1 INFO-VAX	Fri, 15 Dec 2000	Volume 2000 : Issue 699       Contents:+ Re: "process crash" vs. "application crash" + Re: "process crash" vs. "application crash" C Re: %INSTALL-I-NONSHRADR, image installed ignoring '/SHARE=ADDRESS' C Re: %INSTALL-I-NONSHRADR, image installed ignoring '/SHARE=ADDRESS' C %MOUNT-F-DIFVOLMNT, different volume already mounted on this device G Re: %MOUNT-F-DIFVOLMNT, different volume already mounted on this device G Re: %MOUNT-F-DIFVOLMNT, different volume already mounted on this device 9 %SYSINIT-E, error mounting system device, status=0072832C = Re: %SYSINIT-E, error mounting system device, status=0072832C = Re: %SYSINIT-E, error mounting system device, status=0072832C 0 Re: A wonderful message from an antivirus vendor0 Re: A wonderful message from an antivirus vendor0 Re: A wonderful message from an antivirus vendor Re: ALTPRI in stead of SETPRI ! Re: Cluster advertising in the UK  Re: INITIALIZE/MAXIMUM_FILES0 Re: It's 20 years since I first logged onto VMS! Re: Logging OPCOM messages	 Max Users 
 Re: Max Users * Re: Mouse hang: AlphaStation 1000A VMS 6.2* Re: Mouse hang: AlphaStation 1000A VMS 6.2 Re: Mozilla 0.6  Re: Mozilla 0.6  Re: Mozilla 0.6  Re: Mozilla 0.6  Re: Mozilla 0.6 % Re: NI-style cluster upgrade question   Re: OpenVMS Hobbyist TCP/IP Help Pathworks 32 Press Release from ISD Press Release from U.S. Design  Strange access violation problem$ Re: Strange access violation problem Re: Sun Cluster  Re: Sun Cluster  SYS$CANCEL and CMU-IP stack  Re: TCP Bug - MULTINET + UCX Re: TCP Bug - MULTINET + UCX  Re: Virtual I/O Cache and Oracle Re: Weird I/O problem  Re: Weird I/O problem   F ----------------------------------------------------------------------    Date: 15 Dec 2000 11:58:31 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> 4 Subject: Re: "process crash" vs. "application crash"H Message-ID: <y4wvd29ciw.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  M It already has all privileges, so turn on PCB$V_NODELET and wait for the next @ opportunity to analyse the then hopefully hung process with SDA:   	Jan   ------------------------------  # Date: Fri, 15 Dec 2000 12:20:46 GMT = From: system@SendSpamHere.ORG (Brian Schenkenberger, VAXman-) 4 Subject: Re: "process crash" vs. "application crash"0 Message-ID: <009F49F0.017BB8DF@SendSpamHere.ORG>   In article <y4wvd29ciw.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>, Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de> writes:N >It already has all privileges, so turn on PCB$V_NODELET and wait for the nextA >opportunity to analyse the then hopefully hung process with SDA:  >  >	Jan   I Not such sage advice.  A process which incurs an unhandled exception will I eventually be handled by the system by terminating the process.  If it is H immortal by virtue of the PCB$V_NODELET bit being set, you can have thisI process looping and consuming a large amount of CPU cycles because you've I created a moral dilema for the process -- Kill myself?  But I'm immortal! ( I've seen such problems all too often.    H I find it difficult to believe that there is no trace information at allH from the original poster's process.  I'd suggest rebuilding the program H and produce full-blown .LISting and .MAP files.  