1 INFO-VAX	Mon, 22 Sep 2003	Volume 2003 : Issue 525       Contents:0 Re: A flood of spams - another virus on the way?0 Re: A flood of spams - another virus on the way?0 Re: A flood of spams - another virus on the way?0 Re: A flood of spams - another virus on the way?0 Re: A flood of spams - another virus on the way?0 Re: A flood of spams - another virus on the way?0 Re: A flood of spams - another virus on the way?$ Re: Advanced Server software support$ Re: Advanced Server software support$ Re: Advanced Server software support2 Announcing Spring 2003 VMS Sig Tape (CD, actually)" automatically mounting shadow sets  Re: Cleaning up license database< RE: Fork() and Unix compatibility work timetable and scope ? How to uninstall?  Re: merging queue databases  Re: merging queue databases  Re: merging queue databases  Re: merging queue databases  Re: merging queue databases  Re: merging queue databases  Re: merging queue databases  Re: merging queue databases  Re: merging queue databases  Re: merging queue databases & Multichrom support on OpenVMS, anyone?0 Re: OpenVMS hobbyyest - free to join or not ????3 Preserving decwindows session through disconnection  Re: Prove this corrective patch  Re: Prove this corrective patch 6 Re: strangeness with temporary text files used by MAIL+ Re: TCP/IP Services for VMS - sloppy output  TCPIP  Reject processing  Re: The vultures are circulating  RE: The vultures are circulating  F ----------------------------------------------------------------------  % Date: Sun, 21 Sep 2003 10:52:20 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> 9 Subject: Re: A flood of spams - another virus on the way? ' Message-ID: <3F6DC933.C3A4B9F8@fsi.net>    Wayne Sewell wrote:  >  > >>> > >> Install MX and use that instead of the SMTP in the stack. > >> > > N > >I was looking at Hunter's site http://www.madgoat.com/mx042.html.  Does theW > >kit on the freeware 4.0 have those patches included or do they need to be installed?  > >  > J > I think the patches refer to security issues rather than spam rejection.  G Seems to me the trouble with this round of worm-mail is that the sender F is different on the majority of instances. The Sender name and subjectC line seem to be chosen at random from a fixed set of choices. Seems C you'd need to be able to filter on sender name and/or subject line.   D My inbox at Eartlink has been filling twice or more daily since thisH began. Earthlink web-mail doesn't offer a choice to block by sender nameH or subject, only by sender address. Oh, well... My other address has not gotten "loose in the wild" yet.    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  # Date: Sun, 21 Sep 2003 17:48:24 GMT & From: Lee Mah <lytmah@telusplanet.net>9 Subject: Re: A flood of spams - another virus on the way? : Message-ID: <Ivlbb.12004$Vr3.308965@news1.telusplanet.net>  G I've never received such a deluge of crud like this in the past.  It's  
 taken someK work, but my major email filter account is now working properly and junking K the MS 'updates' and the few spam.  However, I still see the junk coming in D before it gets shunted to the garbage bin.  The torrent of MS email 	 makes one I appreciate real spam.  If it's like this now, what's it going to be like   MondayI and beyond.  How much worse will it get before the ISP's take some of the % actions as recommended in this group?   J I just complained to my ISP and suggested cutting off email at the origin.   --   Lee    lytmah@telusplanet.net   ------------------------------  % Date: Sun, 21 Sep 2003 16:03:55 -0500 ( From: Wayne Sewell <wayne@tachysoft.com>9 Subject: Re: A flood of spams - another virus on the way? / Message-ID: <00A263E2.6C4FC243.7@tachysoft.com>    > H >Seems to me the trouble with this round of worm-mail is that the senderG >is different on the majority of instances. The Sender name and subject D >line seem to be chosen at random from a fixed set of choices. SeemsD >you'd need to be able to filter on sender name and/or subject line. > E >My inbox at Eartlink has been filling twice or more daily since this I >began. Earthlink web-mail doesn't offer a choice to block by sender name I >or subject, only by sender address. Oh, well... My other address has not   >gotten "loose in the wild" yet. >     M Fortunately, MX allows you to filter on just about anything in the header.  I E have based rejection rules on most of them at some time or another.     G Sometimes the spammers help you out by including headers that could not  possibly occur in normal mail.  
 For instance:    Rule ID: 205$     Header: X-Mailer: *Direct Email*M     Added: 13-DEC-2001 09:08:37.68, Ref count: 1, Last ref date: 24-OCT-2002    
 Rule ID: 1242      Header: X-Mailer: The Bat!* M     Added: 21-APR-2003 20:41:28.04, Ref count: 10, Last ref date: 14-SEP-2003     K One of my favorite moron groups includes those spammers who forge Received: M headers which appear to come from your own address block.  I guess they think M that makes the mail look legitimate or something.  But that makes their shit  N *very* easy to recognize and filter, because the legitimate headers always use3 the domain name of the machine, not the ip address.     
 Rule ID: 1542 (     Header: Received: from xxx.yyy.zzz.*N     Added: 24-JUL-2003 10:37:13.82, Ref count: 105, Last ref date: 20-SEP-2003 ---   
 Rule ID: 1256      Header: * by xxx.yyy.zzz.*M     Added: 28-APR-2003 10:21:22.97, Ref count: 72, Last ref date: 24-JUL-2003     > Where xxx.yyy.zzz.* is the subnet in which my machines reside.    J A variant is where the message appears to come from your own domain.  OnlyH thing is, the legitimate server addresses are always NODE.tachysoft.com.   Rule ID: 778*     Header: Received: from tachysoft.com *N     Added:  8-NOV-2002 11:50:15.83, Ref count: 11, Last ref date: 12-SEP-2003         K A lot of spam appears to come from a foreign country or actually is relayed L though one.  Since I receive very little mail from other countries, I rejectN most countries across the board, especially middle eastern, east european, andH south american.  This includes the "from" and "by" fields of "received:"* headers, which filters out the relays too.   Rule ID: 31      Sender: *@*.CN+     Message:    Everything from .cn is spam O     Added: 28-JAN-2001 19:47:31.59, Ref count: 275, Last ref date:  7-SEP-2003   ---    Rule ID: 738!     Header: Received: from *.pt * M     Added: 16-OCT-2002 09:11:21.99, Ref count: 4, Last ref date: 19-SEP-2003     ---  Rule ID: 41 &     Header: Received: from * by *.tw *M     Added: 25-MAY-2001 19:34:05.72, Ref count: 7, Last ref date: 14-DEC-2002      O =============================================================================== N Wayne Sewell, Tachyon Software Consulting  (281)812-0738   wayne@tachysoft.com; http://www.tachysoft.com/www/tachyon.html and wayne.html    O =============================================================================== N Butler:"Gentlemen!"  Curly(as he and other Stooges look around):"Who came in?"   ------------------------------  % Date: Sun, 21 Sep 2003 22:31:47 +0200  From: Dirk Munk <munk@home.nl>9 Subject: Re: A flood of spams - another virus on the way? 2 Message-ID: <bkl245$3cr$1@news2.tilbu1.nb.home.nl>   Paul Sture wrote: 8 > More spams. Is this another virus / worm on the loose? > Q > Since 13:44 CET yesterday I have received some 114 spam messages (oops, another % > one just came in) in this  account.  > H Since Friday I must have received close to 1000 of these stupid e-mails.  Q Now what if we all send Bill Gates a thank-you e-mail message for disrupting the  P world economy for the 20th (?) 50th (?) time in succession over the last years, O and sign it with some Arab sounding name, with the addition "Al-Qaida fighter".   Q This may just trigger George W. to bomb the whole M$ headquarters to smithereens  * and redeem us from the Windooz pestilence.   ------------------------------  % Date: Sun, 21 Sep 2003 16:13:20 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> 9 Subject: Re: A flood of spams - another virus on the way? ' Message-ID: <3F6E1470.8AEF95D6@fsi.net>    Dirk Munk wrote: >  > Paul Sture wrote: : > > More spams. Is this another virus / worm on the loose? > > S > > Since 13:44 CET yesterday I have received some 114 spam messages (oops, another ' > > one just came in) in this  account.  > > J > Since Friday I must have received close to 1000 of these stupid e-mails. > R > Now what if we all send Bill Gates a thank-you e-mail message for disrupting theQ > world economy for the 20th (?) 50th (?) time in succession over the last years, Q > and sign it with some Arab sounding name, with the addition "Al-Qaida fighter".  > R > This may just trigger George W. to bomb the whole M$ headquarters to smithereens, > and redeem us from the Windooz pestilence.   ...and lead us not   into Gatesdom, but deliver us from  Windows.   Amen.    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------    Date: 21 Sep 2003 16:53:16 -0700- From: critters+google@vcn.bc.ca (Google Andy) 9 Subject: Re: A flood of spams - another virus on the way? = Message-ID: <82459d76.0309211553.5c5f356f@posting.google.com>   Z p_sture@elias.decus.ch (Paul Sture) wrote in message news:<dZLFASE9Oxar@elias.decus.ch>...8 > More spams. Is this another virus / worm on the loose? >   5 Yes!!! These are viruses.  You can get more info here D http://us.mcafee.com/virusInfo/default.asp?id=helpCenter&hcName=swen   Andy   ------------------------------  % Date: Mon, 22 Sep 2003 06:11:07 +0100 I From: "Dr. David Kirkby" <see-my-signiture-for-email-address@nowhere.com> 9 Subject: Re: A flood of spams - another virus on the way? + Message-ID: <3F6E846B.A5C9968C@nowhere.com>    Larry Kilgallen wrote: > W > In article <dZLFASE9Oxar@elias.decus.ch>, p_sture@elias.decus.ch (Paul Sture) writes: : > > More spams. Is this another virus / worm on the loose? > > S > > Since 13:44 CET yesterday I have received some 114 spam messages (oops, another ' > > one just came in) in this  account.  > 7 > Sorry, those executables are not compatible with VMS.  > J > You better hurry up and buy a Microsoft machine, because from the volumeK > of those messages it would appear that enormous numbers of the _existing_ I > Microsoft users are already ahead of you, getting the benefits ahead of  > you.  D Well I don't use VMS, but are contemplating it. But neither do I use@ Microsoft's so-called operating systems. However, this is reallyH screwing up my Solaris machine. I only have a 56 k modem and despite theA fact the Sun runs 24/7, the modem is unable to download the stuff H sufficiently fast to stop my mailbox filling up. I need to find a way toG delete the files on the server without downloading them. Spam filtering H is effective at dumping them to /dev/null (UNIX language), but it is not preventing me hell.   C However, I am glad I'm not the only one. I thought this was quite a 6 personal thing aimed at me, but appears not to be so.    --  A "The day Microsoft makes something that doesn't suck is probably  > the day they start making vacuum cleaners." -Ernst Jan Plugge.   Dr. David Kirkby,  Senior Research Fellow,  Department of Medical Physics, University College London," 11-20 Capper St, London, WC1E 6JA., Website: http://www.medphys.ucl.ac.uk/~davek- Author of 'atlc' http://atlc.sourceforge.net/    ------------------------------  % Date: Sun, 21 Sep 2003 10:53:52 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> - Subject: Re: Advanced Server software support ' Message-ID: <3F6DC990.1F6FFADE@fsi.net>    Tom Simpson wrote: > [snip]F > This is really crappy contract support!  Thank goodness our customerL > engineers are a whole LOT more dependable than the folks that are supposed > to be supporting them...  H I forwarded this to Mark and Rich at hp. Richard forwarded it internallyE and Cc'd me on the forward. Hope that brings some results. Good luck.    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  # Date: Sun, 21 Sep 2003 16:12:33 GMT 8 From: "Tom Simpson" <thomas.simpson1@nospam.comcast.net>- Subject: Re: Advanced Server software support . Message-ID: <R5kbb.399812$cF.122922@rwcrnsc53>  B I'm in Jacksonville FL.  The contract office is in Alpharetta, GA.   Thank you, David and Kerry.    Regards, Tom   < "David J. Dachtera" <djesys.nospam@fsi.net> wrote in message! news:3F6DC990.1F6FFADE@fsi.net...  > Tom Simpson wrote:
 > > [snip]H > > This is really crappy contract support!  Thank goodness our customerE > > engineers are a whole LOT more dependable than the folks that are  supposed > > to be supporting them... > J > I forwarded this to Mark and Rich at hp. Richard forwarded it internallyG > and Cc'd me on the forward. Hope that brings some results. Good luck.  >  > --   > David J. Dachtera  > dba DJE Systems  > http://www.djesys.com/ > * > Unofficial Affordable OpenVMS Home Page:! > http://www.djesys.com/vms/soho/    ------------------------------  % Date: Sun, 21 Sep 2003 11:41:29 -0700 * From: "Jack Peacock" <peacock@simconv.com>- Subject: Re: Advanced Server software support 2 Message-ID: <VladndFQDstHbfCiXTWJhQ@mpowercom.net>  C "Tom Simpson" <thomas.simpson1@nospam.comcast.net> wrote in message ) news:Bz%ab.392648$Oz4.178583@rwcrnsc54... F > This is really crappy contract support!  Thank goodness our customerL > engineers are a whole LOT more dependable than the folks that are supposed > to be supporting them... > L Perhaps you could call India directly and bypass the branch offices in NorthC America.  Do you have someone on your staff that speaks Hindustani?     Jack Peacock    ------------------------------  % Date: Sun, 21 Sep 2003 17:56:28 -0400 . From: Glenn Everhart <Everhart-nospam@gce.com>; Subject: Announcing Spring 2003 VMS Sig Tape (CD, actually) + Message-ID: <bkl84l$3si$1@bob.news.rcn.net>   D The Spring 2003 VMS Sig tapes have been closed and the first several. have been sent out to net distribution points.  A While Dar Schumann (darnkatt@feist.com) has been the nominal tree F operator for some time I am not sure his address is correct any more, J though the materials I send him are apparently delivered and never bounce.  = If someone wishes access to the material, especially for LUG  J distributions, please let me know where to send copies (Everhart@gce.com).  F Copies are still owed to some other chapters and to Encompasserve (so < those folks should be a bit patient; you will get CDs soon.)  H Thanks to all those who have developed new material, especially new VMS ( material, over the past half year or so.  I Dar's mail address (Farm Credit Bank of Wichita, 245 N. Waco, Wichita KS  0 67201) can be tried however for access to tapes.  @ The material is distributed mostly on a single CD in VMS format, which is rather full.    Glenn C. Everhart  Everhart@gce.com  $ The contents abstract is as follows:   enjoy...    ' Spring 2003 VMS/L&T SIG Tape Collection ' ------ ---- --- --- --- ---- ----------   * This material is all freely distributable.    A Once again the VMS and L&T SIGs present their semiannual software G collection, this one for Spring 2002. The following are brief abstracts 
 of the areas.   A (Note: Since the distribution is getting pretty varied, the split D   between "VMS" and "L&T" parts is not done this time. It was always   artificial.)  F The L&T and VMS Systems SIGs are again producing a combined collection6 for Spring 2002 to reduce duplication between the two.  = The tape is in VMS Backup format in three areas: one each for > [03AVMSLT...], [VMSLT03A...], and [DECUSLIB...].  The first isH the index saveset (03AVMSLT), comprising around 17000 blocks.  The otherF is the DECUS library catalog in machine readable format (DECUSLIB) andE the VAX/L&T collection for Spring 2002 [VMSLT03A...] comprising about  1270000 blocks.    INDEX Directory ([03AVMSLT]) ----- ---------  ---------- C          Indices and descriptions of what is on the tapes. [.tools] C          subdirectory contains tools needed to decompress or unpack           some packed items.   ) DECUS LIBRARY CATALOG ONLINE ([DECUSLIB]) ( ----- ------- ------- ------  ----------C          Current machine readable catalog of DECUS library contents >          as straight text. The BROWSER fulltext retrieval tool?          is included. Also the DECUS Catalog Information System           tool is included.  ? The following are contents of the respective directories in the    [VMSLT03A...] tree.   B [GCE]           IDEZR - read/write disks that have blocksize otherD                  than 512 bytes or that normal DKdriver can't handleE [GNU]           New Gnu software including Gcc 3.3.1, Glibc, Gnucash, %                  Gnumeric, Make, etc. D [IQRDEV]        Message driven queue system that does reliable interJ                  process messaging / scheduling something like DECMessageQ                  but with src.