a total of sixGb of memory.E >  > What else? > A > AlphaServer 8400 4-5/625/4, Console V5.5-4 21-JUL-1999 16:27:00 @ > SROM V3.1, OpenVMS PALcode V1.21-1, Tru64 UNIX PALcode V1.23-18 > System Serial = qv, OS = VMS,  7:43:27  April 20, 2000 >  > P00>>>show memoryI@ > Set   Node   Size        Base Address         Intlv   Position@ > ---   ----   ----      -------- --------      -----   --------< >  A      5    2048 Mb   00000000 00000000      4-Way      0< >  A      6    2048 Mb   00000000 00000000      4-Way      1< >  B      7    2048 Mb   00000001 00000000      2-Way      0 >  >  > Steve Spires > VMS System Manager > BT/Yellow Pages  >  >  > [Information] -- PostMaster:F > This transmission is intended solely for the addressee(s) and may beF > confidential. If you are not the named addressee, or if the message 
 > has beenE > addressed to you in error, you must not read, disclose, reproduce, 0 > distribute ors > use this transmission. > > > Delivery of this message to any person other than the named  > addressee is notF > intended in any way to waive confidentiality.  If you have received  > this@ > transmission in error please contact the sender or delete the 
 > message. >  > Thank you.   David R Barnes dbproductions@juno.com3 =-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-m# Netbsd 1.3 / Linux / Unix, VMS guru-  @ ________________________________________________________________( YOU'RE PAYING TOO MUCH FOR THE INTERNET!% Juno now offers FREE Internet Access!e? Try it today - there's no risk!  For your FREE software, visit:e  http://dl.www.juno.com/get/tagj.   ------------------------------  # Date: Thu, 20 Apr 2000 23:24:58 GMTe. From: Michael Austin <maustin@nc.prestige.net>8 Subject: Re: Upgraded (hardware) 8400 System not booting/ Message-ID: <38FF9EFD.B1CC9CF3@nc.prestige.net>   , This is a multi-part message in MIME format.& --------------5D0B6439A4960BF4855D63E0* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit,  R I suppose that is why companies buy the installation and have the local FE installS it and fix any problems that may be associated.  Can you aafford the downtime while-, relying on a newsgroup of real-time support?   Michael Austin Ex-DEC Field Engineeri  % Steve.Spires@yellowpages.co.uk wrote:a   > cc:i > bcc:? > Contact:   Tel: 3063  -  VSSG, 1st Floor, Bridge Street Plaza  > - > Upgraded (hardware) 8400 System not bootingu >a > Dear All,j > N > We have just upgraded an 8400 replacing the existing single CPU with two newN > ones, and adding another 2 GB of memory. When trying to boot the system, the > following occurs;l >i > jumping to bootstrap codeS >e > CPU 0 halted >   halt code = 2  >   kernel stack not valid halt  >   PC = fffffffffffffffcA >DP > I suspect this may be a memory related error, but haven't seen this particular > message before.  > L > Can anyone help to resolve this? I will supply more information if needed.A > System is running VMS 7.1-2 and has a total of sixGb of memory.I >  > What else? > A > AlphaServer 8400 4-5/625/4, Console V5.5-4 21-JUL-1999 16:27:00 @ > SROM V3.1, OpenVMS PALcode V1.21-1, Tru64 UNIX PALcode V1.23-18 > System Serial = qv, OS = VMS,  7:43:27  April 20, 2000 >  > P00>>>show memory0@ > Set   Node   Size        Base Address         Intlv   Position@ > ---   ----   ----      -------- --------      -----   --------< >  A      5    2048 Mb   00000000 00000000      4-Way      0< >  A      6    2048 Mb   00000000 00000000      4-Way      1< >  B      7    2048 Mb   00000001 00000000      2-Way      0 >e > Steve Spires > VMS System Manager > BT/Yellow Pagesb >n > [Information] -- PostMaster:F > This transmission is intended solely for the addressee(s) and may beN > confidential. If you are not the named addressee, or if the message has beenR > addressed to you in error, you must not read, disclose, reproduce, distribute or > use this transmission. > N > Delivery of this message to any person other than the named addressee is notJ > intended in any way to waive confidentiality.  If you have received thisH > transmission in error please contact the sender or delete the message. >e > Thank you.  & --------------5D0B6439A4960BF4855D63E0- Content-Type: text/x-vcard; charset=us-ascii;   name="maustin.vcf"  Content-Transfer-Encoding: 7bitt, Content-Description: Card for Michael Austin  Content-Disposition: attachment;  filename="maustin.vcf"!   begin:vcard  n:Austin;Michael - tel;work:704-947-1089- x-mozilla-html:FALSE org:Michael Austin, Inca
