1 INFO-VAX	Thu, 20 Apr 2000	Volume 2000 : Issue 220       Contents:$ RE: Backup Strategies for Enterprise$ Re: Backup Strategies for Enterprise$ Re: Backup Strategies for Enterprise$ RE: Backup Strategies for Enterprise$ Re: Backup Strategies for Enterprise BACKUP, file names, file IDs  Re: BACKUP, file names, file IDs Buy a Refurbished AS1000a ! % Calculating Discordian dates from DCL * Command Line History and C++ input streams. Re: Command Line History and C++ input streams DEC (ES40) MS610-EA $9999  EMC Disk Storage RE: EMC Disk Storage Re: EMC Disk Storage Re: EMC Disk Storage Re: EMC Disk Storage Re: EMC Disk Storage Re: EMC Disk Storage Re: EMC Disk Storage Re: EMC Disk Storage Re: EMC Disk StorageK Re: How can you find the exact size of a file in bytes using a VMS command?  Re: HSJ80 experience?  Re: Mozilla M15's out... Re: Mozilla M15's out...* Re: OpenVMS mail question: Attached files?! Re: Scheduler & PSDC alternatives > Re: Sending MIME attachments in OpenVMS 7.1-H1 and TCPware 5.3 Re: Using DiskMizer with Oracle  Re: Using DiskMizer with Oracle  Re: Using DiskMizer with Oracle  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: 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 Re: Web server for VAX/VMS What's this option?  RE: What's this option?  RE: What's this option?  RE: What's this option?  RE: What's this option?  RE: What's this option?  RE: What's this option?  RE: What's this option?  RE: What's this option?  RE: What's this option? 5 Why does Pathworks(Mac) change mod date of .dir file?   F ----------------------------------------------------------------------  % Date: Wed, 19 Apr 2000 15:01:39 -0300 1 From: "Boyle, Darren" <boyledj@bankofbermuda.com> - Subject: RE: Backup Strategies for Enterprise K Message-ID: <F150836441C5D311A11700508B6FF01A6D57DF@bdant024.bda.bobda.com>   F Just one thought, stay away from ABS, we attempted to put it in but itI caused so many problems we scrapped it before it really went live, we are  now following the Legato route.  - Darren   > ----------2 > From: 	Don Rogstad[SMTP:Don.Rogstad@dalsemi.com]* > Sent: 	Wednesday, April 19, 2000 2:35 PM > To: 	Info-VAX@Mvb.Saic.Com, > Subject: 	Backup Strategies for Enterprise > L > Our current environment has lots of Sun boxes, Digital Unix boxes, WindowsE > NT boxes and of course, OpenVMS VAX and Alpha machines.  We are now K > looking for ways to consolidate the backup process from multiple machines  > / platforms. > L > Does anyone consolidate backups from various platforms?   What products doI > you use?   Given the reliability of OpenVMS backup, I would like to use J > OpenVMS as the backup server, but this is not necessarily a requirement. > F >     It looks like Compaq has two solutions.   One, ABS and the otherD > Legato.  Has anyone used Legato Networker?  Wasn't ABS sold to CA?7 > Anyone used either one for multiple platform backups?  > 8 >    Your thoughts and ideas on backups are appreciated. > 	 > Thanks, 
 > Don Rogstad  > Dallas Semiconductor > 972.371.4570 > Don.Rogstad@dalsemi.com  >     F **********************************************************************C This message and any files transmitted with it are confidential and J may be privileged and/or subject to the provisions of privacy legislation.M They are intended solely for the use of the individual or entity to whom they L are addressed. If the reader of this message is not the intended recipient, B please notify the sender immediately and then delete this message.I You are notified that reliance on, disclosure of, distribution or copying  of this message is prohibited.   Bank of Bermuda F **********************************************************************   ------------------------------  % Date: Wed, 19 Apr 2000 15:34:25 -0400 0 From: JF Mezei <jfmezei.spamnot@vl.videotron.ca>- Subject: Re: Backup Strategies for Enterprise / Message-ID: <38FE0A3D.7AB9F046@vl.videotron.ca>    Don Rogstad wrote: > ] > Our current environment has lots of Sun boxes, Digital Unix boxes, Windows NT boxes and of  H >course, OpenVMS VAX and Alpha machines.  We are now looking for ways to consolidate the backup  , >process from multiple machines / platforms.  K Think of this: when your VAX disk busts and needs to be replaced with a new K virgin one. Will you be able to restore your backup that was done with some 2 sort of network software through another machine ?   ------------------------------  + Date: Wed, 19 Apr 2000 19:43:47 +0000 (   ) 3 From: Christopher Smith <chriss@Mufasa.pubserv.com> - Subject: Re: Backup Strategies for Enterprise I Message-ID: <Pine.LNX.4.05.10004191943040.4848-100000@Mufasa.pubserv.com>   $ On Wed, 19 Apr 2000, JF Mezei wrote:   > Don Rogstad wrote:  _ > > Our current environment has lots of Sun boxes, Digital Unix boxes, Windows NT boxes and of  J > >course, OpenVMS VAX and Alpha machines.  We are now looking for ways to > consolidate the backup  . > >process from multiple machines / platforms.  M > Think of this: when your VAX disk busts and needs to be replaced with a new M > virgin one. Will you be able to restore your backup that was done with some 4 > sort of network software through another machine ?  8 Yes, if that other machine is another VAX (or alpha?) ;)   Regards,   Chris   O =============================================================================== @ "My two cents"			(http://rootworks.com/twocentsworth.cgi?128562)= Christopher Smith(chriss@pubserv.com)			Prgramer^W Programmer  Prime Synergy of Champaign, IL. % ------------------------------------- I "Where a calculator on the ENIAC is equipped with 18,000 vacuum tubes and H weighs 30 tons, computers in the future may have only 1,000 vacuum tubes; and weigh only 1.5 tons." -- Popular Mechanics, March 1949  O -------------------------------------------------------------------------------    ------------------------------  % Date: Wed, 19 Apr 2000 15:22:35 -0500 + From: Don Rogstad <Don.Rogstad@dalsemi.com> - Subject: RE: Backup Strategies for Enterprise H Message-ID: <EA5FE99F5DEAD311A6CB00805F199992BCA807@misnts1.dalsemi.com>   JF Mezei wrote:    >Don Rogstad wrote:  >>  ^ >> Our current environment has lots of Sun boxes, Digital Unix boxes, Windows NT boxes and of I >>course, OpenVMS VAX and Alpha machines.  We are now looking for ways to  consolidate the backup  - >>process from multiple machines / platforms.  > L >Think of this: when your VAX disk busts and needs to be replaced with a newL >virgin one. Will you be able to restore your backup that was done with some3 >sort of network software through another machine ?    Which was one of reasons to find out what others are doing.  I know the reliability of OpenVMS Backup.  I have heard of the un-reliability of both Un*xes Dump and NT's attempts at backup.  Actually Seagate Backup Exec on my Windows NT workstation seems to backup pretty reliable, but I have never needed to do a restore yet.  (At least it runs daily and gives a log of what it could and could not back up).  But we are looking for an "enterprise" solution to help manage all the tapes, tape drives, man power cur? rently needed.   But we want a solution that is STILL RELIABLE.    Don Rogstad  Dallas Semiconductor   ------------------------------  # Date: Wed, 19 Apr 2000 23:49:42 GMT . From: "Curtis Rempel" <curtis.rempel@home.com>- Subject: Re: Backup Strategies for Enterprise : Message-ID: <qCrL4.69712$W5.632890@news1.rdc1.ab.home.com>  K We run ABS (still a Compaq product so far) for VMS and UNIX backups - works L great (after we sent the incompetent Digital consultant out the door and didG it ourselves).  ABS is touted as an enterprise backup solution for VMS, E UNIX, and NT.  If your enterprise contains only VMS and UNIX, it will H suffice.  ABS is NOT a good solution for NT - it knows nothing about theL registry or open files etc.  The non-VMS agent is just GNU tar which runs onH UNIX and NT.  We went with Backup Exec for NT and let ABS do the VMS and UNIX boxes.   J ABS 2.x was a dog to install and configure but 3.x is much better and doesJ not rely on the DECscheduler component in the 2.x series.  You can plug inL your own scheduler in 3.x and free yourself from the claws of the Evil SatanD of software vendors that bought it from Digital (no, not MS, worse).  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 looking C for ways to consolidate the backup process from multiple machines / 
 platforms. > L > Does anyone consolidate backups from various platforms?   What products doG you use?   Given the reliability of OpenVMS backup, I would like to use H OpenVMS as the backup server, but this is not necessarily a requirement. > F >     It looks like Compaq has two solutions.   One, ABS and the otherK Legato.  Has anyone used Legato Networker?  Wasn't ABS sold to CA?   Anyone . used either one for multiple platform backups? > 8 >    Your thoughts and ideas on backups are appreciated. > 	 > Thanks, 
 > Don Rogstad  > Dallas Semiconductor > 972.371.4570 > Don.Rogstad@dalsemi.com    ------------------------------   Date: 19 Apr 2000 17:46:46 GMT* From: helbig@astro.rug.nl (Phillip Helbig)% Subject: BACKUP, file names, file IDs . Message-ID: <8dkre6$53i$1@info.service.rug.nl>  ? Does BACKUP at any stage use file IDs as opposed to file names?   G I'm still looking for a definitive answer---and obviously one can't do  H very much experimenting here---as to whether BACKUP/SINCE=BACKUP can be H used on one disk and one machine and then the files restored to another F disk on another machine which was originally a copy of the other disk.  B In other words, I work at site A and site B but (at least for the E directories in question), files are only created on one machine at a  I time.  I'm looking for a convenient way to keep the two consistent.  The  F plan is to start off with the disks being copies of each other.  When F moving from A to B, I take the latest BACKUP/SINCE=BACKUP (which I do F nightly anyway) and restore it incrementally, i.e. only the new files 1 get copied to disk AND DELETED FILES GET DELETED.   I This MIGHT be a problem if BACKUP uses file IDs instead of file names at   some point.   I Of course, doing an /IMAGE BACKUP of the disk and transporting this back  G and forth would be a possibility, but a) it seems like overkill and b)  B means I have more disks (I would rather use redundancy for volume  shadowing---I'm not THAT rich).   F If files are automatically deleted, of course, then this is a) what I G want but b) I have to make sure that, for the directories in question,  3 files are not created simultaneously at both sites.    ------------------------------   Date: 20 Apr 2000 01:55:39 GMT* From: bleau@umdsp.umd.edu (Lawrence Bleau)) Subject: Re: BACKUP, file names, file IDs ) Message-ID: <8dlo2r$el0$1@hecate.umd.edu>   [ In article <8dkre6$53i$1@info.service.rug.nl>, helbig@astro.rug.nl (Phillip Helbig) writes:  [snip]C >In other words, I work at site A and site B but (at least for the  F >directories in question), files are only created on one machine at a J >time.  I'm looking for a convenient way to keep the two consistent.  The G >plan is to start off with the disks being copies of each other.  When  G >moving from A to B, I take the latest BACKUP/SINCE=BACKUP (which I do  G >nightly anyway) and restore it incrementally, i.e. only the new files  2 >get copied to disk AND DELETED FILES GET DELETED.  L I managed to do this at my site to keep two data sets in sync.  I didn't useN BACKUP, either, but "rolled my own" utility.  Basicly, I have a .COM file thatO sets up two f$search streams, A dna B, with identical patterns, but one of them M is on another node.  The .com file uses the local node as the master copy and " updates the remote node as needed.  B The .com file reads the two in tandem and compares them.  The onlyL possibilities are that they are equal, or that one of them lexicographically precedes the other.   M If the filenames (plus type, version, directory) from A and B match, the file O exists in both locations, so leave it alone and advance to the next filename in  both search streams.  O If a file from A precedes the file from B, that means it exists on A but not on N B.  Copy the file to the remote node, then fetch the next filename from stream