Then, look at the loca-D tions in any trace that is available and reconcile with these files.   --O VAXman- OpenVMS APE certification number: AAA-0001     VAXman(at)TMESIS(dot)COM              O city, n., 1. a place where trees are cut down and streets are named after them.    ------------------------------  % Date: Fri, 15 Dec 2000 16:06:32 -0000 - From: wspencer@ap.nospam.org (Warren Spencer) L Subject: Re: %INSTALL-I-NONSHRADR, image installed ignoring '/SHARE=ADDRESS'/ Message-ID: <t3kgc8lktndr87@news.supernews.com>   - helbig@astro.rug.nl (Phillip Helbig) wrote in $ <91c3c6$7a7$1@info.service.rug.nl>:   " >Is this something to worry about? > @ >%ADX-I-INSTALLDONE, OpenVMS Alpha Debug V7.2X installation done@ >%VMSINSTAL-I-MOVEFILES, Files will now be moved to their target> >directories... %INSTALL-I-NONSHRADR, image installed ignoringA >'/SHARE=ADDRESS' DISK$ALPHASYS:<S YS0.SYSCOMMON.SYSLIB>TRACE.EXE H >-SYSTEM-F-NOPRIV, insufficient privilege or object protection violation >  >   K I'm not familiar with TRACE.EXE, but I would be concerned.  Are you logged  / into the SYSTEM account when doing the install?    ws   --  3 << What if there were no hypothetical questions? >>    Warren Spencer Senior Software Engineer The Associated Press  ? ** My employer does not necessarily agree with my statements **    ------------------------------   Date: 15 Dec 2000 16:37:59 GMT* From: helbig@astro.rug.nl (Phillip Helbig)L Subject: Re: %INSTALL-I-NONSHRADR, image installed ignoring '/SHARE=ADDRESS'. Message-ID: <91dhd7$l51$1@info.service.rug.nl>  F In article <t3kgc8lktndr87@news.supernews.com>, wspencer@ap.nospam.org (Warren Spencer) writes:    $ > >Is this something to worry about? > > B > >%ADX-I-INSTALLDONE, OpenVMS Alpha Debug V7.2X installation doneB > >%VMSINSTAL-I-MOVEFILES, Files will now be moved to their target@ > >directories... %INSTALL-I-NONSHRADR, image installed ignoringC > >'/SHARE=ADDRESS' DISK$ALPHASYS:<S YS0.SYSCOMMON.SYSLIB>TRACE.EXE J > >-SYSTEM-F-NOPRIV, insufficient privilege or object protection violation > M > I'm not familiar with TRACE.EXE, but I would be concerned.  Are you logged  1 > into the SYSTEM account when doing the install?   F No, but a SETPRV with /PRIV=ALL enabled.  Has always worked before....   ------------------------------   Date: 15 Dec 2000 08:59:59 GMT* From: helbig@astro.rug.nl (Phillip Helbig)L Subject: %MOUNT-F-DIFVOLMNT, different volume already mounted on this device. Message-ID: <91cmif$clb$1@info.service.rug.nl>  C I get this error from a .COM file after moving from 7.1 to 7.2 VAX.   C I have a .COM file which mounts disks /CLUSTER if it hosts them and F /SYSTEM otherwise.  The idea is that each node executes this file whenH it comes up, and thus mounts /SYSTEM all disks which are already mountedB in the rest of the cluster, and /CLUSTER all disks which it itself hosts.     Worked fine before.   H Now, I have a problem with a CD.  It is the 7.1 VMS documentation on theH CD drive on an ALPHA.  The only problem exists for this disk; all othersD are OK.  I get the fatal error message in the subject when trying to? mount if from the VAX.  Note that it is the same .COM doing the A mounting, so I can't have mis-typed the volume label or whatever.   