- [MOREAU]        Timidity music editor patches H [NET]           Bochs pc emulator, SIMH cpu emulators including for Vax,I                  Gnunet p2p, Xpaint, Wine windows emulator, Blassic Basic +                  interpreter, OpenSSL, more G [OPENOFFICE]    Source for OpenOffice 1.1 RC3 complete. MS Office clone D                  in source. Works on unix, windows, linux. Some work9                  needed for getting it running under VMS. D [SAMBA]         VMS Version 2.2.8a, serve disks to Windows shares orE                  printers to windows, access Windows shares from VMS. / [SEC]           A few docs about security holes F [TK]            CLEAR_ERRORS.ZIP Clear device error count on alpha VMSG                  DCL-EXPAND.ZIP DCL template expansion utility GRAB.ZIP J                  VMS search util with wildcards, many extensions IDEZR.ZIPC                  Read non-512-byte CD drives and a few other tricks F                  LWW-DECWDISP.ZIP like SHOW DISPLAY LWW-FUNCT-KEYS.ZIP5                  Display keymap, define function keys E                  LWW-MODIFY-ATTRIB.ZIP Modify certain file attributes E                  LWW-PERSONA-DECW.ZIP Create new DECterm logged in as D                  someone else LWW-PERSONA.ZIP Log in as another userG                  (rewrite of GLOGIN) LWW-USERINFO.ZIP Display user info H                  from SYSUAF MORE.ZIP more/less for VMS - file displayerI                  TRF-LIB$SEARCH.ZIP C routine to simplify calling $search 6 [VU]            Things that work with VMS. Among them:@                  Alphaahb4 - Alpha cpu emulator (may not run anyC                  interesting OSs) am - math program, symbolic math, J                  claimed portable BOCHS - x86 emulator. Emulates a PC withD                  peripherals, enough to run Windows or Linux BRNCH -F                  Branches. A complete operator in a box. If you have aA                  tape or disk jukebox it can automatically mount, H                  dismount, catalog media and mount them automatically asH                  needed.  Only one is mounted per drive at a time but itI                  can catalog thousands of media depending only on size of I                  your jukebox. Full sources too, from Acorn Software/Dick F                  Munroe BZIP2 - very good compress/decompress utility;J                  does not preserve VMS attributes.  CDRECORD - burn CDs onH                  VMS CDRTOOLS - ditto, more elaborate.  CSWING - variousI                  your jukebox. Full sources too, from Acorn Software/Dick F                  Munroe BZIP2 - very good compress/decompress utility;J                  does not preserve VMS attributes.  CDRECORD - burn CDs onH                  VMS CDRTOOLS - ditto, more elaborate.  CSWING - variousE                  fixes for CSWING utility on VMS 7.3-1. Several folks E                  edited the code and fixed known problem symptoms but J                  there were somewhat different ideas about beautifying theH                  code.  CURL - ftp or http files from/to other locationsH                  from command line, restart partial transfers from pointG                  of interruption, etc. etc.  DFU027A - disk utility for J                  many disk filestructure maintenance operations GNV - UnixF                  like commands for VMS EXPAT - a web search engine forI                  VMS.  FREETYPE - handle TrueType fonts in VMS GAWK - Gnu G                  AWK for VMS Gnupg - crypto utility for lots of systems H                  Ghostscript - free Postscript interpreter. Does displayJ                  postscript too.  Gnuplot - plots graphs on most anything.B                  Graphicsmagick - converts image formats from mostB                  anything to most anything els$ GTK - yes, the GnuJ                  Toolkit, for VMS, graphics support.  GD - chart generatorH                  Imagemagick - converts image formats from most anythingG                  to most anything else Jasper - software for coding and G                  manipultation of images (even huge ones) Mozilla - Web G                  browser, modern successor to Netscape MySQL - complete G                  relational DBMS Nethack - networked famous game Ods2 - F                  portable read-ods2 (vms) disks code Python - portableJ                  language, many uses; somewhat string oriented.  STunnel -J                  Encrypted tcp/ip tunnel between VMS and other systems VMSI                  FAQ, all parts VB031 - Virtual Branches. Disk collection F                  in a box. Works with a jukebox of disks, lets many ofJ                  them (up to all) appear to be mounted at once even thoughH                  there are actually only 1 to a few drives.  Manages allD                  the volumes, even reduces fragmentation as they areE                  written. Works with CD or r/w disks, works perfectly F                  across clusters (and is not subject to the 500 volumeJ                  limit of MSCP server) even for jukeboxes (libraries) withA                  thousands of disks in them. From Acorn Software. I                  Xlockmore - lots of display hacks for locking a display. >                  Beautiful!  XV - view many image formats on XB [WWW]           WASD HTTP 1.1 server for VMS with many plugins and"                  extensions, V8.3.   ------------------------------  + Date: Sun, 21 Sep 2003 19:48:42 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)+ Subject: automatically mounting shadow sets $ Message-ID: <bkkvaq$kan$1@online.de>  F For a long time now, I've been using code like this (gleaned from this@ very newsgroup!) to mount disks (not only) at startup (this is aB subroutine called once for each disk in a hard-coded list with theF parameters physical device name, volume label and (optionally) logical name):     $MNT: SUBROUTINE$ $    IF .NOT. F$GETDVI(DEV,"EXISTS")	 $    THEN  $      EXIT 	 $    ELSE  $      IF F$GETDVI(DEV,"MNT")  $      THEN 
 $        EXIT  $      ENDIF
 $    ENDIF4 $!   If we host the device then mount it clusterwide< $    IF F$GETDVI(DEV,"HOST_NAME") .EQS. F$GETSYI("NODENAME")	 $    THEN  $      QUAL = "/CLUSTER"	 $    ELSE  $      QUAL = "/SYSTEM" 
 $    ENDIF
 $    SET NOON " $    ON SEVERE_ERROR THEN CONTINUEN $    IF F$GETDVI(DEV,"AVL") THEN MOUNT /NOASSIST 'QUAL' 'DEV' 'LABEL' 'LOGNAM'	 $    EXIT  $ENDSUBROUTINE  C This is located on DISK$USER.  DISK$USER is a shadow set mounted by 3 hard-coded commands in SYLOGICALS.COM on each node.   ? When I created my first shadow sets, I just entered stuff like    )    DSA100:/SHADOW=($2$DKA100:,$6$DKB200:)   F for the physical device name and, somewhat to my surprise, this works 0 fine with the F$GETDVI lexical (bug or feature).  D What doesn't work properly with shadow sets is the generation of theC /QUAL qualifier.  Since a shadow set is not MSCP served (though its G individual disks are in my case), F$GETDVI(DEV,"HOST_NAME") will always  return the current node name.   I The above logic mounts it cluster-wide if the node hosts the disk, since  G other nodes could not have already mounted it, and just system-wide if  F it is hosted by another node, with the assumption that it has already   been mounted on all other nodes.  E Is there any way to implement the same logic---mount only as much as  I necessary---with a shadow set (where the members can be on the same node  D or different nodes)?  Even if the node in question hosts one of the E members, the shadow set might be mounted with one less member on all  4 other nodes, in which case a MOUNT/SYSTEM is enough.  G On the other hand, perhaps ALWAYS using MOUNT/CLUSTER would be better;  8 the overhead can't be that high (but it is not elegant).  H I would also like a more elegant method of determining if a device is a H shadow set then just saying "DSA means shadow set" (which is true in my G case).  Surprisingly, there doesn't seem to be an argument of F$GETDVI  < to do this.  Is there anything better than checking whether 0 F$GETDVI(DEV,"SHDW_NEXT_MBR_NAME") is not empty?   ------------------------------  # Date: Sun, 21 Sep 2003 12:45:26 GMT 6 From: peter@langstoeger.at (Peter 'EPLAN' LANGSTOEGER)) Subject: Re: Cleaning up license database 5 Message-ID: <G3hbb.292334$ef4.3154843@news.chello.at>   i In article <_Cp8b.4653$CB2.1692@news.cpqcorp.net>, "Mark Buda" <buda@tabasco.zko.dec.no.spam.com> writes: D >"Peter 'EPLAN' LANGSTOEGER" <peter@langstoeger.at> wrote in message. >news:0Zn7b.12004$ef4.142165@news.chello.at...2 >> >I will add this to a WISH LIST for the future. >>H >> Super. And while you are there, please add LICENSE UNLOAD * (or /ALL) >as well...  >  >OK.   Thanks.   D >Please share with me how you plan on using it and what is gained by >implementing it.   J Sorry, I've to let this be unclear for the moment. (despite the fact, that it seems logical of course).  E I can't remember *exactly* what problem I had because of my vacation.   F My PAKs were to terminate during my vacation and I renewed them beforeD the vacation and loaded them. After my vacation the machine servicesD had nevertheless stopped because of terminated licenses. It seems, ID weren't able to load successfully the new licenses (though I think II remember to have seen no error messages) or I did load both licenses (the C old ones were at this time not yet terminated) and services stopped I because *part of the licenses* terminated (would be strange) or something  else.   J I tried to solve the problem then with LICENSE LOAD again, but obviously II couldn't replace the terminated licenses with the newer ones *in memory*. H After a few minutes I gave up and rebooted, and everything was ok again.  G And during this time, I really hated to have no "LICENSE UNLOAD/ALL" or D a "LICENSE UNLOAD/AUTH=*DECUS*" or a "LICENSE UNLOAD/ALL/CONFIRM"...   Does this help ?   --   Peter "EPLAN" LANGSTOEGER % Network and OpenVMS system specialist  E-mail  peter@langstoeger.atF A-1030 VIENNA  AUSTRIA              I'm not a pessimist, I'm a realist   ------------------------------  % Date: Sun, 21 Sep 2003 18:20:02 -0400 ' From: "Main, Kerry" <kerry.main@hp.com> E Subject: RE: Fork() and Unix compatibility work timetable and scope ? R Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB0C7C51@tayexc19.americas.cpqcorp.net>    Simon,    Re: UNIX compatibility -  0 http://h71000.www7.hp.com/portability/index.html  
 Check out:, http://h71000.www7.hp.com/portability/9.html   CRTL: / http://h71000.www7.hp.com/portability/crtl.html    Roadmaps for fork() etc:> http://h71000.www7.hp.com/openvms/roadmap/openvms_roadmaps.htm   File system enhancements: 6 http://h71000.www7.hp.com/portability/file_system.html  
 Future plans: 7 http://h71000.www7.hp.com/portability/future_plans.html      Portability "Jackets":; http://h71000.www7.hp.com/openvms/products/ips/porting.html    Regards     
 Kerry Main Senior Consultant  HP Services Canada Voice: 613-592-4660  Fax: 613-591-4477  Email: kerryDOTmainAThpDOTcom . (remove the DOT's and AT for email address)=20   -----Original Message----- From: Simon Clubley 8 [mailto:clubley@remove_me.eisner.decus.org-Earth.UFP]=20 Sent: September 8, 2003 7:21 AM  To: Info-VAX@Mvb.Saic.Com   E What is the current timetable for the availability of fork() on VMS ? 9 Will it have full Unix semantics or a restricted subset ?   G Also, what is the current scope and timetable of the Unix compatibility  work ?B Will you be able to, say, run configure against the standard Emacs$ source and then build Emacs on VMS ?  C BTW, I know that someone is working on porting a current version of G Emacs to VMS. My question is about been able to take the standard Emacs F source code and build it on this future version of VMS without any VMS# specific changes to the Emacs code.   H In other words, I am using Emacs as an example of something that I thinkH a full Unix compatibility layer should be able to build with very few or no changes.    Thanks for any information,    Simon.   --=20 D Simon Clubley, clubley@remove_me.eisner.decus.org-Earth.UFP      =20B VMS advocate: One who makes a Mac advocate look like a beginner at	 advocacy.    ------------------------------  # Date: Sun, 21 Sep 2003 13:54:49 GMT , From: Romax MANIAX <romax.maniax@pandora.be> Subject: How to uninstall?: Message-ID: <J4ibb.32068$9F.1635454@phobos.telenet-ops.be>  	 Hi gurus,   F I've installed Netlib on my super cool VAX4000-500 rackable model with
 vmsinstal.B Now I want to uninstall it, can you please tell me how to do that?   Thanks in advance.   ------------------------------  % Date: Sun, 21 Sep 2003 11:04:06 -0500 1 From: "David J. Dachtera" <djesys.nospam@fsi.net> $ Subject: Re: merging queue databases' Message-ID: <3F6DCBF6.49DC902E@fsi.net>   ( (Re-sequenced some items for clarity...)  / Phillip Helbig---remove CLOTHES to reply wrote:  > / > Let me try to state the problem more briefly.  > G > Node 3 (VAX) has no queue manager at all.  Nodes 1 and 2 have a queue J > manager, and everything is the default configuration.  Essentially, nodeH > 1 (ALPHA) has TCPIP queues and a batch queue and node 2 has just TCPIP	 > queues.  > I > What I want is one queue manager for all three nodes with all the files  > on a non-system disk.   F Somewhere on the 'web I believe there are instructions for how to moveB the queue database, if not in the docset. I don't recall the exactD steps, but it seems to me you'd have to shut down the queue manager,@ copy the queue database to the new location, then restart the QM; pointing to the new db-loc. See HELP START /QUEUE /MANAGER.   " > The TCPIP queues look like this: > ) > Generic server queue TCPIP$SMTP_NODE_00 O >   /GENERIC=(TCPIP$SMTP_NODE_01) /OWNER=[SYSTEM] /PROTECTION=(S:M,O:D,G:R,W:S)  >   /SCHEDULE=(NOSIZE) > H > Server queue TCPIP$SMTP_NODE_01, idle, on NODE::, mounted form DEFAULTA >   /BASE_PRIORITY=4 /DEFAULT=(FEED,FORM=DEFAULT) /OWNER=[SYSTEM] @ >   /PROCESSOR=TCPIP$SMTP_SYMBIONT /PROTECTION=(S:M,O:D,G:R,W:S) > G > My question: on such a "fully clustered" system, what should the SMTP J > queues look like?  Should they continue to look like they do above, i.e.H > a generic queue and an execution queue ON EACH NODE, named as they are< > above?  (In other words, TCPIP queues would continue to be > node-specific.)   C Speaking from my Multinet experience, you really only need a single E generic queue and node-specific execution queues on each node, at the 
 very most.  C > It might be easier to just delete everything, define the logicals J > appropriately, and start over, re-creating the queues I need, as opposedH > to merging the two databases I already have (I don't even know if this > is possible).   C Depends on how many forms you have defined (should be in a .COM for @ recovery, IMO), retained entries that have some importance, etc.  G > Would it be better to recreate the queues by hand, or should I re-run  > TCPIP$CONFIG?   F Dunno, but I'd expect TCPIP$CONFIG to be up to the task. Can't say for
 sure, though.    --   David J. Dachtera  dba DJE Systems  http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Sun, 21 Sep 2003 13:03:41 -0400 * From: JF Mezei <jfmezei.spamnot@istop.com>$ Subject: Re: merging queue databases) Message-ID: <3F6DD9EB.F965A879@istop.com>   / Phillip Helbig---remove CLOTHES to reply wrote:lI > What I want is one queue manager for all three nodes with all the filesr > on a non-system disk.o  L What you want are 3 queue managers, one on each node, sharing the same queue> manager databases in some directory accessible to all 3 nodes.  C > It might be easier to just delete everything, define the logicals J > appropriately, and start over, re-creating the queues I need, as opposed- > to merging the two databases I already haves  E No, take the most complete queue database, move the files to whateverkM directory you want them to me, then define the logicals on each node pointinge= to that directory, then start the queue manager on each node.n  N Then, you can repopulate the queue manager database with queues that belong toK each of the other 2 nodes. You define the queue to be /ON=node so the queueaG manager knows that when you submit a job on that queue, it is the queuey* manager of that node that handles the job.  L In terms of your SMTP queues, it depends if you want to do load balancing orK not. You could have one server queue pointing to 2 execution queues, one onmT each node. This way, traffic continues to flow even if one of the 2 nodes goes down.   ------------------------------  % Date: Sun, 21 Sep 2003 13:33:33 -0400r* From: JF Mezei <jfmezei.spamnot@istop.com>$ Subject: Re: merging queue databases) Message-ID: <3F6DE0E9.6C7BCDB6@istop.com>   I > > Would it be better to recreate the queues by hand, or should I re-runp > > TCPIP$CONFIG?   K The startup code simply issues TCPIP START MAIL to get the queues started.  H TCPIP HELP SET CONF SMTP reveals that there is a /QUEUES= qualifier that9 accepts a number of execution queues for that node only. d  9 Currently, do you have 2 totally separate SMTP servers ? v  N Perhaps the best bet would be to take the central database containing the SMTPN queues for node 1, and then manually recreate the SMTP queues for node 2. ThisN way, each node will retain its own separate SMTP queues and the change will be" transparent to the TCPIP software.   ------------------------------  + Date: Sun, 21 Sep 2003 18:46:18 +0000 (UTC)mP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)$ Subject: Re: merging queue databases$ Message-ID: <bkkrlq$g4n$1@online.de>  ; In article <3F6DCBF6.49DC902E@fsi.net>, "David J. Dachtera"r  <djesys.nospam@fsi.net> writes:   I > > Node 3 (VAX) has no queue manager at all.  Nodes 1 and 2 have a queue L > > manager, and everything is the default configuration.  Essentially, nodeJ > > 1 (ALPHA) has TCPIP queues and a batch queue and node 2 has just TCPIP > > queues.- > > K > > What I want is one queue manager for all three nodes with all the filese > > on a non-system disk.  > H > Somewhere on the 'web I believe there are instructions for how to moveD > the queue database, if not in the docset. I don't recall the exactF > steps, but it seems to me you'd have to shut down the queue manager,B > copy the queue database to the new location, then restart the QM= > pointing to the new db-loc. See HELP START /QUEUE /MANAGER.e  F That is my impression as well.  What I'm worried about is MERGING the  two default-style databases.  E > Speaking from my Multinet experience, you really only need a singleoG > generic queue and node-specific execution queues on each node, at thed > very most.   Sounds logical, but...  I > > Would it be better to recreate the queues by hand, or should I re-runh > > TCPIP$CONFIG?  > H > Dunno, but I'd expect TCPIP$CONFIG to be up to the task. Can't say for > sure, though.a  F ...I've never examined the TCPIP queues on a properly clustered system@ before.  Thus, I'd like to know what to expect.  I couldn't findF (admittedly, after just a quick look) where the TCPIP queues actually  get created.   ------------------------------  + Date: Sun, 21 Sep 2003 18:51:25 +0000 (UTC) P From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)$ Subject: Re: merging queue databases$ Message-ID: <bkkrvd$g4n$2@online.de>  2 In article <3F6DD9EB.F965A879@istop.com>, JF Mezei$ <jfmezei.spamnot@istop.com> writes:   K > > What I want is one queue manager for all three nodes with all the filesr > > on a non-system disk.e > N > What you want are 3 queue managers, one on each node, sharing the same queue@ > manager databases in some directory accessible to all 3 nodes.  F Perhaps.  However, the docs talk about the queue MANAGER being set up F to fall over to another node etc, so it sounds like I just need one.   See HELP START/QUE/MANA/ON .  A > In terms of your SMTP queues, it depends if you want to do load A > balancing or not. You could have one server queue pointing to 2eI > execution queues, one on each node. This way, traffic continues to flow1( > even if one of the 2 nodes goes down.   H I'm more concerned with failover than with load balancing for the TCPIP  queues.    ------------------------------    Date: 21 Sep 2003 14:36:50 -0700  From: nmanser@progis.de (Manser)$ Subject: Re: merging queue databases= Message-ID: <2178d61f.0309211336.38d0593d@posting.google.com>   | helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) wrote in message news:<bkjr2t$eup$1@online.de>.../ > Let me try to state the problem more briefly.o > H > Node 3 (VAX) has no queue manager at all.  Nodes 1 and 2 have a queue K > manager, and everything is the default configuration.  Essentially, node aI > 1 (ALPHA) has TCPIP queues and a batch queue and node 2 has just TCPIP s	 > queues.  > " > The TCPIP queues look like this: > ) > Generic server queue TCPIP$SMTP_NODE_00 O >   /GENERIC=(TCPIP$SMTP_NODE_01) /OWNER=[SYSTEM] /PROTECTION=(S:M,O:D,G:R,W:S)  >   /SCHEDULE=(NOSIZE) > H > Server queue TCPIP$SMTP_NODE_01, idle, on NODE::, mounted form DEFAULTA >   /BASE_PRIORITY=4 /DEFAULT=(FEED,FORM=DEFAULT) /OWNER=[SYSTEM]r@ >   /PROCESSOR=TCPIP$SMTP_SYMBIONT /PROTECTION=(S:M,O:D,G:R,W:S) > J > What I want is one queue manager for all three nodes with all the files  > on a non-system disk.e  < it should be a disk which is accessible at least to 2 nodes  (for quorum purposes).   example 1) r* a dual pathed disk, over a HSx controller  which is connected to 2 nodes.  
 example 2)? a shadowed disk with at least 2 members (1 member in one node,   the other in the other node)   > D > It might be easier to just delete everything, define the logicals K > appropriately, and start over, re-creating the queues I need, as opposed  I > to merging the two databases I already have (I don't even know if this - > is possible)..  ! before deleting everything do a :E: $show queue/all/full/output=queues_node.dat on every node 
 $mc sysman set env/cluster= do stop/queue/manager/clustert exit $u   Here are the steps:w  $ 1) in sylogicals.com (on all nodes) K    insert a line to mount the disk which contains the common queue database=  F 2) define the following logicals in sylogicals.com and at the $prompt.  ' $if f$getdvi("<the common disk>","MNT")m $then 4 $  define /system /exec sys$cluster disk:<directory>0 $  define /system /exec qman$master sys$cluster: $  ....c $endif   3) restart the queue manager  9 $start/queue/manager/on=(node1,node2,..) /new qman$mastern   4) control e   $sh queue /manager/fullt   $sh queue/manager/full  1 Master file:  $100$DKA0:[CLUSTER]QMAN$MASTER.DAT;r  5 Queue manager SYS$QUEUE_MANAGER, running, on VMAL05::t	   /ON=(*)n)   Database location:  $100$DKA0:[CLUSTER]a     $dir qman$master  # the following files should be theret   QMAN$MASTER.DATl SYS$QUEUE_MANAGER.QMAN$JOURNAL SYS$QUEUE_MANAGER.QMAN$QUEUESc  4 5) recreate the queues which are in the saved files.F    it is better to write the queues creation in systartup_vms.com for .    every node specific queue (print and batch)  . I would not recommend merging the 2 databases. > G > My question: on such a "fully clustered" system, what should the SMTP-J > queues look like?  Should they continue to look like they do above, i.e.H > a generic queue and an execution queue ON EACH NODE, named as they are< > above?  (In other words, TCPIP queues would continue to be > node-specific.)  > H > Would it be better to recreate the queues by hand, or should I re-run  > TCPIP$CONFIG?   . the tcpip smtp queues look like the following:  	 $sh queues  ) Generic server queue TCPIP$SMTP_VMAL05_00i  J Server queue TCPIP$SMTP_VMAL05_01, busy, on VMAL05::, mounted form DEFAULT  ) Generic server queue TCPIP$SMTP_VMAL06_00n  J Server queue TCPIP$SMTP_VMAL06_01, idle, on VMAL06::, mounted form DEFAULT       Hope this helps.   ------------------------------  % Date: Sun, 21 Sep 2003 15:59:43 -0500d1 From: "David J. Dachtera" <djesys.nospam@fsi.net> $ Subject: Re: merging queue databases' Message-ID: <3F6E113F.CC810B08@fsi.net>r  / Phillip Helbig---remove CLOTHES to reply wrote:  > 4 > In article <3F6DD9EB.F965A879@istop.com>, JF Mezei% > <jfmezei.spamnot@istop.com> writes:. > M > > > What I want is one queue manager for all three nodes with all the filesM > > > on a non-system disk.y > >/P > > What you want are 3 queue managers, one on each node, sharing the same queueB > > manager databases in some directory accessible to all 3 nodes. > G > Perhaps.  However, the docs talk about the queue MANAGER being set up3F > to fall over to another node etc, so it sounds like I just need one. > See HELP START/QUE/MANA/ON .  G What ya gotta watch out for is connectivity to the disk where the queue  database resides.   A If you're talking LAVC, then the disks are likely MSCP-served viaoE Ethernet. If the QM runs on the node that serves up the Q DB disk and-H you lose the QM's host node, you lose access to the Q DB. So, having the) QM move to another node is not an option.e  H If the Q DB disk is accessing via HSD/J/Z/G or EVA, then you can specifyG the QM to run on any node in a list you specify with START/QUE/MAN/ON=,n$ for example: /ON=(preferred_node,*).  