 A, but not B.   N If a file from A follows the file from B, that means it exists on B but not onN A.  In other words, the file was deleted from the local system.  In this case,N delete the file from the remote node, then fetch the next filename from stream
 B, but not A.   O You'll need some special case code to handle when one search stream ends before 
 the other.  O This assumes that directories won't be created or deleted, btw.  If you want to K handle that case, it could be done by the .com file firing up a task on the M remote end to perform directory create or delete operations and passing to it O the name of the directory and any other options.  It's much simpler in a stable  directory tree, though.   H For your application, you'd run this after doing your file create/deleteL operations on a node, but before you do any file create/delete operations onL the other node.  If you forget, it'll delete newly created work on the other node!   I For me, though, it's a real time saver.  I maintain two copies of data on N separate standalone systems, of about six different types of data (all locatedI in different directories).  I just activate the procedure and pass it the I search pattern unique to that data set, and the remote node gets updated.   = If anyone is really interested I can post a copy of the code.    Lawrence Bleau University of Maryland" Physics Dept., Space Physics Group 301-405-6223 bleau@umdsp.umd.edu    ------------------------------  % Date: Wed, 19 Apr 2000 20:28:41 -0400 + From: "David Turner" <d_b_turner@yahoo.com> $ Subject: Buy a Refurbished AS1000a !- Message-ID: <sfsjjsiin215@corp.supernews.com>   ( We have a completely refurbished (by us)  
 AS1000a 5/400 " KZPAC-CA 3 Channel Ultra SCSI RAID IC-KZPAA-AA Narrow SCSI " S3Trio64 2mb video or Elsa Synergy Q-Logic QLA1040 (OEM KZPBA)  RRD46-AA 12x SCSI CD-ROM DE500-AA or BA (Depends on OS) Internal 4gb Wide disk   $3300   # System comes with 3 months warranty % (long enough to be put on a contract) ) Can be custom built per your requirements     ( VMS or UNIX license (NEW) additonal COST   Call or email for more info    -- David Turner Island Computers US Corporation  2700 Gregory Street  Savannah GA 31404  Tel: 912 447 6622  Fax:912 201 0096 sales@islandco.com   ------------------------------  % Date: Wed, 19 Apr 2000 21:22:46 +0100 . From: mpatt644 <mpatt644@netscapeonline.co.uk>. Subject: Calculating Discordian dates from DCL4 Message-ID: <38FE1596.FB176413@netscapeonline.co.uk>   Hi, H 	Working in the finance sector I keep getting asked for the current dateG in Discordian format, I couldn't find any routines or lexicals to do it C for me so I hacked together the attached bit of DCL. I though't I'd G share it here in case anyone else find's it usefull. Obviously the moon $ phases part is just a bit of fun ;-)   Martyn.    --- Cut here --- $save_verify = f$verify(0)G $! Procedure to calculate the Discordian date from the Gregorian/Julian  calendar? $! then display the discordian date and the current moon phase. F $! For more info see http://www.erisx.com/jd/disco/calendar.html among others.  $!A $! The Discordian bit is based on code written by Rev JohnDouglas  WaySide.G $! The moon phases bit is hacked from a post in COMP.OS.VMS, and when I  find0 $! out who wrote it I'll include a named credit. $!  H $! Written by Martyn Pattison on the 36th day of Discord, Anno Mung 3166 $! $    if p1 .nes. "" 	 $    then  $! $! Validate the date $!" $       check_date  = f$cvtime(p1) $       if .not. $status $       then; $          Write sys$output "You specified an invalid date"t $          goto way_out  $       endif    o
 $    endif $e $    Ostring = ""  $    Ostring1 = "" $    Gdate = p1iD $    Gadddays = ",0,3,3,6,8,11,13,16,19,21,24,26,29" !number of days toadd to calc the Julian datef $ C $    DDays =  "Sweetmorn,Boomtime,Pungenday,Prickle-Prickle,Settings Orange" !Days of the week G $    DMonths = "Chaos,Discord,Confusion,Bureaucracy,Aftermath" !SeasonsMH $    Apostledays = "Mungoday,Mojoday,Syaday,Zaraday,Maladay" !Holiday on the 5th of each seasonG $    Seasondays = "Chaoflux,Discoflux,Conflux,Bureflux,Afflux" !Holidaye on the 50th of each season $z $!$ $! Get all the Gregorian date values $!0 $    Gyear = f$cvtime(Gdate,"comparison","year"). $    Gday = f$cvtime(Gdate,"comparison","day")2 $    Gmonth = f$cvtime(Gdate,"comparison","month") $  $! Work out Julian datenH $    Jdate =     ((Gmonth-1)*28) + f$element(gmonth,",",Gadddays) + Gday $! $! St Tibb's day doesn't exist s $!0 $    if Gday .eqs. "29" .and. Gmonth .eqs. "02" 	 $    then 9 $       write sys$output "Oh Joy of Joys 'tis St. Tibb's"e8 $       write sys$output "Let's Eat,drink and be merry."; $       write sys$output "and you can have the day off ;-)"e $       goto way_out
 $    endif $a $! Now get the Discordian date $ - $    DYear = Gyear +  1166      !the easy bit  $    DMonthnum = (Jdate/73)y/ $    Dmonth  = f$element(Dmonthnum,",",DMonths) / $    Dweekdaynum = (Jdate-1) -  ((Jdate-1)/5*5)t, $    Dday = f$element(Dweekdaynum,",",Ddays)) $    Ddaynum =  Jdate - ((Jdate-1)/73*73)i $ , $    if Ddaynum .gt. 9 .and. Ddaynum .lt. 20	 $    then 9 $       remainder = 9 ! Dontcha love the English languaget	 $    else - $       remainder = Ddaynum - (Ddaynum/10*10)i
 $    endif $    -' $    Ostring = f$fao("Today is !AS, the-? !UL!0UL!1%Cst!2%Cnd!3%Crd!%Eth!%F day of !AS, Anno Mung !UL;" -r4                ,Dday,Ddaynum,remainder,Dmonth,Dyear) $: $    if Ddaynum .eq. 5 r	 $    then : $       Apostleday =  f$element(Dmonthnum,",",Apostledays)F $       Ostring1 = Ostring1 +   "it is the Holyday of ''Apostleday', "
 $    endif $u $    if Ddaynum .eq. 50l	 $    thenp8 $       Seasonday =  f$element(Dmonthnum,",",Seasondays)G $       Ostring1 = Ostring1 +   "it is the Seasonday of ''Seasonday', "o
 $    endif $      $ gosub moon_phases  $p $    write sys$output Ostringf $    write sys$output Ostring1	 $way_out:t $e $exit !'f$verify(save_verify)r $  $moon_phases:  $s $!H $! Correct for leapyears now, we didn't do it earlier 'cos of St Tibb's. $!7 $    if Gyear .eqs. (Gyear/4*4) .and. Gmonth .gts. "02"o	 $    thenh $       Jdate = Jdate + 1pH $       if Gyear .eqs. (Gyear/100*100) .and. Gyear .nes. (Gyear/400*400) $       then $          Jdate = Jdate - 1
 $       endife
 $    endif% $ yr = f$cvtime(p1,"absolute","year") A $ cent = 20 ! Stay fixed in this century, by procedure's limits. u $ x = yr/19 + $ golden = yr - 19*x + 1    ! Metonic index - $ epact = ( 11*golden ) + 20  ! Golden number E $ epact = epact + ( ( ( 8*cent ) + 5 )/25 ) - 5 ! Clavian correction eB $ epact = epact - ( ( ( 3*cent )/4 ) - 12 ) ! Leap yr. correction  $ x = epact/30' $ epact = epact - 30*x    ! Take mod 30 E $ if ( epact .eq. 25 ) .and. ( golden .gt. 11 ) then epact = epact+1  > $ if epact .eq. 24 then epact = epact+1  ! Epact is age of the: $ phase = ( ( jDATE-1 ) + epact )  !  moon on January 1st.G                                    ! (Jdate-1) + EPACT gives moon's ages today.! $ x = phase/59    ! 59 = 29.5 x 2  $ phase = phase - 59*xH $ if phase .ge. 29 then phase = ( phase - 29 ) ! PHASE=(PHASE mod 29.5) B                                                ! Age of moon today $ PHASE0 = "new" $ PHASE1 = "new" $ PHASE2 = "new" $ PHASE3 = "waxing crescent" $ PHASE4 = "waxing crescent" $ PHASE5 = "waxing crescent" $ PHASE6 = "waxing crescent"! $ PHASE7 = "in its first quarter"*! $ PHASE8 = "in its first quarter"-! $ PHASE9 = "in its first quarter"e" $ PHASE10 = "in its first quarter" $ PHASE11 = "waxing gibbous" $ PHASE12 = "waxing gibbous" $ PHASE13 = "waxing gibbous" $ PHASE14 = "full" $ PHASE15 = "full" $ PHASE16 = "waning gibbous" $ PHASE17 = "waning gibbous" $ PHASE18 = "waning gibbous" $ PHASE19 = "waning gibbous" $ PHASE20 = "waning gibbous"! $ PHASE21 = "in its last quarter"a! $ PHASE22 = "in its last quarter"d! $ PHASE23 = "in its last quarter"t! $ PHASE24 = "in its last quarter"  $ PHASE25 = "waning crescent"t $ PHASE26 = "waning crescent"  $ PHASE27 = "waning crescent"  $ PHASE28 = "waning crescent"t $ PHASE29 = "new"  $ PHASE = PHASE'PHASE'3 $ Ostring1 = Ostring1 + "and the moon is ''phase'."- $  $! $return    ------------------------------  # Date: Wed, 19 Apr 2000 18:30:21 GMTu From: cgunhouse8375@my-deja.comB3 Subject: Command Line History and C++ input streamsN) Message-ID: <8dktv8$404$1@nnrp1.deja.com>b  , I am using DEC C++ Ver 6.2 on an OpenVMS 6.2. Alpha system.  The problem I am having is that0 when I press the cursor up key on the keyboard I/ get the last command entered in the VMS commandV- line passed to cin instead of the ANSI escapew. sequence.  If set the terminal setting to "set. term/pasthru" prior to loading my program, the) escape sequences get passed to my programo, properly.  Is there a way of doing this from. within my C++ program?  Actually my problem is- with CURSES, but the problem manifests itselfa	 with cin.i     Craig,    & Sent via Deja.com http://www.deja.com/ Before you buy.d   ------------------------------  # Date: Wed, 19 Apr 2000 19:15:54 GMT== From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)t7 Subject: Re: Command Line History and C++ input streamsv0 Message-ID: <009E8D9A.9ED68D92@SendSpamHere.ORG>  K In article <8dktv8$404$1@nnrp1.deja.com>, cgunhouse8375@my-deja.com writes:t- >I am using DEC C++ Ver 6.2 on an OpenVMS 6.2n/ >Alpha system.  The problem I am having is thata1 >when I press the cursor up key on the keyboard I.0 >get the last command entered in the VMS command. >line passed to cin instead of the ANSI escape/ >sequence.  If set the terminal setting to "set-/ >term/pasthru" prior to loading my program, ther* >escape sequences get passed to my program- >properly.  Is there a way of doing this from  >within my C++ program?D    ? Yes!  It's the same as with any program in any language on VMS.   I Call SYS$QIO with IO$_SETMODE function and use/specify the characteristic H TT2$M_PASTHRU.  Check the documentation in the I/O User's Guide for fur-H ther details.  If you think I'm gonna write a quick ditty in C++ to showK you how, you're sadly mistaken.  I use only readable programming languages.      >  Actually my problem isr. >with CURSES, but the problem manifests itself
 >with cin. >u >a >Craig >a >o' >Sent via Deja.com http://www.deja.com/h >Before you buy.  G If I was using C++, I have a problem with CURSES too!  F***ing C++ is ao cin.  ;) --N VAXman- OpenVMS APE certification number: AAA-0001           VAXman@TMESIS.COM  L GNU Freeware -- What does the GNU *really* stand for?  Garbage!  Not Usable!   ------------------------------  % Date: Wed, 19 Apr 2000 20:21:28 -0400t+ From: "David Turner" <d_b_turner@yahoo.com> " Subject: DEC (ES40) MS610-EA $9999/ Message-ID: <sfsj6b21in2134@corp.supernews.com>s  + We have a new, unused DEC MS610-EA in stocke   Price is $9999 List price is $27,000 !a   Call if interested  	 toll freeE USA 912 447 6622   -- David Turner Island Computers US Corporation- 2700 Gregory Streetp Savannah GA 31404  Tel: 912 447 6622p Fax:912 201 0096 sales@islandco.com   ------------------------------  % Date: Wed, 19 Apr 2000 15:00:52 -0400n. From: "Paul M Boynton" <pmboynton@us.hsbc.com> Subject: EMC Disk StorageS4 Message-ID: <852568C6.00688C18.00@mail2.us.hsbc.com>  F We are in the process of evaluating EMC disk storage to replace Compaq StorageWorksN controllers and disk on an OpenVMS CI cluster. Does anyone have any experience" good or bad with EMC disk storage?     Paul Boynton HSBC Bank, USA Buffalo, NYs Pmboynton@us.hsbc.comi   ------------------------------  % Date: Wed, 19 Apr 2000 15:15:44 -0500N+ From: Don Rogstad <Don.Rogstad@dalsemi.com>n Subject: RE: EMC Disk StoragebH Message-ID: <EA5FE99F5DEAD311A6CB00805F199992BCA806@misnts1.dalsemi.com>  < >----- Paul M Boynton [mailto:pmboynton@us.hsbc.com] wrote:  >Subject: EMC Disk Storage >nG >We are in the process of evaluating EMC disk storage to replace Compaqs