 adr:;;;;;; version:2.1r+ email;internet:michaelaustininc@hotmail.com" title:PresidentS x-mozilla-cpt:;0 fn:Michael E. Austin	 end:vcardi  ( --------------5D0B6439A4960BF4855D63E0--   ------------------------------  % Date: Fri, 21 Apr 2000 12:34:10 +1200 * From: "Ryan C. Price" <pricerc@ihug.co.nz>3 Subject: Re: VAX CI storage vs served fibre channel * Message-ID: <8do7mf$sbh$1@news.ihug.co.nz>  . CI = 140Mb/s, shared, maximum distance 30m (?)5 FC = 1000Mb/s, switched, maximum distance 2km (IIRC).S  E The math is fairly simple. FC also gives you HS-to-HS shadowing (at ao
 price), soD you can have multi-site shadow sets without host-based shadowing (if that appeals to you).  E Your biggest problem is getting data from FC to your VAXen. I suspect- that leavingF your existing connections (CI and FDDI), would allow you to get access	 to the FCeC disks just about as fast as you can get to your existing disks (thej Alphas won't addD much latency). You'll still want some CI attached stuff for booting, but you indicate" that you'd be keeping that anyway.   /Ryanl  : "John Nixon" <jorlnixon@worldnet.att.net> wrote in messageA news:L%DL4.25856$WF.1002460@bgtnsc04-news.ops.worldnet.att.net... C > I know Fibre channel is faster than CI, but can someone give me aD
 subjective, > description of how much faster and better? >X' > My current configuration consists of:0? >  T hree CI based VAX 7860s (each with four CIXCDS),  FDDI and  10baseT 
 > EthernetE >  One Alhpa 8400 with four CIPCAs, FDDI, 10baseT and a bunch of SCSI0 adapters2 >  One Alpha 2100 5/300 with  dual Faster Ethernet > > > The three VAXes and the Alpha 8400 are attached to four Star couplers with a 2 > farm full of HSJ52s and (ugh) MTI Stingray IIIs. >0B > I am trying to decide if I should replace the MTI Stingrays with	 HSJ82s so0, > that my VAXes can have direct connections,C > or should I get Fibre Channel adapters for the Alphas and connectt into our# > developing Compaq SAN that we aremF > installing for our Unix systems (Tru64 and Solaris).  (Actually, the dataF > would be moved from the HSJ52s to the new storage, and the data from theF* > StingRays would be moved to the HSJ52s). >-F > Our eventual goal is to get rid of the VAXes and move all processing to theE > Alphas, which would mean Fibre Channel would be the only way to go,a but IeB > will have to make sure I don't hinder the VAX performance for at least an
 > year or so.u >  >  >c   ------------------------------  # Date: Fri, 21 Apr 2000 01:17:39 GMT / From: "John Nixon" <jorlnixon@worldnet.att.net>p3 Subject: Re: VAX CI storage vs served fibre channelpG Message-ID: <T_NL4.25554$fV.1427898@bgtnsc05-news.ops.worldnet.att.net>/   Ryan,b  G How do I get multi-site shadow sets without host based shadowing.  Thati3 would appeal to me, but I have never heard of that.e  K As for the math,  that part is simple, if you believe that in real life theeJ math numbers hold up.  You only get the theoretical thruput numbers if youH eliminate every bottleneck.  With CI I have come close and  can actuallyI sustain 7MB/second, but I am not sure what kind of config I would need ton maintain FC thruput maximums.A    5 "Ryan C. Price" <pricerc@ihug.co.nz> wrote in message $ news:8do7mf$sbh$1@news.ihug.co.nz...0 > CI = 140Mb/s, shared, maximum distance 30m (?)7 > FC = 1000Mb/s, switched, maximum distance 2km (IIRC).- > G > The math is fairly simple. FC also gives you HS-to-HS shadowing (at a  > price), soF > you can have multi-site shadow sets without host-based shadowing (if > that appeals