C > > In terms of your SMTP queues, it depends if you want to do load C > > balancing or not. You could have one server queue pointing to 2aK > > execution queues, one on each node. This way, traffic continues to flown) > > even if one of the 2 nodes goes down.e > I > I'm more concerned with failover than with load balancing for the TCPIPr	 > queues.n  F Except in the case of a mail server, I've never seen load balancing to be an issue.   -- x David J. Dachteraf dba DJE Systemsd http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  + Date: Sun, 21 Sep 2003 21:49:49 +0000 (UTC)iP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)$ Subject: Re: merging queue databases$ Message-ID: <bkl6dt$rgr$1@online.de>  ; In article <3F6E113F.CC810B08@fsi.net>, "David J. Dachtera"   <djesys.nospam@fsi.net> writes:   I > What ya gotta watch out for is connectivity to the disk where the queued > database resides.  > C > If you're talking LAVC, then the disks are likely MSCP-served viaoG > Ethernet. If the QM runs on the node that serves up the Q DB disk andeJ > you lose the QM's host node, you lose access to the Q DB. So, having the+ > QM move to another node is not an option.   F LAVC is correct.  However, the disk with the "cluster-common" files is9 DISK$USER.  This is a shadow set (mounted on each node ineB SYLOGICALS.COM) of SCSI disks with members locally mounted on two G different nodes.  I would have to simultaneously lose both nodes, both  H disks or one node and the disk on the other to lose connection to this; F I think it's pretty robust.  (In general, in my LAVC, I like to mount I shadow sets on two different nodes, except for system disks---although I oH do mount them on all nodes, I figure that if I lose a node I can afford H to lose both members of its shadow set (and without dual-ported disks I F have to locally mount both members of a system-disk shadow set on the  same node anyway).)   J > If the Q DB disk is accessing via HSD/J/Z/G or EVA, then you can specifyI > the QM to run on any node in a list you specify with START/QUE/MAN/ON=,m& > for example: /ON=(preferred_node,*).  3 Can I do this with my SCSI LAVC as described above?m  E > > > In terms of your SMTP queues, it depends if you want to do load2E > > > balancing or not. You could have one server queue pointing to 2jM > > > execution queues, one on each node. This way, traffic continues to flow@+ > > > even if one of the 2 nodes goes down.n > > K > > I'm more concerned with failover than with load balancing for the TCPIP  > > queues.  > H > Except in the case of a mail server, I've never seen load balancing to > be an issue.  B I've now done the following: On the node with no queue manager, I ; defined QMAN$MASTER to point to the new location and did a  D START/QUE/MANAGER/NEW/ON=(just_that_node).  After shutting down and ? starting up TCPIP, the "standard" SMTP queues were created and d everything seems to work.   G As far as I can tell, I'm now running a queue manager on each node with > its own set of database files.  Presumably, with /ON specifiedH correctly, it would be OK to run it on just one node, after I solve the 2 issue of merging the three sets of database files.  G With everything but the SMTP queues, I'm happy to delete all queues and@G re-create them, and also just run one queue manager.  Before I do this, H however, I'd like to know what the SMTP queues should look like.  At the= moment, each node has a generic queue and an execution queue.5H Presumably, I need just one generic queue, rather than one on each node,D and an execution queue on each node.  (Actually, I suppose just one G execution queue which can fail over is all I need---as long as I'm not  I worried about load-balancing---but TCPIP seems to want the generic queue uG pointing to 1 execution queue, so I don't want to change that by hand.)   @ Is anyone running HP TCP/IP Services in a cluster with ONE queue- manager?  What do your SMTP queues look like?.   ------------------------------  + Date: Sun, 21 Sep 2003 22:09:37 +0000 (UTC)bP From: helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply)$ Subject: Re: merging queue databases$ Message-ID: <bkl7j1$s14$1@online.de>  = In article <2178d61f.0309211336.38d0593d@posting.google.com>,p# nmanser@progis.de (Manser) writes: s   > example 2)A > a shadowed disk with at least 2 members (1 member in one node, v > the other in the other node)   OK, that's my case.h  F > > It might be easier to just delete everything, define the logicals M > > appropriately, and start over, re-creating the queues I need, as opposed eK > > to merging the two databases I already have (I don't even know if this e > > is possible).n  # > before deleting everything do a :m< > $show queue/all/full/output=queues_node.dat on every node  > $mc sysman > set env/clustera > do stop/queue/manager/cluster  > exit > $  >  > Here are the steps:o > & > 1) in sylogicals.com (on all nodes) M >    insert a line to mount the disk which contains the common queue databasel > H > 2) define the following logicals in sylogicals.com and at the $prompt. > ) > $if f$getdvi("<the common disk>","MNT")e > $then-6 > $  define /system /exec sys$cluster disk:<directory>2 > $  define /system /exec qman$master sys$cluster:	 > $  ....G > $endif >  > 3) restart the queue manager > ; > $start/queue/manager/on=(node1,node2,..) /new qman$master   
 Sounds OK.  
 > 4) control t >  > $sh queue/manager/full > 3 > Master file:  $100$DKA0:[CLUSTER]QMAN$MASTER.DAT;s > 7 > Queue manager SYS$QUEUE_MANAGER, running, on VMAL05::  >   /ON=(*)m+ >   Database location:  $100$DKA0:[CLUSTER]i  E Presumably, you get the same output on each node.  What puzzles me is H that the documentation says that the default location WILL NOT WORK in aE cluster with more than one system disk.  In my case, it works.  OK, I A have different queues on each node and I want to make things moretH node-independent, which is why I am looking into this, but it does seem F to work OK as it is.  (I haven't tried creating a queue with the same = name on two nodes, though I don't see why it shouldn't work.)I  0 > I would not recommend merging the 2 databases.   Is it even possible?  I > > My question: on such a "fully clustered" system, what should the SMTPeL > > queues look like?  Should they continue to look like they do above, i.e.J > > a generic queue and an execution queue ON EACH NODE, named as they are> > > above?  (In other words, TCPIP queues would continue to be > > node-specific.)  > > J > > Would it be better to recreate the queues by hand, or should I re-run  > > TCPIP$CONFIG?  > 0 > the tcpip smtp queues look like the following: >  > $sh queue  > + > Generic server queue TCPIP$SMTP_VMAL05_00a > L > Server queue TCPIP$SMTP_VMAL05_01, busy, on VMAL05::, mounted form DEFAULT > + > Generic server queue TCPIP$SMTP_VMAL06_00g > L > Server queue TCPIP$SMTP_VMAL06_01, idle, on VMAL06::, mounted form DEFAULT  H This looks like my system at the moment, though on each node I just see  the queues on that node.  E What's the point of having a generic SMTP queue on each node, rather  E than just one which can fail over?  If one did have just one generic ,G SMTP queue, what is the point, other than load-balancing, in having an nE execution queue on each node?  Could one theoretically have just one  + execution queue, with fail-over capability?e   ------------------------------  % Date: Sun, 21 Sep 2003 21:01:07 -0500k1 From: "David J. Dachtera" <djesys.nospam@fsi.net> $ Subject: Re: merging queue databases' Message-ID: <3F6E57E3.DC583678@fsi.net>a  / Phillip Helbig---remove CLOTHES to reply wrote:  > = > In article <3F6E113F.CC810B08@fsi.net>, "David J. Dachtera" ! > <djesys.nospam@fsi.net> writes:  > K > > What ya gotta watch out for is connectivity to the disk where the queue  > > database resides.  > >eE > > If you're talking LAVC, then the disks are likely MSCP-served viatI > > Ethernet. If the QM runs on the node that serves up the Q DB disk andCL > > you lose the QM's host node, you lose access to the Q DB. So, having the- > > QM move to another node is not an option.s > H > LAVC is correct.  However, the disk with the "cluster-common" files is; > DISK$USER.  