 >StorageWorks O >controllers and disk on an OpenVMS CI cluster. Does anyone have any experiencel# >good or bad with EMC disk storage?E >5# >----------------------------------.   Paul:o      We have a EMC "Symmetrix - Integrated Cache Disk Array" disk system connect to an Alpha 2100.  We have found that they are not nearly as flexible as StorageWorks disks.  A couple of points:      1) Service must be provided by EMC.  They want a "dial-up" modem line connected directly to their box for service.  In our environment, that's a big NO-NO so our service has been somewhat sporadic.      2) The console front end is a laptop PC.  To load new firmware, the engineer (just out of high school) brought a box full of 3.5" floppies (I think there were 15 of them), and proceeded to copy them onto the PC.  Then he had to run a program on the PC to download them to the disk controller.  He had problems and we had to hook a modem for remote support to dial into the laptop and do the procedure over.  I think the problem turned out to be a misplaced file.   Needless to say, it took them over 2 days to  upgrade the firmware.     3) Their idea of fail-over redundancy is to have two seperate controllers connecting to the same disks.  This means that you need two seperate controllers on OpenVMS and the same disks show up on the two controllers, e.g. DKA100 and DKB100.  Even though you have two devices, NEVER mount both of them or data corruption will naturally occur.  To fail-over a device, you must dismount the device from the first controller and then mount it on the second controller.r      4) EMC insists on being the only ones to configure the box.  You tell them your storage requirements and they try to configure it.  I don't think they configure all the space we bought, but I was not working here when the box was bought and configured.      5) Until the latest firmware release, the EMC box each disk mount took 90 seconds to occur during boot.  The EMC box would not spin up the disks until a mount request was given.  This cause out boot up time to increase noticeably.  (You never realize how *long* a minute is until you are staring at a screen waiting for mount completion ;-)  This was supposed to be one of the firmware fixes, but we haven't rebooted to confirm this.  y  On the PLUS side, the disks do show up as normal ODS-2 disks and OpenVMS has never had a problem using them.  Fortunately, we have not had to try their fail-over plan.  Overall, if the purchase hadn't been political, I doubt we would have bought the box.  Our current plans are to migrate as much as possible to StorageWorks and use the EMC box as a big scratch disk storage.5   Don Rogstad  Dallas Semiconductor   ------------------------------  % Date: Wed, 19 Apr 2000 14:47:42 -0700l  From: "Bob Fera" <bob@serve.net> Subject: Re: EMC Disk Storagec2 Message-ID: <1KpL4.183$VW1.411@newsfeed.slurp.net>  J We are in the process of installing a Symmetrix right now. In fact, the CEK is here working on the config as I write this. We researched both platforms 2 and came to a few (hopefully correct) conclusions.  B The Storageworks system would be more convenient because it's userH configurable. However, the amount of data protection in the StorageworksL system doesn't seem to come close to what the Symmetrix can do. Each disk inI the Symm has multiple SCSI paths to it, where the SW has only one. So, if G you lose a disk channel on the SW, you lose each drive on it -- and if,tK heaven forbid, you have two disks in one RAID set on the same SCSI channel,lF you lose it all. The Symm also does a lot of proactive error checking,I attempting to detect failing devices before they become a problem, and it K has the ability to contact service automatically when it detects a problem.d  L Also, I have a sneaking suspicion that the Storageworks controllers are madeH by CMD (any confirmation here?). I have used CMD controllers for severalJ years here and, well, let's just say I've had more trouble with controllerI firmware failing than with disks going bad. The quality of the componentsa/ isn't even in the same league as the Symmetrix.p  L Overall, then, I think that the inconvenience of not being able to configureL the system myself is a small price to pay for what looks like a huge gain inH data protection capability. Of course, the Symm isn't hooked up yet! :-)+ I'll know more in the next couple of weeks.l   Regards,   Bob Fera  Zenith Information Systems, Inc. Los Angeles, CA   9 "Paul M Boynton" <pmboynton@us.hsbc.com> wrote in message . news:852568C6.00688C18.00@mail2.us.hsbc.com... >e >n >eH > 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 anyh
 experience$ > good or bad with EMC disk storage? >e >r > Paul Boynton > HSBC Bank, USA
 > Buffalo, NYn > Pmboynton@us.hsbc.com  >e >r   ------------------------------  % Date: Wed, 19 Apr 2000 15:07:51 -0700h  From: "Bob Fera" <bob@serve.net> Subject: Re: EMC Disk Storagei3 Message-ID: <W0qL4.203$VW1.1266@newsfeed.slurp.net>u   A few specific comments:  H >    2) The console front end is a laptop PC.  To load new firmware, theI engineer (just out of high school) brought a box full of 3.5" floppies (IiK think there were 15 of them), and proceeded to copy them onto the PC.  Then J he had to run a program on the PC to download them to the disk controller.J He had problems and we had to hook a modem for remote support to dial intoK the laptop and do the procedure over.  I think the problem turned out to bey@ a misplaced file.   Needless to say, it took them over 2 days to >  upgrade the firmware.  K Not optimal, but, as I understand it, there would be no down time involved.    >1B >    3) Their idea of fail-over redundancy is to have two seperateG controllers connecting to the same disks.  This means that you need twopE seperate controllers on OpenVMS and the same disks show up on the two G controllers, e.g. DKA100 and DKB100.  Even though you have two devices,rE NEVER mount both of them or data corruption will naturally occur.  To1J fail-over a device, you must dismount the device from the first controller+ and then mount it on the second controller.  >n  L You only need this configuration to handle a *host* SCSI controller failure.C Redundant paths to the disks are built in to the Symmetrix, but areeL invisible to the host. Any component failure within the Symmetrix is handledD transparently to the operating system. Also, I believe that VMS V7.2% supports host SCSI channel fail over.-   Regards,   Bob Fera  Zenith Information Systems, Inc. Los Angeles, CAn   ------------------------------  # Date: Wed, 19 Apr 2000 22:12:52 GMTs From: Dirk Munk <munk@home.nl> Subject: Re: EMC Disk Storagei' Message-ID: <38FE2F62.18576637@home.nl>s   Paul M Boynton wrote:  > H > We are in the process of evaluating EMC disk storage to replace Compaq > StorageWorksP > controllers and disk on an OpenVMS CI cluster. Does anyone have any experience$ > good or bad with EMC disk storage?  C CI of course is veeerrrryyyyyy slow. It was a beautiful concept but-A should be replaced now, because the max. speed is about 8.5 MB CIi adapter.  > Take a look at fibre channel storage works with the new ModulaG disks/shelfs. It is much faster than CI and EMC, and a lot cheaper than.G EMC. This summer there will be 15000 rpm modula drives... Of course youb2 will need VMS 7.2-1 and suitable Alpha systems.      >  > Paul Boynton > HSBC Bank, USA
 > Buffalo, NY. > Pmboynton@us.hsbc.com!   ------------------------------  # Date: Wed, 19 Apr 2000 22:22:25 GMTn From: itjck01@my-deja.comf Subject: Re: EMC Disk Storage ) Message-ID: <8dlbih$jnu$1@nnrp1.deja.com>d  4 In article <852568C6.00688C18.00@mail2.us.hsbc.com>,1   "Paul M Boynton" <pmboynton@us.hsbc.com> wrote:i >s >oH > We are in the process of evaluating EMC disk storage to replace Compaq > StorageWorksE > controllers and disk on an OpenVMS CI cluster. Does anyone have any 
 experience$ > good or bad with EMC disk storage? >u > Paul Boynton > HSBC Bank, USA
 > Buffalo, NY  > Pmboynton@us.hsbc.comB >t >, .g .a .r! <bias flame on, with good reason>y  A Bad experience in my shop.  VERY VERY BAD EXPERIENCE!  We HAD twofE AlphaServer 4100's and an AlphaServer 2100 hooked to EMC storage with,F KZPSAs.  They all are NOW hooked to StorageWorks.  The problems we had% in no particular order of importance:o  E 1) EMC drives would go into mount verify and offline.  (This happened   about 4 times in about 3 years.)F 2) EMC drives would corrupt INDEXF.SYS and BITMAP.SYS.  (This happened  about 6 times in about 3 years.)D 3) EMC service would reconfigure/reallocate disk storage incorrectlyF and/or inappropriately such that we lost data and had to go to backupsB with much downtime.  (This happened twice to us in about 3 years.)@ 4) EMC controller died and we lost data on VMS, NT, Solaris, andB Novell; which was never suppose to happen according to EMC.  (This& happened once to us in about 3 years.)G 5) Had EMC onsite with mount verify and drive offline problem occurringtE for about 2 weeks straight, and they (about 4 folks and remote staff) C were never able to fix or come up with a reason or solution.  (This"9 happen once in about 3 years, and was last straw for us.)tH 6) EMC drive performance was bad. (Shared storage -> shared performance)F 7) No backup client for OpenVMS that worked.  EDM client caused systemE to crash.  They never came up with good software over the 2 1/2 years + or so that we had EMC storage with OpenVMS.m@ 8) EMC upgraded Symmetrix firmware, but they never tested if EMCG controllers and I/O sub-system would work/continue to work with OpenVMS E and KZPSAs.  We could not demostrate that the firmware upgrade caused D any problems, but the fact that it came out that they did no testing6 did not leave us with a warm and fuzzy feeling at all.  C I would never go to EMC storage for OpenVMS.  And if had to do over B would reconsider ever going at all with shared storage and lack ofB control over hardware.  If you think you are going to be I/O boundE versus CPU or memory bound, then shared disk storage is only going tou provide worst performance.  F In my opinion, EMC storage is very expensive for what you get.  And if@ the EMC storage is not expensive, then it will be with expensiveC service contract that you have to have.  EMC sales is known to dropm? initial price and then charge like h_ll on service maintenance.   D We went to StorageWorks fibre storage with HSG80s and KGPSAs, and weG saw about a twofold increase in I/O performance over EMC Symmetrix.  WefG brought a StorageWorks configuration with only 66 drives as compared to G the 188 drives we came off of on EMC Symmetrix, and we still saw bettero performance.  E I have it on pretty good word that EMC tried to qualify their storageeF with Compaq OpenVMS Alpha systems, but failed.  I think there are goodC technical reasons why they failed.  I just can't tell you what theycG are, I can only tell of the end results that we experienced at my site.   8 Some questions you might ask of yourself and others are:  G I know there are OpenVMS/EMC shops out there, but how many?  And do you B want to be one of only a few with dwindling/non-support?  How many@ OpenVMS reference accounts can EMC provide?  How many can CompaqD provide?  How well can you manage finger pointing between vendors if you have a problem?5   JKoska@bender.com  Matthew Bender & Co., Inc.