 > to you). > G > Your biggest problem is getting data from FC to your VAXen. I suspectd > that leavingH > your existing connections (CI and FDDI), would allow you to get access > to the FC"E > disks just about as fast as you can get to your existing disks (then > Alphas won't addF > much latency). You'll still want some CI attached stuff for booting, > but you indicate$ > that you'd be keeping that anyway. >t > /Ryant > < > "John Nixon" <jorlnixon@worldnet.att.net> wrote in messageC > news:L%DL4.25856$WF.1002460@bgtnsc04-news.ops.worldnet.att.net...-E > > I know Fibre channel is faster than CI, but can someone give me a  > subjective. > > description of how much faster and better? > > ) > > My current configuration consists of:.A > >  T hree CI based VAX 7860s (each with four CIXCDS),  FDDI and 	 > 10baseTb > > EthernetG > >  One Alhpa 8400 with four CIPCAs, FDDI, 10baseT and a bunch of SCSIe
 > adapters4 > >  One Alpha 2100 5/300 with  dual Faster Ethernet > >o@ > > The three VAXes and the Alpha 8400 are attached to four Star > couplers with ae4 > > farm full of HSJ52s and (ugh) MTI Stingray IIIs. > >LD > > I am trying to decide if I should replace the MTI Stingrays with > HSJ82s soA. > > that my VAXes can have direct connections,E > > or should I get Fibre Channel adapters for the Alphas and connect 
 > into our% > > developing Compaq SAN that we areIH > > installing for our Unix systems (Tru64 and Solaris).  (Actually, the > dataH > > would be moved from the HSJ52s to the new storage, and the data from > theT, > > StingRays would be moved to the HSJ52s). > > H > > Our eventual goal is to get rid of the VAXes and move all processing > to theG > > Alphas, which would mean Fibre Channel would be the only way to go,T > but IGD > > will have to make sure I don't hinder the VAX performance for at	 > least aV > > year or so.E > >  > >t > >  >l >e   ------------------------------  + Date: Thu, 20 Apr 2000 14:10:55 -0500 (CDT)u# From: Ron <RROCKWELL@CCTR.UMKC.EDU>E Subject: Re: Verify of Backups. Message-ID: <01JOGLE19BRAQVNO4E@CCTR.UMKC.EDU>  F   Just curious, anyone, but taking this point seriously and attemptingJ to apply it to a system with multiple users receiving mail at all hours...E How can data be compared when the snapshot is obsolete by the time its has finally finished?   G   We have a TZ89 tape backup job that is giving me other problems rightiD now, but as soon as those other problems are solved my next questionF will be how to verify that the backup was successful. The job attemptsA to backup 16 separate savesets of 16 different disks with severaltC hundred user accounts on each disk onto one tape. All of the users iI receive e-mail all of the time. Users are online 24 by 7. Mail comes in, iI and mail is also deleted, not to mention other files as well, but in the  G past few attempts at using verify, mail.mai and external .mai files aretA usually missing or changed and show errors. So, again, how can itaE be possible to automate such a job and have it check its own success?e  M   Just in case anyone has any ideas about the "other problems" i mentioned...IG It seems that no matter how we try (under different usernames, or under N the system username/quotas, varying the ordering of which disks are backed up,F the job consistently runs fine up to a cpu-time of ~23 minutes, i/o ofE about 1.6 million and then lags into a very lazy state. I/O continues.G very infrequently, no OPCOM messages or tape problems are apparent, the M job just slows to a near dead crawl on the fourth disk every time, regardlessnF of which disk we setup to be the fourth in the ordering. We also triedE turning off/on compaction to see if it would lag on a different disk.u= All sixteen disks contain around 35 Gig of data. Our current ,E hypotheses is that the process is bumping a quota somehow. Any ideas?e  J OpenVMS V7.2  on node NODE03  19-APR-2000 19:58:36.74  Uptime  50 12:47:34M   Pid    Process Name    State  Pri      I/O       CPU       Page flts  PagesiO 5E202257 STDNT7 (4/16)   LEF      5  1623428   0 00:23:10.50      2693    220 Bq      H =>From:	IN%"pricerc@ihug.co.nz"  "Ryan C. Price" 20-APR-2000 03:49:25.11  =>To:	IN%"Info-VAX@Mvb.Saic.Com" =>CC:	 =>Subj:	RE: Verify of Backups   B =>My point was merely that it's pointless to make a backup without =>checking it. =>       Ron Rockwell   Academic Computing Services-'   University of Missouri at Kansas CityA   ------------------------------    Date: 20 Apr 2000 15:21:47 -04004 From: "Robert Deininger" <rdeininger@mindspring.com> Subject: RE: Verify of Backups+ Message-ID: <B524D10D-2398D@165.247.26.226>S  I On Wed, Apr 19, 2000 6:22 PM, Malcolm Dunnett <dunnett@mala.bc.ca> wrote:S0 >In article <8dl6ku$8lk$1@info.service.rug.nl>, 1 >    helbig@astro.rug.nl (Phillip Helbig) writes:   H >> Why/how?  