This is a shadow set (mounted on each node inCC > SYLOGICALS.COM) of SCSI disks with members locally mounted on two H > different nodes.  I would have to simultaneously lose both nodes, bothI > disks or one node and the disk on the other to lose connection to this;IG > I think it's pretty robust.  (In general, in my LAVC, I like to mount J > shadow sets on two different nodes, except for system disks---although II > do mount them on all nodes, I figure that if I lose a node I can affordtI > to lose both members of its shadow set (and without dual-ported disks IhG > have to locally mount both members of a system-disk shadow set on the  > same node anyway).)p  4 Well, understand: DSA devices canNOT be MSCP-served!   For example:   DSA1:c   NODEA$DKA100:i   NODEB$DKC200:a  3 If NODEA and NODEB both go down, DSA1 "disappears".m  L > > If the Q DB disk is accessing via HSD/J/Z/G or EVA, then you can specifyK > > the QM to run on any node in a list you specify with START/QUE/MAN/ON=, ( > > for example: /ON=(preferred_node,*). > 5 > Can I do this with my SCSI LAVC as described above?y  H Sure, but you'll need to limit yourself to /ON=(NODEA,NODEB) or whatever= two nodes have a direct SCSI path to the cluster-common disk.   G > > > > In terms of your SMTP queues, it depends if you want to do load G > > > > balancing or not. You could have one server queue pointing to 2 O > > > > execution queues, one on each node. This way, traffic continues to flow - > > > > even if one of the 2 nodes goes down.t > > > M > > > I'm more concerned with failover than with load balancing for the TCPIP 
 > > > queues.S > >OJ > > Except in the case of a mail server, I've never seen load balancing to > > be an issue. > C > I've now done the following: On the node with no queue manager, In< > defined QMAN$MASTER to point to the new location and did aE > START/QUE/MANAGER/NEW/ON=(just_that_node).  After shutting down andp@ > starting up TCPIP, the "standard" SMTP queues were created and > everything seems to work.n > I > As far as I can tell, I'm now running a queue manager on each node witht! > its own set of database files. i  H If SHOW QUE/MAN/FU gives the same queue manager name (SYS$QUEUE_MANAGER)? on each node, then yes: you have three separate queue managers.V   -- E David J. DachteraT dba DJE SystemsG http://www.djesys.com/  ( Unofficial Affordable OpenVMS Home Page: http://www.djesys.com/vms/soho/    ------------------------------  % Date: Sun, 21 Sep 2003 19:46:47 -0500p0 From: Keith Brown <kbrown2720@NOSPAMcomcast.net>/ Subject: Multichrom support on OpenVMS, anyone?s0 Message-ID: <3ZGdnRScFtrZ2vOiXTWJiQ@comcast.com>   To all,n  O My company has for many years (about 15) used an application called Multichrom .K (runs on OpenVMS VAX and Alpha) to do gas chromatography. Ownership of the  I application has been passed around over the years. It now resides with a eM company called LabSystems. LabSystems says they will discontinue support for iO Multichrom in the near future. This is an attempt to force us to buy their new tI Windows GC solution.  Is anyone aware of a 3rd party who does Multichrom  O support? It has been estimated that there may be thousands of Multichrom sites  P still out there in the world. Is there anyone who will take my support dollar$?  BTW, we are located in the US.     -- d Keith Browns   ------------------------------  % Date: Sun, 21 Sep 2003 13:17:13 +0100t0 From: "Dr. David Kirkby" <drkirkby@ntlworld.com>9 Subject: Re: OpenVMS hobbyyest - free to join or not ????l, Message-ID: <3F6D96C9.FA46BF00@ntlworld.com>   Graham Burley wrote: >  > Dr. David Kirkby wrote:s > >nC > > I just thought I'd say I was able to get a DECUS ID number from  > > Encompass in the USA > >t3 > > http://www.encompassus.org/membership/join.html  > >  > > for zero cost. > >nJ > > That seems somewhat better value than paying the 52.88 (~ $88) to get > > it fromo > >lC > > http://www.hp-interex.org/site/shop/shopsublistnonmem.asp?id=91r > >t > > in the UK. > >  > B > I really don't understand what the UK HPUG are on about. They'reC > advertising 6 types of membership this year, only 3 with Hobbyist @ > entitlement. The Hobbyist License membership is qualified with? > "one renewal only", and costs more than subscription to theiru= > technical journal. I hope this means that they are making aL1 > generous contribution to the folks at Montagar.l  D I had quite a lengthy email from someone who is a member of the HPUGE Interim Board, which was copied to the co-chairman of the user group.gE Apparently hobbyist membership used to be free in the UK too. It said > that due to various changes that happened and various problemsE associated with getting funding from sponsors (including Digital), ito? became necessary to charge for membership. He said that withoutDC charging, there could be no usergroup and hence no OpenVMS HobbyistEA Scheme. It was also stated that it actually requires just as muchEE administration at the time of renewal as any other membership grade. E  F Whatever the case, I'm certainly not going to join the UK HPUG, when IH can get what I want (i.e. the DECUS number) for nothing from the USA viaF Encompass. I've not compared the pros/cons of Encompass vs. UK HPUG asH regards to other benefits, but for what I personally want (which is just6 the number), it makes no sense to get it from the UK.      > > Dr. David Kirkby,d > > Senior Research Fellow, " > > Department of Medical Physics, > > University College London, > H > Ah! You might well find VMS media easier than you think - someone from6 > UCL Mech Eng posts to comp.os.vms from time to time.  A Cheers, when I get some disks sorted out I'll probably looking atrH contacting them. I've been looking on eBay UK to buy some disks, but theF prices seem silly. There was a ultra-modern U320 144 Gb SCSI disk thatG sold for about 137 pounds a few days back, yet 18 Gb disks are fetchingdA thirty odd pounds for a semi-modern 1" high ones. I actually needsG several disks - two for the Alpha, another for a Sun and perhaps one orsH two more for my HP or RS/6000. I might try asking around to see if I canH find someone who has several, as I don't want to pay postage on each andG every disk, which makes the smaller disks very unattractively priced. IaG might just buy two large disks (144 Gb each) put them in my Sun and I'd?H have a load of smaller disks (1 x 73, 3 x 36) to move to other machines.E I have my external disks in Sun 611 boxes, which don't do a very goodaG job of cooling the things, so I'd like to move the disks inside the SunOC where they get better cooling. Once I get this all sorted out, I'lli install OpenVMS!   > Graham   -- kA "The day Microsoft makes something that doesn't suck is probably t> the day they start making vacuum cleaners." -Ernst Jan Plugge.   Dr. David Kirkby,  Senior Research Fellow,a Department of Medical Physics, University College London," 11-20 Capper St, London, WC1E 6JA., Website: http://www.medphys.ucl.ac.uk/~davek- Author of 'atlc' http://atlc.sourceforge.net/e   ------------------------------  % Date: Sun, 21 Sep 2003 14:12:00 -0400)* From: JF Mezei <jfmezei.spamnot@istop.com>< Subject: Preserving decwindows session through disconnection) Message-ID: <3F6DE9EA.C9AD1A1D@istop.com>n  Y Lets say I have a decwindows (vax vms) session established to a X windows server/display.e  N If I turn off the X windows display (or quit the x emulation software), my VMSM session is killed and I must login again and reopen, place etc windows to get  to where I was before.  N Is there a way to preserve an x windows session such that if the X server goes? away, when it comes back, I could get all the previously openedt; applications/windows to redisplay the exact way they were ?.   ------------------------------  % Date: Sun, 21 Sep 2003 14:53:13 -0700* From: "ynotssor" <"ynotssor">a( Subject: Re: Prove this corrective patch, Message-ID: <3f6e1e44_2@corp.newsgroups.com>  7 "Jem Berkes" <jem@users.pc9.EXTRA_org> wrote in message 1 news:Xns93FDA93C4E162jbuserspc9org@130.179.16.24 o  < > Got over 2,000 just today, and scrapped that email address > temporarily. e > G > I created a new USENET posting address, within 1 hour of using it gotdC > another 50. This page http://vil.nai.com/vil/content/v_100662.htmaH > says "Within the list of servers carried in the worm are multiple NNTPG > servers. Analysis is currently ongoing to determine exactly how theseL > are used"g > : > Umm... I think I know how those NNTP servers are used :)  . Notice how "SWEN" is "NEWS" spelled backwards?     -- e% use hotmail com for any email replies     > -----= Posted via Newsfeeds.Com, Uncensored Usenet News =-----A http://www.newsfeeds.com - The #1 Newsgroup Service in the World!D? -----==  Over 100,000 Newsgroups - 19 Different Servers! =-----T   ------------------------------   Date: 21 Sep 2003 21:38:11 GMT* From: Jem Berkes <jem@users.pc9.EXTRA_org>( Subject: Re: Prove this corrective patch9 Message-ID: <Xns93FDA93C4E162jbuserspc9org@130.179.16.24>0  G > Any one else get more than 548 versions of this in the last two days?n  G Got over 2,000 just today, and scrapped that email address temporarily.9  F I created a new USENET posting address, within 1 hour of using it got A another 50. This page http://vil.nai.com/vil/content/v_100662.htmeG says "Within the list of servers carried in the worm are multiple NNTP LJ servers. Analysis is currently ongoing to determine exactly how these are  used"t  8 Umm... I think I know how those NNTP servers are used :)   -- e