 518-4878-3255t   <flame off>=   I feel better now.    & Sent via Deja.com http://www.deja.com/ Before you buy.=   ------------------------------  % Date: Wed, 19 Apr 2000 19:06:45 -0700n- From: David R Barnes <dbproductions@juno.com>8 Subject: Re: EMC Disk StorageH: Message-ID: <20000419.191136.500.0.dbproductions@juno.com>   Hello Paul;1    A 	I have an Alphaserver 8400 and Alphaserver 4100 connected to EMC G Symmetrix , about 400+ GB on the 8400 and 200+GB on the 4100.  I highlybI recommend it!  Very very fast!!!!  I am close to you also... in  Buffalo,  NY.n     							David Barnes  							dbproductions@juno.comi  I On Wed, 19 Apr 2000 15:00:52 -0400 Paul M Boynton <pmboynton@US.HSBC.COM>  writes:tB > We are in the process of evaluating EMC disk storage to replace  > Compaq > StorageWorksF > controllers and disk on an OpenVMS CI cluster. Does anyone have any  > experience$ > good or bad with EMC disk storage? >  >  > Paul Boynton > HSBC Bank, USA
 > Buffalo, NY  > Pmboynton@us.hsbc.com-   David R Barnes dbproductions@juno.com3 =-=-=-=-=--=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-j# Netbsd 1.3 / Linux / Unix, VMS guruu  @ ________________________________________________________________( YOU'RE PAYING TOO MUCH FOR THE INTERNET!% Juno now offers FREE Internet Access!t? Try it today - there's no risk!  For your FREE software, visit:c  http://dl.www.juno.com/get/tagj.   ------------------------------  # Date: Wed, 19 Apr 2000 22:42:50 GMTe From: itjck01@my-deja.comm Subject: Re: EMC Disk Storager) Message-ID: <8dlcp7$l27$1@nnrp1.deja.com>m   My comments in-line.  2 In article <1KpL4.183$VW1.411@newsfeed.slurp.net>,#   "Bob Fera" <bob@serve.net> wrote:EE > We are in the process of installing a Symmetrix right now. In fact,v the CEC > is here working on the config as I write this. We researched both-	 platformsa4 > and came to a few (hopefully correct) conclusions. >HD > The Storageworks system would be more convenient because it's user= > configurable. However, the amount of data protection in then StorageworksF > system doesn't seem to come close to what the Symmetrix can do. Each disk in D > the Symm has multiple SCSI paths to it, where the SW has only one.  F Not sure how you thought about configuring your StorageWorks, but mineF is dual redundant throughout, and could more redundant if I thought itF was worth it.  I have 2 paths to each of my datasets.  I have 2 KGPSAsC (Emulex), 4 Brocade switches, a total of 8 HSG80s in dual redundanteF mode, and dual power supplies all round.  I do both hardware mirroring! and host-based volumen shadowing.   G I proved to myself I could drop a path and it would failover flawlesslyl? several times before putting into my development and productions
 environments.T   So, ifE > you lose a disk channel on the SW, you lose each drive on it -- and  if,hD > heaven forbid, you have two disks in one RAID set on the same SCSI channel,H > you lose it all. The Symm also does a lot of proactive error checking,D > attempting to detect failing devices before they become a problem, and itD > has the ability to contact service automatically when it detects a problem. >u  : Ah, DECevent with DSNlink does this also with StoragWorks.  E > Also, I have a sneaking suspicion that the Storageworks controllers  are madeB > by CMD (any confirmation here?). I have used CMD controllers for several A > years here and, well, let's just say I've had more trouble with 
 controller@ > firmware failing than with disks going bad. The quality of the
 components1 > isn't even in the same league as the Symmetrix.r >sD > Overall, then, I think that the inconvenience of not being able to	 configureiF > the system myself is a small price to pay for what looks like a huge gain iniA > data protection capability. Of course, the Symm isn't hooked upa yet! :-)- > I'll know more in the next couple of weeks.a >@
 > Regards, > 
 > Bob Fera" > Zenith Information Systems, Inc. > Los Angeles, CA0 >n   JKoska@bender.como Matthew Bender & Co., Inc. 518-487-3255    & Sent via Deja.com http://www.deja.com/ Before you buy..   ------------------------------  % Date: Wed, 19 Apr 2000 20:20:01 -0400p" From: Dan Sugalski <dan@sidhe.org> Subject: Re: EMC Disk Storagee8 Message-ID: <4.3.1.0.20000419200743.01d139b0@24.8.96.48>  * At 02:47 PM 4/19/00 -0700, Bob Fera wrote:C >The Storageworks system would be more convenient because it's user I >configurable. However, the amount of data protection in the Storageworkse@ >system doesn't seem to come close to what the Symmetrix can do.  ( "Seem" would be the operative word here.  
 >Each disk inuC >the Symm has multiple SCSI paths to it, where the SW has only one.k  F This isn't quite right unless the drives on the EMC system are really L custom. There's only one SCSI interface coming off the back of the disk, so H if the shelf it's plugged into goes, the drives all go. Luckily shelves L generally don't fail, since they're pretty much passive doodads. (You could K take one out with a sledgehammer, I suppose, but that's why you use volume v shadowing too... :)   H Now, if you mean each StorageWorks controller has a single path to each I disk then you're right. That's why you run with dual controllers--if one dH controller loses a scsi connection the other picks up the drives on it. K This is likely the same as what the EMC gadget's doing, though it could be sI they have a single controller with two paths, in which case your failure  H exposure's actually greater with the EMC system. (Losing the controller F would lose all your drives, and controllers do get fried occasionally)   >So, ifoH >you lose a disk channel on the SW, you lose each drive on it -- and if,L >heaven forbid, you have two disks in one RAID set on the same SCSI channel, >you lose it all.e  H Which is why you have dual-redundant controllers and keep your RAID and K mirrorsets on different shelves. (And, of course, shadow your storage to a vK system in another building, just in case your main cabinet takes a bad hit)   6 >The Symm also does a lot of proactive error checking,J >attempting to detect failing devices before they become a problem, and itL >has the ability to contact service automatically when it detects a problem.  H The StorageWorks systems don't do much proactive drive probing, so they E won't generally see a bad spot or problem until you run across it in tH practice. It's been my experience that drives either fail gradually, in I which case you'll see it before it dies, or just fail abruptly, in which lL case it doesn't matter much. Besides, that's why you have the things RAIDed D or mirrored in the first place--when something does start to go the I controller swaps the bad drive out and brings one of your spares on-line   for you.  K SWCC seems pretty good about noticing things like drive failures. Besides, sK you do walk your computer room looking for the amber warning lights anyway g just to be sure, right? :)   					Dan  I --------------------------------------"it's like this"-------------------o2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and eveno;                                       teddy bears get drunka   ------------------------------  # Date: Thu, 20 Apr 2000 01:04:12 GMT ( From: Terry Kennedy <terry@gate.tmk.com> Subject: Re: EMC Disk Storageo' Message-ID: <FtAIz0.BDp@spcuna.spc.edu>x  . Paul M Boynton <pmboynton@us.hsbc.com> writes:H > We are in the process of evaluating EMC disk storage to replace Compaq > StorageWorksP > controllers and disk on an OpenVMS CI cluster. Does anyone have any experience$ > good or bad with EMC disk storage?  E   Years ago, EMC bid on a memory subsystem for a VAX-11/70 at my thentI employer (SPC) and was the low bidder. After sending them a PO, no memory H appeared. After weeks of calling and getting various runarounds, we wereG told that the sales rep was fired for sending out lowball bids and they F had no intention of honoring their quote. Since then I have refused to do business with them.  K   At a datacenter where I have equipment, I observed an "all hands on deck"tJ EMC emergency service where EMC folks from all over were camped out at theI site trying to get a customer back up. This ran for at least the 4 nightseJ that I was there doing unrelated work. I believe they got the customer go-H ing again by the time it was over. This was a truly *huge* customer withG terabytes of storage. I don't have any details on what the problem was,tJ but it was apparently major - senior management from both the customer andK EMC were there 24 hours/day until it was fixed. As far as I know, that cus-I tomer is not a VMS shop.  - 	Terry Kennedy             http://www.tmk.comy5         terry@tmk.com             Jersey City, NJ USAg   ------------------------------  # Date: Wed, 19 Apr 2000 18:59:44 GMT * From: Steven Whatley <swhatley@blkbox.com>T Subject: Re: How can you find the exact size of a file in bytes using a VMS command?1 Message-ID: <956170783.819249@newstest.texas.net>u   Rick <rw712@my-deja.com> wrote:i@ > The reason I asked the question was to determine if a file wasD > transfered to a VMS box correctly using FTP. It was an Oracle dumpF > file, and the individual transfered the file using ASCII mode. Well,H > that did a bit of translation on it's own, which would modify the fileH > (such as converting CRNL to NL), but not so much as to change the sizeH > of the file by a number of blocks. How could I tell easily if the fileI > had been altered? Oh, I did not expect this person to transfer a binary H > file use text mode... It took quite a while to figure out that was the0 > cause of the odd messages in loading the dump.  D I'm constantly fighting this problem of users sending ASCII files inF binary (from a PC or UNIX) or binary files sent in ASCII.  I have justG learned what what errors are caused by using the wrong transfer mode.  n  H Had a user say, "I've always FTPed ASCII and binary files in binary modeH from a Mac for years with no problems."  Yes, FTPing from a Mac seems toC work fine but Windows/DOS and UNIX use different characters as line 2 separators.  The users are finally catching on. :)   Later, Steven   ------------------------------  # Date: Thu, 20 Apr 2000 02:26:54 GMTt$ From: Ed Wilts <ewilts@mediaone.net> Subject: Re: HSJ80 experience?, Message-ID: <38FE6AEE.2F5D8FAA@mediaone.net>   Bob Kaplow wrote:  > U > In article <38F5197C.EE02CD2E@mediaone.net>, Ed Wilts <ewilts@mediaone.net> writes: L > > The CI is propietary technology but you can purchase a license to do theH > > same things the HSJ does.  I ran a pair of CMD Hawks for a while andH > > they were rock solid.  While the early HSJ-40 controllers were being > - > Ever check the CIPCA? They are made by CMS.o  F Typo I assume.  They're made by CMD.  I was a CMD Hawk customer at theE time they were developing the Hawk and was actually given a choice ofiC buying one from them or from Digital.  CMD could only sell to theirl existing customers back then.  n   	.../Edt   -- e Ed Wilts Mounds View, MN, USA mailto:ewilts@mediaone.net   ------------------------------  # Date: Wed, 19 Apr 2000 17:51:12 GMTc' From: Colin Blake <colin@theblakes.com>h! Subject: Re: Mozilla M15's out...i- Message-ID: <38FDF20B.F9EEAF3F@theblakes.com>r   Tim Shoppa wrote:f  D > It's good news that it's out, but why is the VMS release 30 Mbytes9 > while the other (same version) releases are 5-7 Mbytes?e  Y For a start, all the images are linked /traceback. This helps when debugging but it has a @ cost. Its surprising just how much bigger this makes the images.  : The OpenVMS kit includes GTK+. The other platforms do not.  U The other kits are in zip format. If you zip the OpenVMS PCSI file its only 17 Mbytesn0 instead of 30Mb for the self-extracting version.   ------------------------------   Date: 19 Apr 2000 17:59:20 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)! Subject: Re: Mozilla M15's out...d6 Message-ID: <8dks5o$kg2$2@mailint03.im.hou.compaq.com>  W In article <38fde767$1@news.kapsch.co.at>, eplan@kapsch.net (Peter LANGSTOEGER) writes: P :MOZILLA is at least available on OpenVMS, but why isn't the NETSCAPE beta, too?N :I can't really follow the arguments, that NETSCAPE didn't release the code...  