 Any ideas?   ------------------------------  % Date: Fri, 15 Dec 2000 10:40:45 +0000 4 From: John Laird <john@laird-towers.freeserve.co.uk>P Subject: Re: %MOUNT-F-DIFVOLMNT, different volume already mounted on this device8 Message-ID: <91tj3topa0kiiqb4gbumua5q93t887ab1g@4ax.com>  H On 15 Dec 2000 08:59:59 GMT, helbig@astro.rug.nl (Phillip Helbig) wrote:  D >I get this error from a .COM file after moving from 7.1 to 7.2 VAX. > D >I have a .COM file which mounts disks /CLUSTER if it hosts them andG >/SYSTEM otherwise.  The idea is that each node executes this file when I >it comes up, and thus mounts /SYSTEM all disks which are already mounted C >in the rest of the cluster, and /CLUSTER all disks which it itself  >hosts.  >  >Worked fine before. > I >Now, I have a problem with a CD.  It is the 7.1 VMS documentation on the I >CD drive on an ALPHA.  The only problem exists for this disk; all others E >are OK.  I get the fatal error message in the subject when trying to @ >mount if from the VAX.  Note that it is the same .COM doing theB >mounting, so I can't have mis-typed the volume label or whatever. >  >Any ideas?   F There have been lots of problems in this area - I recall it arrived atE 7.0 or 7.1 and affected 6.2 machines.  I don't think we ever fixed it B completely, despite applying every likely ECO.  Take a look at the2 available patches - it'll probably be a MOUNT kit.  * Bit of a sorry state of affairs, this bug.   	John  --  
 John Laird   ------------------------------  % Date: Fri, 15 Dec 2000 12:28:30 +0100   From: Paul Sture <paul@sture.ch>P Subject: Re: %MOUNT-F-DIFVOLMNT, different volume already mounted on this device+ Message-ID: <VA.000001df.0550542f@sture.ch>   D In article <91cmif$clb$1@info.service.rug.nl>, Phillip Helbig wrote:, > From: helbig@astro.rug.nl (Phillip Helbig) > Newsgroups: comp.os.vms N > Subject: %MOUNT-F-DIFVOLMNT, different volume already mounted on this device  > Date: 15 Dec 2000 08:59:59 GMT > E > I get this error from a .COM file after moving from 7.1 to 7.2 VAX.  > E > I have a .COM file which mounts disks /CLUSTER if it hosts them and H > /SYSTEM otherwise.  The idea is that each node executes this file whenJ > it comes up, and thus mounts /SYSTEM all disks which are already mountedD > in the rest of the cluster, and /CLUSTER all disks which it itself	 > hosts.   >  > Worked fine before.  > J > Now, I have a problem with a CD.  It is the 7.1 VMS documentation on theJ > CD drive on an ALPHA.  The only problem exists for this disk; all othersF > are OK.  I get the fatal error message in the subject when trying toA > mount if from the VAX.  Note that it is the same .COM doing the C > mounting, so I can't have mis-typed the volume label or whatever.  >  > Any ideas? > , Couple of points, which may or may not help.  M 1. The order in which devices are delivered is not guaranteed using f$device   (if you're using that).   M 2. In their efforts to ensure that important parameters are preserved across  O the upgrade, the upgrade procedures typically turn modparams.dat into somewhat  K of a mess. Duplicate entries are appended, and it is easy to fall into the  M trap of amending an early entry without realising that it's duplicated lower  N down with the old value. You'll probably want to revise your modparams.dat in = the light of this, even if it isn't the cause of the problem.  ___ 
 Paul Sture Switzerland    ------------------------------   Date: 15 Dec 2000 15:59:25 GMT* From: helbig@astro.rug.nl (Phillip Helbig)B Subject: %SYSINIT-E, error mounting system device, status=0072832C. Message-ID: <91df4t$kd3$1@info.service.rug.nl>  : Any idea what could cause the message in the subject line?  G This is a 1 GB disk on a VAXstation 4000.  I've successfully booted off G of it several times.  Initially, I had a 2 GB disk, and also booted off F of it successfully several times, and still can.  I made an image copyH to the smaller disk, booted off of it (a few times), then installed some? layered products.  