 Jem Berkes http://www.sysdesign.ca/   ------------------------------    Date: 21 Sep 2003 19:40:30 -0700. From: spamsink2001@yahoo.com (Alan E. Feldman)? Subject: Re: strangeness with temporary text files used by MAILd= Message-ID: <b096a4ee.0309211840.7042c4d0@posting.google.com>u  | helbig@astro.multiCLOTHESvax.de (Phillip Helbig---remove CLOTHES to reply) wrote in message news:<bkid82$b98$2@online.de>...J > Each user has SYS$SCRATCH defined as DISK$SCRATCH:[USERNAME].  MAIL has H > EDT as the editor.  From most accounts, after exiting a file composed G > with SEND/EDIT to send the message, I see something like this on the l	 > screen:v > 4 >    DISK$SCRATCH:[USERNAME]MAIL_203470C7_SEND.TMP;1 > J > However, from one account I don't.  The file IS written to disk and the @ > message is sent, though.  Output of SHOW TERMINAL is the same. >  > Any ideas?     Well, one theory:o  A This user has EDTSYS or a similar logical name defined to his EDTkA startup file which contains the SET NOSUMMARY command. (I haven'taA tried this with MAIL, but I would expect the logical name to taketE effect for all EDT sessions that don't explicitly specify otherwise.)    Disclaimer: JMHO Alan E. Feldmanu   ------------------------------  # Date: Mon, 22 Sep 2003 02:20:15 GMTl; From: "John Gemignani, Jr." <jon-nope@thiswontworkossc.net>s4 Subject: Re: TCP/IP Services for VMS - sloppy output< Message-ID: <z%sbb.2210$iT4.1602059@news1.news.adelphia.net>  , "Roy Omond" <Roy@Omond.net> wrote in message# news:3F695D3F.823DCFA4@Omond.net...y > "John Gemignani, Jr." wrote: >- > > [...snip...] > >0L > > That's because the person who coded the TCPIP utility (nee, UCX utility)D > > used SMG routines to deal with the screen.  The entire screen is	 formattedrG > > then WHAM it's put up in one shot, so it contains all of the CR/LFs  acrossD > > multiple lines as opposed to individual line output appearing as separate > > records in the output. >-J > Well, that was a really stupid decision to have made.  Very presumptiousJ > that the output is going to a screen, and also unnecessarily introducing > OS-specific code.F >9= > Any chance of rescinding that decision and fixing the bug ?   K The engineer who did this work is long since gone.  There is no chance that  I will have anythingJ to do with fixing it as I will be long gone in the not-too-distant future.   -Johna   >a > Roy Omond  > Blue Bubble Ltd. >a   ------------------------------  # Date: Mon, 22 Sep 2003 01:50:38 GMTa> From: Michael Austin <maustin@no-more-spam.firstdbasource.com>! Subject: TCPIP  Reject processingt; Message-ID: <Ozsbb.1458$xj1.105@newssvr23.news.prodigy.com>e  = Compaq TCP/IP Services for OpenVMS Alpha Version V5.3 - ECO 2 5    on a AlphaServer 2100 4/200 running OpenVMS V7.3-1t     ... could use some changes:    In STMP.CONFIG:h; Reject-Mail-From: *Microsoft*Services*,*Microsoft*Support*,,( *Microsoft*System*,*Microsoft*Security*,, *Microsoft*Bulletin*,*Microsoft*Assistance*,' *MS*Security*,*MS*Support*,*MS*System*,e+ *MS*Service*,*MS*Bulletin*,*MS*Assistance*,G/ *Microsoft*Customer*Assistance*,*Email*Service*E  3 This seems to only work if the string is in the < > : This latest round of SPAM has the following in the headers  @ FROM: "Microsoft Email Storage Service" <removed@to-protect.com>  F NOTE this may be a good address but the "name" contains what I really 5 need to be looking for ...not just within the < >....e 	w   CORRECT BEHAVIOR $telnet 0/port=25e& %TELNET-I-TRYING, Trying ... 127.0.0.16 %TELNET-I-SESSION, Session 01, host localhost, port 25H 220 mysys.mydomain.com V5.3-18E, OpenVMS V7.3-1 Alpha ready at Sun, 21 S helo; 250 mysys.mydomain.com Hello LOCALHOST, pleased to meet youi> MAIL FROM: "Microsoft Services" <MicrosoftServices@nowhre.com>8 550 "You have been infected with a virus. Call support."     WRONG BEHAVIOR $telnet 0/port=25t& %TELNET-I-TRYING, Trying ... 127.0.0.16 %TELNET-I-SESSION, Session 01, host localhost, port 25H 220 mysys.mydomain.com V5.3-18E, OpenVMS V7.3-1 Alpha ready at Sun, 21 S helo; 250 mysys.mydomain.com Hello LOCALHOST, pleased to meet you = MAIL FROM: "Microsoft Services" <somepoorschmuck@nowhere.com>a RCPT TO: met. 250 <somepoorschmuck@nowhere.com>... Sender OK 250 < me>... Recipient OKe DATA, 354 Start mail input; end with <CRLF>.<CRLF> this is a testA . << this line changed from single .(period) so email would work.h 250 OK quit; 221 mysys.mydomain.com Service closing transmission channel - %TELNET-S-REMCLOSED, Remote connection closed/6 -TELNET-I-SESSION, Session 01, host localhost, port 250 New mail from SMTP%"somepoorschmuck@nowhere.com"   ------------------------------  % Date: Sun, 21 Sep 2003 14:01:39 -0400s* From: JF Mezei <jfmezei.spamnot@istop.com>) Subject: Re: The vultures are circulatingp( Message-ID: <3F6DE77E.26A8F27@istop.com>   "Main, Kerry" wrote:H > The Sector7 folks live-n-breathe on change. If everyone was happy with9 > their current OS vendor, they would be out of business.e > J > They are well known internally to HP (and I leave it to you to determine" > whether this is good or bad) :-)  H I have a feeling that some parfts of HP love thse folks if they can helpL convert some VMS customers to HP-UX. Deep down, that is what HP would really want, isn't it ?    J (I realise that this doesn't apply to the VMS group, but I have a distinct2 feeling it applies above and around the VMS group.   ------------------------------  % Date: Sun, 21 Sep 2003 18:42:05 -0400a' From: "Main, Kerry" <kerry.main@hp.com> ) Subject: RE: The vultures are circulatingYR Message-ID: <FD827B33AB0D9C4E92EACEEFEE2BA2FB0C7C52@tayexc19.americas.cpqcorp.net>   >=20 > -----Original Message-----6 > From: JF Mezei [mailto:jfmezei.spamnot@istop.com]=20" > Sent: September 21, 2003 2:02 PM > To: Info-VAX@Mvb.Saic.Com  >=20 > "Main, Kerry" wrote:B > > The Sector7 folks live-n-breathe on change. If everyone was=20 > happy with=20i; > > their current OS vendor, they would be out of business.- > >=20E > > They are well known internally to HP (and I leave it to you to=20i. > > determine whether this is good or bad) :-) >=20? > I have a feeling that some parfts of HP love thse folks if=20oB > they can help convert some VMS customers to HP-UX. Deep down,=20/ > that is what HP would really want, isn't it ?s >=20 >=20? > (I realise that this doesn't apply to the VMS group, but I=20rD > have a distinct feeling it applies above and around the VMS group. >=20 >u  3 Re: HP wanting to convert VMS Customers to HP-UX ..t  ? Well, while I would say there was certainly a very strong HP-UXoF corporate influence at the time of the merger, but my own $.02 is thatG it seems that the overall approach internally has matured a great deal.lF It is now much more of a "better together" approach than expecting any< one single OS to dominate and/or be right for every Customer environment.=20   A And this does not just mean a better understanding of the OpenVMSiH Customer environment, but also the Linux, NSK and Windows world as well.   Regards-  
 Kerry Main Senior Consultantn HP Services Canada Voice: 613-592-4660v Fax: 613-591-4477d Email: kerryDOTmainAThpDOTcom . (remove the DOT's and AT for email address)=20   ------------------------------   End of INFO-VAX 2003.525 ************************