K   The Mozilla beta is effectively part of the Netscape beta process -- the  H   Netscape browser is directly based on the Mozilla browser, and we are H   tracking the Mozilla updates, and I expect this will be useful as partJ   of getting over onto the Netscape browser...  I would expect to see the L   Mozilla browser stabilize before the Netscape browser stabilizes...  (And J   no, I don't have a schedule for this, as I have learned not to even try '   to predict schedules in this area...)   N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 19 Apr 2000 15:31:32 -0400E# From: sol gongola <sol@adldata.com>03 Subject: Re: OpenVMS mail question: Attached files?,' Message-ID: <38FE0994.540C@adldata.com>r  G With x windows, you can use the mailer that comes with the vms version )A of netscape. For terminal sessions, the Pine mail program handles  attachment.e  F You can also format a mail message yourself in a program, encoding theE attachment (with uuencode module), including the headers appropriate uA for the type of attachment you are using. Invoke the mail programrB as a system command specifying destination, subject, and preformat2 file name as command line parameters for vms mail.   sol gongolar  	 > Hi all,dL > We have the need to be able to send attached files from our OpenVMS mailerI > to mail recipients.  We are looking at sending HTMP files from within ad% > purchase order system thru OpenVMS.  > , > How do we do it, or isn't  this available? > L > Alternately: Any other mailers availbale for OpenVMS that would be able to > send attached files? >  > Thanx D > Thys de Wet, University of Stellenbosch, Stellenbosch South Africa > Email: thys@it.sun.ac.za   ------------------------------  # Date: Wed, 19 Apr 2000 18:17:14 GMTw, From: kaplow_r@eisner.decus.org (Bob Kaplow)* Subject: Re: Scheduler & PSDC alternatives' Message-ID: <2000Apr19.141714.1@eisner>r  ] In article <8dii52$o42$1@lure.pipex.net>, "Graham HAZLEGREAVES" <qg38@dial.pipex.com> writes: N > First off, I'd like to say a big thank you to all of you who posted replies.N > You saved me a lot of effort trawling the net and phoning companies.  We now' > have a big file for future reference.g >   > Why do I say future reference? > G > Well, I found out today that at some time in the past somebody mady adM > brilliant decision when the 3 products in question were orginally purchasedf2 > and licensed.  They took out perpetual licenses! > L > When the products switched over to CA they then had a choice for support -M > CA or Digital.  Once again somebody's brain was in gear and they decided toi > go with Digital.  J How are you still getting DECpaq support for these products? They (DeQ) noI longer have them on the ConDist. They no longer have the support folks or G the developers. The Unix version of DECscheduler doesn't even exist anyo more!k  B I'd like nothing better than to have all of these back, and on DeQ$ maintenance. And to tell CA to .....   	Bob Kaplow	  E SPAM:	spamrecycle@ChooseYourmail.com	uce@ftc.gov	postmaster@127.0.0.1h   ------------------------------  # Date: Thu, 20 Apr 2000 00:43:02 GMTe- From: "Dave Pampreen" <davepampreen@home.com>tG Subject: Re: Sending MIME attachments in OpenVMS 7.1-H1 and TCPware 5.3b; Message-ID: <qosL4.34725$h01.284613@news1.rdc1.mi.home.com>y  J I did the MIME thing.  The help was a bit lacking...When I enter NEW I get the following:  F %DCL-W-IVVERB, unrecognized command verb - check validity and spelling  \TPU\  K And it just sits there.  CTRL-T tells me I'm running MIME but what gives???a   Dave    ? "Mark Buda" <buda@tabasco.zko.dec.no.spam.com> wrote in messagee% news:8dg591$ofv$1@lead.zk3.dec.com...n" > Take a look at the mime utility. >u > $ MIME :==$sys$system:mime > $ mime >  > Take care,	 >  - markK >aE > Steve Freegard <stevenf@alpha-centori.demon.co.uk> wrote in message== > news:955999311.19338.0.nnrp-11.c1ed965a@news.demon.co.uk...o > > Hi,e > >bG > > I was wondering if anybody knows how to send MIME attachments under 	 > VMS.  ImJ > > have tried MAIL/FOREIGN but this seems to preserve the VMS file format > and-2 > > make text files impossible to read in Windoze. > >lC > > Basically what I am trying to do is send out mail from VMS witht
 > multipleG > > spool files attached, so the recipient can view/print the report in 
 > Windows. > >r > > Thanks in advancel > >o > >a > > Steve Freegard > >n > >  > >	 >v   ------------------------------  # Date: Wed, 19 Apr 2000 16:34:40 GMT2. From: Michael Austin <maustin@nc.prestige.net>( Subject: Re: Using DiskMizer with Oracle/ Message-ID: <38FDECF9.AB652AF5@nc.prestige.net>t  , This is a multi-part message in MIME format.& --------------D096E38122C0344D4E429428* Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bitC  B My general rule of thumb (and some may not agree..) is to only useH database tools to mess with database files. Unless there are no databaseH utilities for those procedures.  You would probably get the same resultsG if you export/imported your tables and create new datafiles when you doRA it.  Remember Database Record fragmention is not the same as filesG fragmentation.  When you create an Oracle file, you generally create itg@ at the maximum size it will be (then you add another file to theE tablespace).   Later versions have the "autoextend" feature, but this  should still not be a problem.E To clean up record fragmentation you will need to do an export/imports anyway.p   Michael Austin# DBA Consultant  -- OracleRdb/7/8/8ii     Ed Burke wrote:a  3 > We are considering using DiskMizer from Symark tor6 > compress some Oracle database disks.  We already use6 > DiskMizer for other storage and have been happy with7 > it, but our management would like to hear about otherr. > people's experiences with Oracle data before2 > committing to using DiskMizer for our databases.6 > Symark could not give me any references because they/ > don't keep track of what types of files theirs8 > customers use DiskMizer for.  So, does anyone have any8 > experiences they can share or know of any gotchas that( > aren't mentioned in the documentation? >n3 > We are running DiskMizer v3.2, Oracle 7.3.2.3 and 1 > 8.0.x on an AlphaServer 8400 running VMS 7.1-2.G >s	 > Thanks,  >r >   Ed Burke >   clegg67@yahoo.comn >a4 > __________________________________________________ > Do You Yahoo!?. > Send online invitations with Yahoo! Invites. > http://invites.yahoo.com  & --------------D096E38122C0344D4E429428- Content-Type: text/x-vcard; charset=us-ascii;v  name="maustin.vcf"o Content-Transfer-Encoding: 7bito, Content-Description: Card for Michael Austin  Content-Disposition: attachment;  filename="maustin.vcf"i   begin:vcard  n:Austin;Michael I tel;work:704-947-1089i x-mozilla-html:FALSE org:Michael Austin, Incr
 adr:;;;;;; version:2.1g+ email;internet:michaelaustininc@hotmail.com  title:Presidentt x-mozilla-cpt:;0 fn:Michael E. Austin	 end:vcardy  ( --------------D096E38122C0344D4E429428--   ------------------------------  + Date: Wed, 19 Apr 2000 12:02:08 -0700 (PDT)u" From: Ed Burke <clegg67@yahoo.com>( Subject: Re: Using DiskMizer with Oracle? Message-ID: <20000419190208.11661.qmail@web3003.mail.yahoo.com>e  / > How does the cost of the compression softwaret compare with addingm6 > additional storage?  Disks are cheap these days.  At first glance, I 2 > think that anything that gets between Oracle and fast disk access could > not be good.  / The issue is partly the cost of the disk (we'rey4 looking at a half-terabyte of growth this year), but5 more the cost of the additional controller, the spaceu3 in the data center, etc.  We already own DiskMizer, 1 and there's no monetary cost to adding additional  DiskMizer volumes.  0 In theory, DiskMizer shouldn't get in the way of0 accessing the disk.  It does the compression and1 decompression on the fly, transparent to VMS.  Ind3 fact, since it reduces I/O, it could even be faster-5 than a regular disk.  (In exchange for using some CPU . to handle the compression, you only write (for4 example) half as much to disk if you are getting 2:16 compression).  We have found this to be true with very2 large text files, and Symark says it would be true6 with Oracle, but some confirmation (or rejection) from4 others who have gone this route could help us decide whether to pursue this further.t     Ed  2 __________________________________________________ Do You Yahoo!?, Send online invitations with Yahoo! Invites. http://invites.yahoo.com   ------------------------------  % Date: Wed, 19 Apr 2000 19:56:32 -0400l' From: "Bill Todd" <billtodd@foo.mv.com>c( Subject: Re: Using DiskMizer with Oracle( Message-ID: <8dlgud$s0v$1@pyrite.mv.net>  - Ed Burke <clegg67@yahoo.com> wrote in messagel9 news:20000419190208.11661.qmail@web3003.mail.yahoo.com...a   ...o  2 > In theory, DiskMizer shouldn't get in the way of2 > accessing the disk.  It does the compression and3 > decompression on the fly, transparent to VMS.  Ine5 > fact, since it reduces I/O, it could even be fasteru7 > than a regular disk.  (In exchange for using some CPUi0 > to handle the compression, you only write (for6 > example) half as much to disk if you are getting 2:18 > compression).  We have found this to be true with very > large text files,f  E Compressing a large sequentially-accessed file is an optimal case fornJ compression:  the compressed data remains contiguous and in the same orderL on disk.  Compressing randomly-accessed data such as database pages is a lotG more iffy:  that data *may* be accessed sequentially in table scans, soaL should ideally remain contiguous and in-order (and Oracle will likely try toL keep it so), but if something else is revectoring the data based on where it? will 'fit' after random-update compression this may not happen.u  H There can certainly be exceptions, but my inclination is to believe thatJ compression is primarily useful for sequentially-accessed data that is not modified in place.  YMMV.t   - bill  !  and Symark says it would be truea8 > with Oracle, but some confirmation (or rejection) from6 > others who have gone this route could help us decide! > whether to pursue this further.  >i >   Ed >c4 > __________________________________________________ > Do You Yahoo!?. > Send online invitations with Yahoo! Invites. > http://invites.yahoo.com   ------------------------------   Date: 19 Apr 2000 17:49:00 GMT* From: helbig@astro.rug.nl (Phillip Helbig) Subject: Re: Verify of Backups. Message-ID: <8dkric$53i$2@info.service.rug.nl>  9 In article <38FDBC5D.5058B2D3@ccinet.ab.ca>, Gord CoulmanW <gcoulman@ccinet.ab.cawrites:   C > I have never seen verify catch any problems other than files thatoB > changed since the backup pass (not really a problem).  When tapeD > problems have occurred (bad tapes, dirty heads, etc) backup alwaysI > bagged an error during the backup pass and never even got to the verifym	 > pass.  i   ??? SET NOON, ON ERROR,...   ------------------------------   Date: 19 Apr 2000 17:53:04 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) Subject: Re: Verify of Backups6 Message-ID: <8dkrq0$kg2$1@mailint03.im.hou.compaq.com>  X In article <38FDEBFC.F32D34A2@ohio.edu>, "Richard D. Piccard" <piccard@ohio.edu> writes: :Hoff Hoffman wrote:L :>   That said, if I have unique or difficult-to-replace data, I usually use= :>   BACKUP /VERIFY -- otherwise, why bother with the BACKUP?l  K :There is one failure scenario that the /VERIFY won't detect:  mis-aligned  I :heads.  In a previous position, we were passing the /VERIFY comparison, lH :but could not restore the tapes on any other drive because of the head  :mis-alignment.k  I   True, but how much of a problem that is depends on your willingness to  F   resort to hardware fixes, assuming that the hardware permits it...    I   This alignment problem was something I saw fairly regularly many years  I   back -- big, hot, power-hungry 9-track 1600/6250 vacuum-column magtape  I   (nagtape? :-) drives -- and just tweaked the heads on the other drives  H   the appropriate distance out of alignment.  (We didn't usually need toI   do this, as we usually picked up on any head alignment problems rather -G   quickly given the frequency with which we swapped tapes around among hL   the various drives...  Long before the drive alignments were incompatible,I   we saw the error counts climbing or saw the drives "maytagging"...  We MH   had the alignment tapes around for "target practice", too...)  (I feel4   older having just written this posting, too... :-)  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  # Date: Wed, 19 Apr 2000 17:50:22 GMTx2 From: kilgallen@eisner.decus.org (Larry Kilgallen) Subject: Re: Verify of Backups' Message-ID: <2000Apr19.135022.1@eisner>o  V In article <2000Apr19.114247.1@eisner>, kaplow_r@eisner.decus.org (Bob Kaplow) writes:  : > BTW, I don't think it's been pointed out that if you useJ > BACKUP/IMAGE/IGNORE=INTERLOCK/VERIFY it is very likely that /VERIFY willN > FAIL. Any file that is either open at backup time, or is changed between theK > backup and verify passes will trigger errors. That renders /VERIFY prettya