Just curious---it still has to read them again to compare; 2 >> isn't /VERIFY doing essentially the same thing? > @ >   It depends on the tape drive and how BACKUP gets back to theH >start of the saveset. I used to use /VERIFY and I found that BACKUP wasJ >rewinding the tape and then skipping forward to the start of the saveset,E >so every save set in front of the one you wanted to verify had to be-J >passed over along the way - slowing things down and creating extra passesD >over the tape. On some drives ( such as the old Exabyte 8200s ) theG >"file skip" function got executed at read speed, so it could literallyb# >take hours to start a verify pass.a  C I do my backups with /verify, with multiple save sets per DLT tape.aD Backup does NOT go back to the beginning of the tape for each verifyF on our systems. It just backs up to the beginning of the last save set and compares it to disk.  C I've heard other complaints of the rewind to beginning problem, but1 I don't know what triggers it.  E Exabyte 8200!  Ack!  At my last site, we had an even dozen of various D exabytes, which were moderately busy.  Over several years, there wasB NEVER a time when there wasn't at least one in for repairs.  NooneE expected to be able to read a tape except on the drive that wrote it.TB If you could, it was a bonus.  Probably a third of the sick drivesF had important tapes stuck in them when we sent them in.  We learned toA only use the luxury-priced Exabyte brand tapes, which were barely0E adequate.  All other brands we tried were equivalent to data suicide. F Helical scan is for the birds. I'll stick to DLTs from now on, thanks.     ---------------------------t Robert Deininger rdeininger@mindspring.com    ------------------------------  % Date: Thu, 20 Apr 2000 15:23:55 -0400n" From: Dan Sugalski <dan@sidhe.org> Subject: Re: Verify of Backups8 Message-ID: <4.3.1.0.20000420152014.01d34340@24.8.96.48>  % At 02:10 PM 4/20/00 -0500, Ron wrote:aO >   Just in case anyone has any ideas about the "other problems" i mentioned...-H >It seems that no matter how we try (under different usernames, or underO >the system username/quotas, varying the ordering of which disks are backed up,bG >the job consistently runs fine up to a cpu-time of ~23 minutes, i/o ofeF >about 1.6 million and then lags into a very lazy state. I/O continuesH >very infrequently, no OPCOM messages or tape problems are apparent, theN >job just slows to a near dead crawl on the fourth disk every time, regardlessG >of which disk we setup to be the fourth in the ordering. We also trieduF >turning off/on compaction to see if it would lag on a different disk.= >All sixteen disks contain around 35 Gig of data. Our currenteF >hypotheses is that the process is bumping a quota somehow. Any ideas?  J Are you running VMS 7.2 and connecting to storage and tape drives hanging I off the same HSJ40 by any chance? There's some wackiness with 7.2's port rK driver or something that will basically lock up the controller for a while oJ until VMS decides that IO's taken too long, at which point it sends out a I hard reset packet and starts back up. (It leaks credits or something, it yK seems) I ran into this one--the workaround was to hang the tape drives off   their own HSJ40.   					Dan  I --------------------------------------"it's like this"------------------- 2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and even-;                                       teddy bears get drunkN   ------------------------------    Date: 20 Apr 2000 15:34:05 -04004 From: "Robert Deininger" <rdeininger@mindspring.com> Subject: Re: Verify of Backups+ Message-ID: <B524D3EF-2E6F0@165.247.26.226>-  I On Thu, Apr 20, 2000 6:10 AM, Phillip Helbig <helbig@astro.rug.nl> wrote: 1 >In article <8dmcr0$kc0$1@n.ruf.uni-freiburg.de>, 7 >gartmann@immunbio.mpg.de (Christoph Gartmann) writes: 0 > 6 >> I thought I explained this already some weeks ago.  >y; >Right, but it was easier to just repost the whole thing...P  G If you're going to repost the questions periodically, at least include  D the previous answers!  We'll never make progress if we have to startF over every time.  Why don't you turn it into a little FAQ-in-progress?    " >> >   o  /MEDIA_FORMAT=COMPACTION  D Doesn't work on our Quantum-brand DLTs.  I think it does work on theG Digital-firmware variety.  We set the compression using the front panelc# push-button when initing each tape.     5 >Should one always use it?  Does it slow things down?   D I would say use it, unless you need to read the tape on a drive thatC doesn't support it.  Depending on the drive and the data, it may go : faster, since compressed data requires less tape movement.    < >Why is the table above non-linear even without compression?  C They must be fitting more bits/square inch on the tape in the newerdF formats.  DLTs are the same way -- more tracks, more bits/inch on each& track.  And occasionally longer tapes.    E >> Makes no sense with DATs. It is used on old 9-track tapes as thesea supportp/ >> various densities (e.g. 1600bpi or 6250bpi).0 >y >What confuses me is >- >$ help init/densC >n0 >     DDS1           Digital Data Storage 1 - 2G0 >     DDS2           Digital Data Storage 2 - 4G3 >     DDS3           Digital Data Storage 3 - 8-10Gt+ >     DDS4           Digital Data Storage 4t    D I think newer versions of VMS can control the density on DLT and DDSG drives - when they understand the drives.  This is another feature that E doesn't work on our quantum-brand DLT drives.  (Actually, the VMS 7.1 E help doesn't list any DLT or DDS densities, so I don't think I shouldc, complain about this until we have upgraded.)       ---------------------------  Robert Deininger rdeininger@mindspring.comu   ------------------------------  + Date: Thu, 20 Apr 2000 15:25:15 -0500 (CDT)n# From: Ron <RROCKWELL@CCTR.UMKC.EDU>i Subject: Re: Verify of Backups. Message-ID: <01JOGOVKICJOQVNO4E@CCTR.UMKC.EDU>  B =>From:	IN%"dan@sidhe.org"  "Dan Sugalski" 20-APR-2000 14:23:34.08E =>To:	IN%"RROCKWELL@CCTR.UMKC.EDU"  "Ron", IN%"Info-VAX@Mvb.Saic.Com"  =>CC:	 =>Subj:	RE: Verify of Backupsl =>' =>At 02:10 PM 4/20/00 -0500, Ron wrote: Q =>>   Just in case anyone has any ideas about the "other problems" i mentioned...mJ =>>It seems that no matter how we try (under different usernames, or underQ =>>the system username/quotas, varying the ordering of which disks are backed up,-I =>>the job consistently runs fine up to a cpu-time of ~23 minutes, i/o of.H =>>about 1.6 million and then lags into a very lazy state. I/O continuesJ =>>very infrequently, no OPCOM messages or tape problems are apparent, theP =>>job just slows to a near dead crawl on the fourth disk every time, regardlessI =>>of which disk we setup to be the fourth in the ordering. We also tried H =>>turning off/on compaction to see if it would lag on a different disk.? =>>All sixteen disks contain around 35 Gig of data. Our currentrH =>>hypotheses is that the process is bumping a quota somehow. Any ideas? =>L =>Are you running VMS 7.2 and connecting to storage and tape drives hanging K =>off the same HSJ40 by any chance? There's some wackiness with 7.2's port rM =>driver or something that will basically lock up the controller for a while sL =>until VMS decides that IO's taken too long, at which point it sends out a K =>hard reset packet and starts back up. (It leaks credits or something, it iM =>seems) I ran into this one--the workaround was to hang the tape drives off g =>their own HSJ40.  J Kind of, sort of... The machine in question serves an HSZ-50 for the disksH in question, the tape machine is hooked up to its MKA on-board SCSI portG though. Its a 2100 so if i am not mistaken, its the same SCSI port that;G handles the internal floppy and CD. The drives are DKBxxx. The page and7E swap are on their own private DRA... raid controller for the internal;F shelf. The HSZ50's disks are served by this machine and two others allG sharing the same FWD SCSI bus. The tape machine is hanging off the back-4 on the 60-pin SCSI port that comes with the machine.  -C  So, it could be the same problem with a different set of hardware, D only for completely reverse reasons.... if that makes any sense. 8^)B The highlights of your message though are "7.2", "storage","tape",C "HSxxxx","lock up" and a couple other keywords which all ring thoseaC ever familiar alarm bells "do you mean i have to call support?" in o my mind again. 8^(       =>
 =>					Dan =>K =>--------------------------------------"it's like this"-------------------u4 =>Dan Sugalski                          even samuraiA =>dan@sidhe.org                         have teddy bears and even = =>                                      teddy bears get drunkt =>     Ron Rockwell   Academic Computing Services('   University of Missouri at Kansas Cityu   ------------------------------  % Date: Thu, 20 Apr 2000 16:36:31 -0400y" From: Dan Sugalski <dan@sidhe.org> Subject: Re: Verify of Backups8 Message-ID: <4.3.1.0.20000420163527.01d748e0@24.8.96.48>  % At 03:25 PM 4/20/00 -0500, Ron wrote: E >  So, it could be the same problem with a different set of hardware, E >only for completely reverse reasons.... if that makes any sense. 8^)eC >The highlights of your message though are "7.2", "storage","tape",tD >"HSxxxx","lock up" and a couple other keywords which all ring thoseC >ever familiar alarm bells "do you mean i have to call support?" inA >my mind again. 8^(   J Well, a call to support's a good thing in this case, but I'd be surprised K if it's the same problem as I ran across. That was specifically because of IJ some bizarre happenstance in a newly rewritten CI port driver, but you're  not using one of those.e   					Dan  I --------------------------------------"it's like this"-------------------n2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and evenA;                                       teddy bears get drunke   ------------------------------    Date: 20 Apr 2000 14:40:40 -0700* From: dunnett@mala.bc.ca (Malcolm Dunnett) Subject: RE: Verify of Backups, Message-ID: <GhfJAgVAdCXV@malvm2.mala.bc.ca>  , In article <B524D10D-2398D@165.247.26.226>, 9    "Robert Deininger" <rdeininger@mindspring.com> writes:a > E > I do my backups with /verify, with multiple save sets per DLT tape.-F > Backup does NOT go back to the beginning of the tape for each verifyH > on our systems. It just backs up to the beginning of the last save set > and compares it to disk. > D     Can it quickly back up over the saveset? Some drives can't ( andI on some drives that can it appears that BACKUP and/or VMS doesn't realizel& this and still does it the slow way ).  C     I originally started using /COMPARE when we had Exabytes ( as IlG mentioned earlier - it could take hours to start a /VERIFY on a savesetbD on those it it wasn't at the start of the tape - and you sure didn't? want to place any trust in a backup on an Exabyte if you didn'tf? do a /VERIFY). When we switched to DAT there didn't seem to be aH much difference between /COMPARE and /VERIFY, in fact I switched back toF doing /VERIFY. Then I replaced the DATs with some DLT2000XTs and foundF that once again /VERIFY was much slower than /COMPARE. It appears thatG even though the DLT2000s support fast file skips BACKUP doesn't use it.    ------------------------------  % Date: Thu, 20 Apr 2000 13:05:20 -1000t- From: Maren Purves <m.purves@jach.hawaii.edu>m Subject: Re: Verify of Backups+ Message-ID: <38FF8D30.14CB@jach.hawaii.edu>i   Ewan Parker wrote: > ? > On 18-Apr-00 22:24:22, L McCann said about Verify of Backups:aJ > > We are running VMS Backup on Alpha servers and tz88 tape drives (hsz50K > > controller) and we do not use the /verify option on the backup command.u > M > > I'd like to get a feel for what other VMS shops do for their backups.  DooM > > you verify them or not?  