 > useless.  G No, I would say that is one of the most useful applications of /VERIFY.oG It makes it apparent that you do not have an accurate copy of your disko  at any particular point in time.   ------------------------------   Date: 19 Apr 2000 17:54:19 GMT* From: helbig@astro.rug.nl (Phillip Helbig) Subject: Re: Verify of Backups. Message-ID: <8dkrsb$53i$3@info.service.rug.nl>  V In article <2000Apr19.114247.1@eisner>, kaplow_r@eisner.decus.org (Bob Kaplow) writes:  : > BTW, I don't think it's been pointed out that if you useJ > BACKUP/IMAGE/IGNORE=INTERLOCK/VERIFY it is very likely that /VERIFY will > FAIL.    ????   SET NOON, ON ERROR,...  H > Any file that is either open at backup time, or is changed between theK > backup and verify passes will trigger errors. That renders /VERIFY pretty 
 > useless.   Not at all:   B $  SEARCH DISK$USER:[ADMIN]XXX.LOG "-E-","-W-"/OUT=SYS$SCRATCH:X.X9 $  MAIL/SUBJ="PROBLEM WITH BACKUP?" SYS$SCRATCH:X.X ADMINe  G This is the first mail I read every morning.  Usually, there are about iF 10--20 "problems".  These are almost always in things like WWW server E log files and MAIL directories, i.e. things which are very likely to fF change.  If anything else has a problem, it's worth checking out.  Of I course, if you're backing up disks with a lot of activity, this approach C won't work.I   ------------------------------   Date: 19 Apr 2000 17:59:17 GMT* From: helbig@astro.rug.nl (Phillip Helbig) Subject: Re: Verify of Backups. Message-ID: <8dks5l$53i$4@info.service.rug.nl>  F In article <2000Apr19.114534.1@eisner>, kaplow_r@eisner.decus.org (Bob Kaplow) writes:    > What is DDS?  G Digital Data Storage---DDS1, DDS2, DDS3 refer to different "levels" of 0F DAT tapes.  Time to post my old list of questions again, as they were  never ALL answered:i  I I'm looking for a good description of the interrelationships between the t% following things regarding DAT tapes:t      o  exact TLZ number  3    o  SCSI1, SCSI2, SCSI3, narrow, wide, fast, slows      o  DDS1, DDS2, DDS3 etc      o  /MEDIA_FORMAT=COMPACTION  "    o  length of the tape in meters  ;    o  capacity of the tape in GB and what this really means   *    o  compression in the tape drive itself  F    o  increase in size of a BACKUP saveset compared to the size of the       stuff on diskM      o  INITIALIZE/DENSITY  (    o  anything else which seems relevant   ------------------------------  # Date: Wed, 19 Apr 2000 18:20:47 GMTa* From: Gord Coulman <gcoulman@ccinet.ab.ca> Subject: Re: Verify of Backups, Message-ID: <38FDF212.71468EE3@ccinet.ab.ca>   Phillip Helbig wrote:1  ; > In article <38FDBC5D.5058B2D3@ccinet.ab.ca>, Gord Coulmans > <gcoulman@ccinet.ab.cawrites:e > E > > I have never seen verify catch any problems other than files thattD > > changed since the backup pass (not really a problem).  When tapeF > > problems have occurred (bad tapes, dirty heads, etc) backup alwaysK > > bagged an error during the backup pass and never even got to the verifyg	 > > pass.a >d > ??? SET NOON, ON ERROR,...  I Yes, I had NOON and error handling, but I refer to a problem with a TLZ07eM tape loader.  When this thing had problems, all the lights on the front wouldzM flash and the backup could not continue no matter what.  I changed the "guts"eG three times under warranty, used lots of cleaning tapes, then used less-M cleaning tapes (on the theory they were prematurely wearing the heads).  SameoH results.  When the warranty period was over it ran for about a year thenL failed again.  This time, one of my techs took it apart to see what he couldK find.  What he found was about a million teeny little parts that now residesH in a ziplock bag in the back of a storage room somewhere.  Too bad, thatB sucker was expensive.  I bought a DLT drive and never looked back.   Gord.h   ------------------------------    Date: 19 Apr 2000 11:19:03 -0700* From: dunnett@mala.bc.ca (Malcolm Dunnett) Subject: Re: Verify of Backups, Message-ID: <D2ssa5fQIQHW@malvm2.mala.bc.ca>  ( In article <2000Apr19.135022.1@eisner>, 7    kilgallen@eisner.decus.org (Larry Kilgallen) writes:i > ; >> BTW, I don't think it's been pointed out that if you useeK >> BACKUP/IMAGE/IGNORE=INTERLOCK/VERIFY it is very likely that /VERIFY will O >> FAIL. Any file that is either open at backup time, or is changed between thekL >> backup and verify passes will trigger errors. That renders /VERIFY pretty >> useless.  > I > No, I would say that is one of the most useful applications of /VERIFY.aI > It makes it apparent that you do not have an accurate copy of your diskn" > at any particular point in time.  C    Or, put another way, it lets you know exactly what is inaccurate-G about your backup - which allows you to make an informed decision about  how useful the backup is.B   ------------------------------  # Date: Wed, 19 Apr 2000 18:22:52 GMTA- From: "Richard D. Piccard" <piccard@ohio.edu>2 Subject: Re: Verify of Backups( Message-ID: <38FDF975.88D335DD@ohio.edu>  [ Yup.  In a one-drive shop, though, it can get pretty far out of alignment, being discoveredaX only when you need to restore an old backup and find that it is unreadable.  In our case[ this was a TS-11 at 1600 bpi.  (Of course, head alignment on helical-scan drives is a wholeo different ballgame.)  #                                 RDPe   Hoff Hoffman wrote:s  Z > In article <38FDEBFC.F32D34A2@ohio.edu>, "Richard D. Piccard" <piccard@ohio.edu> writes: > :Hoff Hoffman wrote:N > :>   That said, if I have unique or difficult-to-replace data, I usually use? > :>   BACKUP /VERIFY -- otherwise, why bother with the BACKUP?. > L > :There is one failure scenario that the /VERIFY won't detect:  mis-alignedJ > :heads.  In a previous position, we were passing the /VERIFY comparison,I > :but could not restore the tapes on any other drive because of the heade > :mis-alignment.i >bJ >   True, but how much of a problem that is depends on your willingness toF >   resort to hardware fixes, assuming that the hardware permits it... >gJ >   This alignment problem was something I saw fairly regularly many yearsJ >   back -- big, hot, power-hungry 9-track 1600/6250 vacuum-column magtapeJ >   (nagtape? :-) drives -- and just tweaked the heads on the other drivesJ >   the appropriate distance out of alignment.  (We didn't usually need toJ >   do this, as we usually picked up on any head alignment problems ratherH >   quickly given the frequency with which we swapped tapes around amongN >   the various drives...  Long before the drive alignments were incompatible,J >   we saw the error counts climbing or saw the drives "maytagging"...  WeJ >   had the alignment tapes around for "target practice", too...)  (I feel6 >   older having just written this posting, too... :-) >aP >  --------------------------- pure personal opinion ---------------------------N >    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   --B ==================================================================B Dick Piccard                           Academic Technology ManagerB piccard@ohio.edu                                 Computer ServicesB http://oak.cats.ohiou.edu/~piccard/                Ohio University   ------------------------------  % Date: Wed, 19 Apr 2000 20:19:21 +0100o8 From: John Macallister <J.Macallister1@physics.ox.ac.uk> Subject: RE: Verify of BackupsL Message-ID: <F02D5A46B8AED311BE4F0090279FA2401E831C@ppnt41.physics.ox.ac.uk>  H Using /VERIFY tells you nothing at all about whether a restore operationL will be successful. The examples given of accidental overwriting of tapes byI using /REWIND every time (!!) or by mounting the wrong tapes are only toorG real to anyone who has had to do regular backups, often under pressure.h  I The only sure way to test your backups is to make regular restore checks.(I This may come about in the normal course of events if you have to restoreaL some files for users from time to time or if you have to recover from a disk failure.  D To protect against accidental overwriting of tape savesets I use theK /TAPE_EXPIRATION and /LABELS=(list)/EXACT_ORDER qualifiers in BACKUP: I hadt# a few accidents before using those.n  J Performing simple checks on the backup jobs such as checking that the tapeH unit records some I/O operations and knowing the normal elapsed times ofL your backups are probably the easiest and most effective checks that you can6 make to ensure that your backups are running normally.   John    B Name: John B. Macallister  E-mail: j.macallister1@physics.ox.ac.ukH Post: Nuclear and Astrophysics Laboratory, Keble Road, Oxford OX1 3RH,UKA Phone: +44-1865-273388 (direct)  273333 (reception)  273418 (Fax)    ------------------------------  # Date: Wed, 19 Apr 2000 19:56:30 GMTd0 From: djlong@wild_wizards.net (David J. P. Long) Subject: Re: Verify of Backups+ Message-ID: <39000f20.40126259@news.mv.com>   % "L McCann" <lmccann@ddpwa.com> wrote:   G >We are running VMS Backup on Alpha servers and tz88 tape drives (hsz50 H >controller) and we do not use the /verify option on the backup command. > J >I'd like to get a feel for what other VMS shops do for their backups.  DoJ >you verify them or not?  I have found them to be reliable whenever I do a  - We did a 2-stage backup in our nightly cycle.   C First, we did disk-to-disk backups of all the database files.  This_F allowed us to get out of the backup section of our nightly cycle ASAP.  F Then, once the cycle was done, we would light off a tape backup of theC on-line backups.  With all that redundancy, we didn't need /verify.tA +----/|-------------------------------------+-------------------+rA |   | |  djlong@wildwizards.net              \                  |mA |  /  |    djlong@msn.com                     \                 |1A | (    )     http://www.wildwizards.net        \ ICQ# 8976662   |eA +--`--' ----------------------------------------+---------------+s   ------------------------------    Date: 19 Apr 2000 13:53:51 -0700* From: dunnett@mala.bc.ca (Malcolm Dunnett) Subject: RE: Verify of Backups, Message-ID: <DldZBxkVuL3O@malvm2.mala.bc.ca>  M In article <F02D5A46B8AED311BE4F0090279FA2401E831C@ppnt41.physics.ox.ac.uk>, D=    John Macallister <J.Macallister1@physics.ox.ac.uk> writes:   J > Using /VERIFY tells you nothing at all about whether a restore operationN > will be successful. The examples given of accidental overwriting of tapes byK > using /REWIND every time (!!) or by mounting the wrong tapes are only toohI > real to anyone who has had to do regular backups, often under pressure.o > @    Rather than use /VERIFY I do all the savesets first, then useH /COMPARE to check them all after all the sets are created. This not onlyG catches errors such as /REWIND overwriting savesets but it is generallyr" faster and saves wear on the tape.   ------------------------------   Date: 19 Apr 2000 20:58:06 GMT* From: helbig@astro.rug.nl (Phillip Helbig) Subject: RE: Verify of Backups. Message-ID: <8dl6ku$8lk$1@info.service.rug.nl>  H In article <DldZBxkVuL3O@malvm2.mala.bc.ca>, dunnett@mala.bc.ca (Malcolm Dunnett) writes:    B >    Rather than use /VERIFY I do all the savesets first, then useJ > /COMPARE to check them all after all the sets are created. This not onlyI > catches errors such as /REWIND overwriting savesets but it is generally $ > faster and saves wear on the tape.  E Why/how?  Just curious---it still has to read them again to compare; _/ isn't /VERIFY doing essentially the same thing?n   ------------------------------    Date: 19 Apr 2000 19:28:54 +0000, From: "Ewan Parker" <ewan@chimera.u-net.com> Subject: Re: Verify of Backups9 Message-ID: <687.144T2734T11686545ewan@chimera.u-net.com>e  = On 18-Apr-00 22:24:22, L McCann said about Verify of Backups:aH > We are running VMS Backup on Alpha servers and tz88 tape drives (hsz50I > controller) and we do not use the /verify option on the backup command.   K > I'd like to get a feel for what other VMS shops do for their backups.  Do K > you verify them or not?  I have found them to be reliable whenever I do aeA > restore but I know that it only takes 1 time to make you sorry.h  O If you are under time constraints, you could mount the tape after the backup isaO complete, and just restore the last file on that tape to a spare disk.  I would  recommend /verify though.    Regards, Ewan.y   --3 Public key:   http://www.chimera.u-net.com/ewan.asco   ------------------------------    Date: 19 Apr 2000 15:22:31 -0700* From: dunnett@mala.bc.ca (Malcolm Dunnett) Subject: RE: Verify of Backups, Message-ID: <Mpo5ZUXvCFvb@malvm2.mala.bc.ca>  / In article <8dl6ku$8lk$1@info.service.rug.nl>, d0     helbig@astro.rug.nl (Phillip Helbig) writes:  J > In article <DldZBxkVuL3O@malvm2.mala.bc.ca>, dunnett@mala.bc.ca (Malcolm > Dunnett) writes: e > C >>    Rather than use /VERIFY I do all the savesets first, then usebK >> /COMPARE to check them all after all the sets are created. This not onlyuJ >> catches errors such as /REWIND overwriting savesets but it is generally% >> faster and saves wear on the tape.i > G > Why/how?  Just curious---it still has to read them again to compare; t1 > isn't /VERIFY doing essentially the same thing?   ?    It depends on the tape drive and how BACKUP gets back to the_G start of the saveset. I used to use /VERIFY and I found that BACKUP was I rewinding the tape and then skipping forward to the start of the saveset,aD so every save set in front of the one you wanted to verify had to beI passed over along the way - slowing things down and creating extra passesrC over the tape. On some drives ( such as the old Exabyte 8200s ) the F "file skip" function got executed at read speed, so it could literally" take hours to start a verify pass.   ------------------------------  # Date: Thu, 20 Apr 2000 00:01:48 GMTs- From: "Dave Pampreen" <davepampreen@home.com>c Subject: Re: Verify of Backups; Message-ID: <MNrL4.34717$h01.283929@news1.rdc1.mi.home.com>s  G The quality of drives and tape units is not important...Tape media justeJ plain sucks.  We use both 4mm and DLT and have had times when either tapesG or drives would go bad.  The only thing that caught it was the /VERIFY.m  L We also take our backup tapes and restore the WHOLE system twice a year.  WeK use a disaster recovery site (Comdisco).  They have a duplicate enviromnent , for a pretty low cost for disaster testing!!  A I agree, you should use SET NOON then do the SEARCH backuplog.log K "-W-","-E-","-F-" then e-mail...I have a job that does this to ALL my batch I jobs every morning at 6:45.  it also includes other errors that have comelE up, like non-vms errors "Error during processing of inventory" etc...a   Dave    7 "Gord Coulman" <gcoulman@ccinet.ab.ca> wrote in messagee& news:38FDBC5D.5058B2D3@ccinet.ab.ca...K > I have never seen verify catch any problems other than files that changede sincelK > the backup pass (not really a problem).  When tape problems have occurrede (badJ > tapes, dirty heads, etc) backup always bagged an error during the backup passD > and never even got to the verify pass.  I still use verify anyway. > L > The worst backup problems I have seen are techs that don't properly rotate theeD > tapes or get too lazy to do the offsite rotation.  Improper backup commands areG > another big one.  I once wrote a DCL procedure that backed up severaln disks toB > the same tape (so far, so good), but used /rewind between disks, effectively.- > overwriting each saveset with the next one!" >g > Gord.c >  > Charlie Hammond wrote: >i. > > In article <8djvr0$6kn$1@news.ihug.co.nz>,0 > > "Ryan C. Price" <pricerc@ihug.co.nz> writes: > > ..F > > >you can't trust your backup unless you've tested it. The only way. > > >to test it is to scan the whole tape. ... > >-F > > I disagree.  The only way to test your backup is to restore it andJ > > check that the restored disk/files is/are correct.  Short of that, anyB > > number of things -- Backup utility, media reliability, storage procedures,e" > > etc. -- can and will go wrong. > >dI > > As for /VERIFY -- if push came to shove I would probably cover myselfE and/K > > advise  to use it.  However, this option exists because the tape drivespF > > typically used for backup a decade or two ago were not paricularly	 reliable.eJ > > They would sometimes report a write operation completed, but not writeL > > anything readable; /VERIFY caught this.  Today's backup devices are much? > > more robust.  The need for and value of /VERIFY is reduced.  > > = > > The need for and value of periodic test restores remains. < > > This also provides a test for your procedures to recover) > > any input lost since the last backup.k > >3 > > --K > >     Charlie Hammond -- Compaq Computer Corporation -- Pompano Beach  FLe USArH > >            (hammond@peek.ppb.dec.com -- remove "@not" when replying)B > >       All opinions expressed are my own and not necessarily my employer's.  >l   ------------------------------  % Date: Wed, 19 Apr 2000 20:21:28 -0400p" From: Dan Sugalski <dan@sidhe.org> Subject: Re: Verify of Backups8 Message-ID: <4.3.1.0.20000419202040.01d26570@24.8.96.48>  / At 12:01 AM 4/20/00 +0000, Dave Pampreen wrote:MH >The quality of drives and tape units is not important...Tape media just
 >plain sucks.C  H Well, sure. That's a given, isn't it? *All* hardware sucks. It's just a  matter of how badly...   					Dan  I --------------------------------------"it's like this"-------------------i2 Dan Sugalski                          even samurai? dan@sidhe.org                         have teddy bears and eveno;                                       teddy bears get drunk    ------------------------------  # Date: Thu, 20 Apr 2000 02:18:34 GMTr$ From: Ed Wilts <ewilts@mediaone.net># Subject: Re: Web server for VAX/VMS:, Message-ID: <38FE68FB.C375B60B@mediaone.net>   Arne Vajhj wrote: > A > Only the ones from Compaq seems to have problems with VMS VAX !b  D Let's be fair here.  Compaq wants Java compatibility, and that's notD technically feasible on the Vax architecture.  I'm sure Compaq couldG release a Java-less web server if they wanted to, but why should they? oD They won't make any money off it, either in support or sales.  TheirG push is on the Alpha, and I certainly won't fault them for that at this  stage of the game.   	.../Edd   -- , Ed Wilts Mounds View, MN, USA mailto:ewilts@mediaone.net   ------------------------------  % Date: Wed, 19 Apr 2000 20:25:12 -0600a% From: Dan O'Reilly <dano@process.com> # Subject: Re: Web server for VAX/VMStB Message-ID: <4.2.0.58.20000419202358.00a84700@pop.clsp.uswest.net>  % At 08:18 PM 4/19/00 , Ed Wilts wrote:  >Arne Vajh=F8j wrote:  > >nC > > Only the ones from Compaq seems to have problems with VMS VAX !  >eE >Let's be fair here.  Compaq wants Java compatibility, and that's notbE >technically feasible on the Vax architecture.  I'm sure Compaq couldiG >release a Java-less web server if they wanted to, but why should they?cE >They won't make any money off it, either in support or sales.  Their H >push is on the Alpha, and I certainly won't fault them for that at this >stage of the game.p  H Why is it "not technically feasible"?  I'm not sure that's the reason so7 much as Compaq would like to kill the VAX architecture.u  J Cynical?  Yeah, but then again, what's the Q done to convince us that they really give a darn?.     ------I +-------------------------------+---------------------------------------+aI | Dan O'Reilly                  |                                       |sI | Principal Engineer            |  "Time flies like an arrow.  Fruit    | I | Process Software Corporation  |   flies like a banana."               |lI | http://www.process.com        |                    -- Groucho Marx    |oI +-------------------------------+---------------------------------------+    ------------------------------  % Date: Wed, 19 Apr 2000 12:44:59 -0600d% From: Dan O'Reilly <dano@process.com>i Subject: What's this option?B Message-ID: <4.2.0.58.20000419124430.00a66dd0@pop.clsp.uswest.net>  % What is a PBXGA-CA/CN?  Anybody know?b   ------I +-------------------------------+---------------------------------------+ I | Dan O'Reilly                  |                                       |tI | Principal Engineer            |  "Time flies like an arrow.  Fruit    |cI | Process Software Corporation  |   flies like a banana."               |.I | http://www.process.com        |                    -- Groucho Marx    | I +-------------------------------+---------------------------------------+k   ------------------------------  % Date: Wed, 19 Apr 2000 15:00:15 -0400e( From: Kurt Schulte <kschulte@GLComp.Com>  Subject: RE: What's this option?G Message-ID: <2948CB0D8867D311AE340000F849735E470A65@glcfs01.glcomp.com>i  ) TGA 24 Plane, w/ Z Buffer. 4M-Pixels, PCII   -----Original Message-----  % What is a PBXGA-CA/CN?  Anybody know?o   ------I +-------------------------------+---------------------------------------+nI | Dan O'Reilly                  |                                       |eI | Principal Engineer            |  "Time flies like an arrow.  Fruit    |-I | Process Software Corporation  |   flies like a banana."               |iI | http://www.process.com        |                    -- Groucho Marx    |FI +-------------------------------+---------------------------------------+r   ------------------------------  % Date: Wed, 19 Apr 2000 13:03:12 -0600 % From: Dan O'Reilly <dano@process.com>d  Subject: RE: What's this option?B Message-ID: <4.2.0.58.20000419130255.00a63f00@pop.clsp.uswest.net>   Supported by VMS?    Thanks for the info!  ) At 01:00 PM 4/19/00 , Kurt Schulte wrote:o* >TGA 24 Plane, w/ Z Buffer. 4M-Pixels, PCI >u >-----Original Message-----a >o& >What is a PBXGA-CA/CN?  Anybody know? >  >------ J >+-------------------------------+---------------------------------------+J >| Dan O'Reilly                  |                                       |J >| Principal Engineer            |  "Time flies like an arrow.  Fruit    |J >| Process Software Corporation  |   flies like a banana."               |J >| http://www.process.com        |                    -- Groucho Marx    |J >+-------------------------------+---------------------------------------+   ------I +-------------------------------+---------------------------------------+gI | Dan O'Reilly                  |                                       |sI | Principal Engineer            |  "Time flies like an arrow.  Fruit    |eI | Process Software Corporation  |   flies like a banana."               |sI | http://www.process.com        |                    -- Groucho Marx    |oI +-------------------------------+---------------------------------------+r   ------------------------------  % Date: Wed, 19 Apr 2000 15:20:12 -0400 ( From: Kurt Schulte <kschulte@GLComp.Com>  Subject: RE: What's this option?G Message-ID: <2948CB0D8867D311AE340000F849735E470A69@glcfs01.glcomp.com>.  J These were standard fare in AS400's and the like.  -CA's are good for VMS, Unix.  -CN's are NT.   -----Original Message-----, From: Dan O'Reilly [mailto:dano@process.com]' Sent: Wednesday, April 19, 2000 3:03 PM- To: Kurt Schulte Cc: Info-VAX@Mvb.Saic.Comn  Subject: RE: What's this option?       Supported by VMS?=   Thanks for the info!  ) At 01:00 PM 4/19/00 , Kurt Schulte wrote: * >TGA 24 Plane, w/ Z Buffer. 4M-Pixels, PCI >g >-----Original Message-----  > & >What is a PBXGA-CA/CN?  Anybody know? >/ >------.J >+-------------------------------+---------------------------------------+J >| Dan O'Reilly                  |                                       |J >| Principal Engineer            |  "Time flies like an arrow.  Fruit    |J >| Process Software Corporation  |   flies like a banana."               |J >| http://www.process.com        |                    -- Groucho Marx    |J >+-------------------------------+---------------------------------------+   ------I +-------------------------------+---------------------------------------+wI | Dan O'Reilly                  |                                       |aI | Principal Engineer            |  "Time flies like an arrow.  Fruit    |bI | Process Software Corporation  |   flies like a banana."               |uI | http://www.process.com        |                    -- Groucho Marx    |oI +-------------------------------+---------------------------------------+X   ------------------------------   Date: 19 Apr 2000 19:14:36 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)  Subject: RE: What's this option?6 Message-ID: <8dl0is$m9b$1@mailint03.im.hou.compaq.com>  j In article <4.2.0.58.20000419130255.00a63f00@pop.clsp.uswest.net>, Dan O'Reilly <dano@process.com> writes: :Supported by VMS?  H   Yes, though it requires the Open3D kit -- check the Open3D kit versionF   and the OpenVMS version to ensure compatibility and support for this   particular widget.  * :At 01:00 PM 4/19/00 , Kurt Schulte wrote:+ :>TGA 24 Plane, w/ Z Buffer. 4M-Pixels, PCI2  >   The PBXGA-C* series is also known as the ZLXp-E3 controller.   :>What is a PBXGA-CA/CN?  F   The ZLXp-E1/E2/E3 PCI Graphics Options Owner's Guide is EK-T2424-OG.  F   Lemme know what resolution(s) you want to run, and I'll send you theC   switchpack settings for it.  There are sixteen settings options, 8E   ranging from 1280 x 1024 x 72 Hz down to 640 x 480 x 60 Hz or so... F   (I'd normally post the whole list, but I'd first have to type it in ;   from the switchpack settings diagram I have access to...)a  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 19 Apr 2000 13:29:50 -0600A% From: Dan O'Reilly <dano@process.com>   Subject: RE: What's this option?B Message-ID: <4.2.0.58.20000419132829.00a6f3a0@pop.clsp.uswest.net>  ) At 01:14 PM 4/19/00 , Hoff Hoffman wrote:-  H >In article <4.2.0.58.20000419130255.00a63f00@pop.clsp.uswest.net>, Dan $ >O'Reilly <dano@process.com> writes: >:Supported by VMS?  > J >   Yes, though it requires the Open3D kit -- check the Open3D kit versionH >   and the OpenVMS version to ensure compatibility and support for this >   particular widget. >-+ >:At 01:00 PM 4/19/00 , Kurt Schulte wrote:-, >:>TGA 24 Plane, w/ Z Buffer. 4M-Pixels, PCI > @ >   The PBXGA-C* series is also known as the ZLXp-E3 controller. >. >:>What is a PBXGA-CA/CN?A >8H >   The ZLXp-E1/E2/E3 PCI Graphics Options Owner's Guide is EK-T2424-OG. > H >   Lemme know what resolution(s) you want to run, and I'll send you theD >   switchpack settings for it.  There are sixteen settings options,G >   ranging from 1280 x 1024 x 72 Hz down to 640 x 480 x 60 Hz or so... G >   (I'd normally post the whole list, but I'd first have to type it inn= >   from the switchpack settings diagram I have access to...)   J 1280x1024 is fine.  I have a 20" Sony monitor, so I think that'll be fine.J Many thanks!  Is that manual by any chance online, or could you fax me the appropriate pages?     ------I +-------------------------------+---------------------------------------+-I | Dan O'Reilly                  |                                       |tI | Principal Engineer            |  "Time flies like an arrow.  Fruit    |lI | Process Software Corporation  |   flies like a banana."               |aI | http://www.process.com        |                    -- Groucho Marx    |/I +-------------------------------+---------------------------------------+a   ------------------------------  # Date: Wed, 19 Apr 2000 19:27:27 GMTE= From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-)v  Subject: RE: What's this option?0 Message-ID: <009E8D9C.3B74D1CD@SendSpamHere.ORG>  k In article <8dl0is$m9b$1@mailint03.im.hou.compaq.com>, hoffman@xdelta.zko.dec.nospam (Hoff Hoffman) writes:2 > k >In article <4.2.0.58.20000419130255.00a63f00@pop.clsp.uswest.net>, Dan O'Reilly <dano@process.com> writes:7 >:Supported by VMS?n >iI >  Yes, though it requires the Open3D kit -- check the Open3D kit versioncG >  and the OpenVMS version to ensure compatibility and support for this  >  particular widget.n  E And it has a bug I've reported and which has *never* been fixed alongtE with the screwed up fonts after running NetScape. :(  None out of two  ain't bad, eh?    G When you open a document in BookReader and use the scroll bar, the texttF is 'smeared out' and unreadable.  You can get by if you use the [^][v]G controls to drag the scroll bar but you can't directly move it with theu mouse or text is smeared.    --N VAXman- OpenVMS APE certification number: AAA-0001           VAXman@TMESIS.COM  L GNU Freeware -- What does the GNU *really* stand for?  Garbage!  Not Usable!   ------------------------------   Date: 19 Apr 2000 19:49:55 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)  Subject: RE: What's this option?6 Message-ID: <8dl2l3$n0b$1@mailint03.im.hou.compaq.com>  j In article <4.2.0.58.20000419132829.00a6f3a0@pop.clsp.uswest.net>, Dan O'Reilly <dano@process.com> writes:   :1280x1024 is fine.M  A   All switches downward (with the module flat, and the component a$   side up) is 1280 x 1024 x 72 Hz...  @   SW1 (nearest of the four switches to the stereo port) up, all &   others down, is 1280 x 1024 x 66 Hz.  C   There are also other 1280 x 1024 settings (60 Hz and 75 Hz), but h=   the above two settings are probably the more common ones...h  ;   You have to power-cycle the box to change the settings...a  = :Is that manual by any chance online, or could you fax me the  :appropriate pages?f  G   Not easily.  Again, if I could have easily posted the list (or faxed  F   it) (without having to retype the whole document, or spend some timeG   to dig around for a copy of the switchpack settings in a more easily e4   "transcribable" format), I would have...  Sorry...  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Wed, 19 Apr 2000 15:56:12 -0600,% From: Dan O'Reilly <dano@process.com>   Subject: RE: What's this option?B Message-ID: <4.2.0.58.20000419155600.00a66700@pop.clsp.uswest.net>   Thanks!1  ) At 01:49 PM 4/19/00 , Hoff Hoffman wrote:a  H >In article <4.2.0.58.20000419132829.00a6f3a0@pop.clsp.uswest.net>, Dan $ >O'Reilly <dano@process.com> writes: >p >:1280x1024 is fine. >dB >   All switches downward (with the module flat, and the component& >   side up) is 1280 x 1024 x 72 Hz... >mA >   SW1 (nearest of the four switches to the stereo port) up, allh( >   others down, is 1280 x 1024 x 66 Hz. >wD >   There are also other 1280 x 1024 settings (60 Hz and 75 Hz), but? >   the above two settings are probably the more common ones...a >a= >   You have to power-cycle the box to change the settings...5 >i> >:Is that manual by any chance online, or could you fax me the >:appropriate pages? >sH >   Not easily.  Again, if I could have easily posted the list (or faxedH >   it) (without having to retype the whole document, or spend some timeH >   to dig around for a copy of the switchpack settings in a more easily6 >   "transcribable" format), I would have...  Sorry... >l5 >  --------------------------- pure personal opinion c > ---------------------------nN >    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------I +-------------------------------+---------------------------------------+oI | Dan O'Reilly                  |                                       |aI | Principal Engineer            |  "Time flies like an arrow.  Fruit    |aI | Process Software Corporation  |   flies like a banana."               |wI | http://www.process.com        |                    -- Groucho Marx    |oI +-------------------------------+---------------------------------------+0   ------------------------------  % Date: Wed, 19 Apr 2000 16:13:01 -0600 % From: Dan O'Reilly <dano@process.com>p  Subject: RE: What's this option?B Message-ID: <4.2.0.58.20000419161158.00a87140@pop.clsp.uswest.net>  ) At 01:14 PM 4/19/00 , Hoff Hoffman wrote:t  H >In article <4.2.0.58.20000419130255.00a63f00@pop.clsp.uswest.net>, Dan $ >O'Reilly <dano@process.com> writes: >:Supported by VMS?e > J >   Yes, though it requires the Open3D kit -- check the Open3D kit versionH >   and the OpenVMS version to ensure compatibility and support for this >   particular widget.    E Is this kit required for all uses of the option, or just for specificlH things (i.e., will it operate as a "normal" DCE environment without it)?  , Also, is a separate Open3D license required?     ------I +-------------------------------+---------------------------------------+eI | Dan O'Reilly                  |                                       |cI | Principal Engineer            |  "Time flies like an arrow.  Fruit    | I | Process Software Corporation  |   flies like a banana."               | I | http://www.process.com        |                    -- Groucho Marx    |aI +-------------------------------+---------------------------------------+@   ------------------------------   Date: 19 Apr 2000 19:36:41 GMT* From: bleau@umdsp.umd.edu (Lawrence Bleau)> Subject: Why does Pathworks(Mac) change mod date of .dir file?) Message-ID: <8dl1s9$6p5$1@hecate.umd.edu>a  N I have a puzzle here with Pathworks(Mac) V1.3A on OpenVMS AXP 7.1-2.  It seemsL that every time a user connects to a Pathworks(Mac) volume, the modificationL date of the .DIR file at the root of that volume is updated and the revisionK count is incremented.  Likewise, when a folder (directory) is opened withinsM that volume, the corresponding .DIR file also has its mod date updated.  This-B happens even when *nothing* is changed in that folder/directory.    L This is additionally puzzling because typically the mod date for a .DIR fileO does *not* change on a regular basis; that is, file creation and deletion go onnN all the time without the .DIR file's mod date changing.  Only when one changesK its protection, adds/removes an ACL, renames the directory, or some similars# operation does the mod date change.e  N I did a DIR/FUL on several of my user directories to confirm this.  Those thatK do not have a Pathworks(Mac) volume have revision counts in the range 1-10;hD those with Pathworks(Mac) volumes have revision counts of over 2000.  M Ordinarily I wouldn't be concerned about this, except for the behavior of VMSoK BACKUP.  When BACKUP (under V7.1-2 at least) detects a .DIR file with a modiE date later than the last backup date, it backs up *all files* in *allgL subdirectories* of that directory.  Since my users have their Pathworks(Mac)O volumes mapped to their login directories, this in effect does a full backup of5D those users' directories every day (I do daily incremental backups).  M I detected this when the system started asking for a 2nd, then a 3rd tape for.J backups.  Usually this system uses three 4mm tapes per month: two for fullL backup and one for all the daily incremental backups.  When it asked for theN 4th tape (2nd incremental tape) I got suspicious and checked the backup logs. N Yep; there are 3313 files in the user's directory tree, and BACKUP copied 3313 files to tape.  K Now to complicate life: This didn't always happen.  Although I noticed thiseO behavior only this month, I checked my older logs, and it's been going on sincetP March 11.  Before that only 20-30 of this user's files were backed up each day.   I I checked with the user, and he's not changed the typical operations he's J done.  The system was rebooted (and Pathworks(Mac) started) on March 1 andL again on Mar 18, so a system reboot didn't cause this.  I've not changed the# system parameters in that period.  -  J HOWEVER, I did add a Pathworks(Mac) volume for a new user on Mar 10.  ThisE very suspiciously coincides with the first date of the strange BACKUPrK behavior.  I checked the MSAF$SERVER.LOG files, and nothing is reported for H that day; the next entry is for Mar 15, when the new user first tries toG connect.  The odd BACKUP behavior had already begun, though, so that in- itself is not the reason.-  F So, what is going on here?  How do I stop it from happening?  How do I; change Pathworks(Mac)'s behavior back to "normal"?  Thanks.?   Lawrence Bleau University of Maryland" Physics Dept., Space Physics Group 301-405-6223 bleau@umdsp.umd.edu    ------------------------------   End of INFO-VAX 2000.220 ************************