I have found them to be reliable whenever I do a@C > > restore but I know that it only takes 1 time to make you sorry.a > Q > If you are under time constraints, you could mount the tape after the backup istQ > complete, and just restore the last file on that tape to a spare disk.  I woulds > recommend /verify though.C  E I wouldn't buy the time constraints as an excuse. Backups can be doneuC at any time, and the only "constraint" I could possibly see is thatpD the tape unit is needed for something else, but in that case I'd buyE another one. We do our backups in batch jobs, with email notificationmE of errors, and verified, most of them after one part of the people godB home and the other part of the people start (that's developers and= such during the day and people using the telescope at night).    Maren Purves   ------------------------------  % Date: Fri, 21 Apr 2000 10:42:34 +0930h/ From: Mark Daniel <mark.daniel@wasd.vsm.com.au>T# Subject: Re: Web server for VAX/VMS / Message-ID: <38FFAB02.E6596738@wasd.vsm.com.au>e  A Yes Arne, as far as I know yahMAIL is operating on many OSU sites ? (from emails I've received).  There doesn't seem to be anythingX; particularly tricky in setting it up apart from getting the B authorization right - in common with any application that requiresD such an environment.  I managed it for my OSU development/testing of yahMAIL.  @ Mike contacted me a few days ago with a problem.  Looking at theF information provided it was probably an authorization issue, suggestedF he re-read the yahMAIL OSU readme (which is not a paradigm of OSU infoD - but then I'm not an OSU admin) and perhaps more importantly, if he? wasn't full-bottle on OSU drop a query to the OSU mailing list.l  @ Pure personal opinion ... I think he'd be happier with yahMAIL's
 capabilities.i  E Postscript: yahMAIL also now works with VMS Apache (1.3.9 BETA at anyp< rate).  No changes to yahMAIL code required, just some minorC accomodations for Apache in the CGILIB code.  Took about two hours. D There's currently a yahMAIL 1.3 BETA out there (no problems reportedF so far), when the final 1.3 is released at the end of this month it'll2 be able to work under WASD, OSU, Apache, Purveyor.   Regards to all, Mark Daniel.   Arne Vajhj wrote:   > Mike Rechtman wrote:> > > On a VAXstation 4000-60, OpenVMS 6.2, OSU V3.6b I've triedB > > installing yahMAIL and could not get it running. Got a messageB > > (from memory, now) something like "ERROR access not permitted"B > > The message came on a HTML formatted page, and I could find no > > reason for it.B > > In the end I gave up, took the source for HYPERMAIL and 'bent'B > > it so it allows me to read VMSmail from a browser. Fairly Q&D,  > > but enough for whats needed. >uD > It is my impression that many has gotten YAHMAIL to work with OSU. >o > Arne  E +-------------------------------------------------------------------+oD  Mark Daniel            Opinions my own ... and on loan from others.E  mailto:Mark.Daniel@wasd.vsm.com.au (Mark.Daniel@dsto.defence.gov.au)oE +-------------------------------------------------------------------+w   ------------------------------   End of INFO-VAX 2000.222 ************************ion to see if it would lag on a different disk.u= All sixteen disks contain around 35 Gig of data. Our current ,E hypotheses is that the process is bumping a quota somehow. Any ideas?e  J OpenVMS V7.2  on node NODE03  19-APR-2000 19:58:36.74  Uptime  50 12:47:34M   Pid    Process
yDWU?jpOTeLCsZjf2}0Y!Au6bQMe4`[C0}Ϗ000.zi6AOlF
C܎gc5[	)AW(U~7?Z0xxwR2WLS3 @:l?rMV:t
v8Q`-V6J`9MZRcKT,;b>p۪#=Tpcl&QY9ͯe^8*bKo:%7
K.Ap=XPml6,+D˄3UdP<mW}PDSvGqmj#̪F>@Iɵ*aƗ~*Jjdyss1|LOwN%ߌ2uƠS.`:\rx-)J E.*眜Ƹ4nGo5}$"wb/[W3F	\̷jvhn!"$wǱysߔnS`bXkjoЛgwlvMmKO~+jqsheѤ"(2b-gO>B][ylQ6nj9eEPK    5|(f1s      lame-3_70/l3side.hIM VMSV       V6.2 IM$ VFAB P     B 	 @    m!Ц IM VFHC ,     ; 
 	p{    IM% VDAT <     % 0ގr0ގr    IM VPRO X     'b1@@  0       IM VRDT      =R 0ގrUmkAbHQR+-$PЂiiYֻ\+kz~B3c&
 H&<˄uP9AX*9}Um<7壄^)Z<ݨuxsi
V<VӺ"v^ܛDd̽4zx4S9ӉFW~_eDX-'B&6XA`ـ!_c68b1YHAdѯ,
<|/7m%R<hu۬3q"Ys.c뵂FG?FB
<P4l9wR?ze2KI9V[j'S}c@L	y$ʔ=qf}-T)W,v1WbcѬN̊TڔO+?2rfY$,U_x>U9DCR\lMpi
J;)v9
Z6p9L\I,LUp)SbS)8#<I-1T@ɞa=zb|2hQz3L#Å{o}ٷ	+%9S41Q\ǘsSjD~VF
MfƺJPpnaF57Sgu