Everything went well.  Before installing the / defragger, I booted from the other disk and did D @SYS$SYSTEM:SETFILENOMOVE.COM to the smaller disk.  I then wanted toC boot from the smaller disk and install the defragger, but get this   strange error message.  E I can still boot off the other disk and thus access the smaller disk.    ------------------------------   Date: 15 Dec 2000 16:03:26 GMT* From: helbig@astro.rug.nl (Phillip Helbig)F Subject: Re: %SYSINIT-E, error mounting system device, status=0072832C. Message-ID: <91dfce$kgs$1@info.service.rug.nl>  B In article <91df4t$kd3$1@info.service.rug.nl>, helbig@astro.rug.nl (Phillip Helbig) writes:    < > Any idea what could cause the message in the subject line? > I > This is a 1 GB disk on a VAXstation 4000.  I've successfully booted off I > of it several times.  Initially, I had a 2 GB disk, and also booted off H > of it successfully several times, and still can.  I made an image copyJ > to the smaller disk, booted off of it (a few times), then installed someA > layered products.  Everything went well.  Before installing the 1 > defragger, I booted from the other disk and did F > @SYS$SYSTEM:SETFILENOMOVE.COM to the smaller disk.  I then wanted toE > boot from the smaller disk and install the defragger, but get this   > strange error message. > G > I can still boot off the other disk and thus access the smaller disk.   H I see that the disk remained mounted from another node in the cluster.  E I guess I should apply all those MOUNT patches before doing anything   else....   ------------------------------  % Date: Fri, 15 Dec 2000 17:35:45 +0100 > From: "Jean-Franois Marchal" <jean-francois.marchal@x9000.fr>F Subject: Re: %SYSINIT-E, error mounting system device, status=0072832C. Message-ID: <91dh0m$q9d$1@reader1.imaginet.fr>  C %MOUNT-F-DIFVOLMNT, different volume already mounted on this device    Cordialement Jean-Franois Marchal  X9000 - LYON (FR)   7 "Phillip Helbig" <helbig@astro.rug.nl> wrote in message ( news:91df4t$kd3$1@info.service.rug.nl...< > Any idea what could cause the message in the subject line? > I > This is a 1 GB disk on a VAXstation 4000.  I've successfully booted off I > of it several times.  Initially, I had a 2 GB disk, and also booted off H > of it successfully several times, and still can.  I made an image copyJ > to the smaller disk, booted off of it (a few times), then installed someA > layered products.  Everything went well.  Before installing the 1 > defragger, I booted from the other disk and did F > @SYS$SYSTEM:SETFILENOMOVE.COM to the smaller disk.  I then wanted toD > boot from the smaller disk and install the defragger, but get this > strange error message. > G > I can still boot off the other disk and thus access the smaller disk.    ------------------------------  # Date: Fri, 15 Dec 2000 09:17:17 GMT ( From: Terry Kennedy <terry@gate.tmk.com>9 Subject: Re: A wonderful message from an antivirus vendor ' Message-ID: <G5Lr4t.8Cv@spcuna.spc.edu>   " Paul Sture <paul@sture.ch> writes:A >>This is a multi-part message in MIME format. If you are reading ? >>this text, you may want to consider changing to a mail reader C >>or gateway that understands how to properly handle MIME multipart  >>messages.  >  > Are they kidding, or what?  I   Why? I use a MIME-aware mail user agent (PMDF MAIL) that doesn't appear F to have any of the Outlook/etc. security flaws of launching executableH attachments without my request or permission. Don't tar all of MIME just4 because some mail user agent authors are clueless...  4         Terry Kennedy             http://www.tmk.com5         terry@tmk.com             Jersey City, NJ USA    ------------------------------  % Date: Fri, 15 Dec 2000 12:28:30 +0100   From: Paul Sture <paul@sture.ch>9 Subject: Re: A wonderful message from an antivirus vendor + Message-ID: <VA.000001de.055052a9@sture.ch>   F In article <OF5252494A.E1E4D649-ON882569B6.00133829@foundation.com>,   wrote:# > From: Shane.F.Smith@Healthnet.com  > Newsgroups: comp.os.vms ; > Subject: Re: A wonderful message from an antivirus vendor ' > Date: Thu, 14 Dec 2000 19:31:06 -0800  >  > K > Heh, their text sucks, but their personal firewall product "ZoneAlarm" is ! > really good. Pity it's PC only.  > L I have to agree that their firewall is good. Indeed it blocked a connection 3 yesterday to a site I only visited once a week ago.   6 > Paul Sture <paul@sture.ch> on 12/14/2000 07:05:35 PM > ! > Please respond to paul@sture.ch  >  > To:   Info-VAX@Mvb.Saic.Com  > cc:  > 8 > Subject:  A wonderful message from an antivirus vendor >  >  > Security ahoy! > H > This message comes from a would be antivirus vendor, www.zonelabs.com. > B > >This is a multi-part message in MIME format. If you are reading@ > >this text, you may want to consider changing to a mail readerD > >or gateway that understands how to properly handle MIME multipart > >messages. >  > Are they kidding, or what? > ___  > Paul Sture
 > Switzerland  >    ___ 
 Paul Sture Switzerland    ------------------------------  % Date: Fri, 15 Dec 2000 12:28:28 +0100   From: Paul Sture <paul@sture.ch>9 Subject: Re: A wonderful message from an antivirus vendor + Message-ID: <VA.000001dd.05504bee@sture.ch>   < In article <G5Lr4t.8Cv@spcuna.spc.edu>, Terry Kennedy wrote: > Newsgroups: comp.os.vms * > From: Terry Kennedy <terry@gate.tmk.com>; > Subject: Re: A wonderful message from an antivirus vendor % > Date: Fri, 15 Dec 2000 09:17:17 GMT  > $ > Paul Sture <paul@sture.ch> writes:C > >>This is a multi-part message in MIME format. If you are reading A > >>this text, you may want to consider changing to a mail reader E > >>or gateway that understands how to properly handle MIME multipart 
 > >>messages.  > >N > > Are they kidding, or what? > K >   Why? I use a MIME-aware mail user agent (PMDF MAIL) that doesn't appeareH > to have any of the Outlook/etc. security flaws of launching executableJ > attachments without my request or permission. Don't tar all of MIME just6 > because some mail user agent authors are clueless... >  Fair point.r   ___u
 Paul Sture Switzerland    ------------------------------   Date: 15 Dec 2000 17:11:02 GMT2 From: hoffman@xdelta.zko.dec.nospam (Hoff Hoffman)& Subject: Re: ALTPRI in stead of SETPRI6 Message-ID: <91djb6$c21$1@mailint03.im.hou.compaq.com>  K In article <3A37D2A0.845FDAB2@momss.nl>, Cor Mom <cor.mom@momss.nl> writes:iI :Can anybody tell be in which version of OpenVMS Alpha and VAX the SETPRIu :privilege was obsolete?  L   Circa OpenVMS VAX V4.0 and later, and all versions of OpenVMS Alpha, IIRC.M   The SETPRI name is accepted as a synonym for ALTPRI, both via the routines TL   that parse the user text input of a privilege name and via programs using :   the $PRVDEF definitions.  (Do I win the kewpie doll? :-)  N  --------------------------- pure personal opinion ---------------------------L    Hoff (Stephen) Hoffman   OpenVMS Engineering   hoffman#xdelta.zko.dec.com   ------------------------------  % Date: Fri, 15 Dec 2000 11:26:06 +0000 - From: Tim Llewellyn <tim.llewellyn@bbc.co.uk>s* Subject: Re: Cluster advertising in the UK) Message-ID: <3A39FFCE.70F565C9@bbc.co.uk>i   Paul Repacholi wrote:e  1 > Tim Llewellyn <tim.llewellyn@bbc.co.uk> writes:e >CI > > I've also got a DHL package from Compaq to pickup. Lucky me, I wonder  > > if it will bounce. >rF > I had one yesterday. Wow, it was only the week before I ordered! Sue% > sorted that one well. Thanks folks.e >5 >c  ? Actually it was my OpenVMS and NT Integration for Dummies book.r     --6 Tim Llewellyn, OpenVMS Infrastructure, Remarcs Project0 MedAS at the BBC, Whiteladies Road, Bristol, UK.A Email tim.llewellyn@bbc.co.uk. Home tim.llewellyn@cableinet.co.uki  A I speak for myself only and my views in no way represent those ofo MedAS or the BBC.o   ------------------------------  # Date: Fri, 15 Dec 2000 09:18:43 GMT $ From: Dave Barlow <uzko@my-deja.com>% Subject: Re: INITIALIZE/MAXIMUM_FILESS) Message-ID: <91cnlg$g3b$1@nnrp1.deja.com>c  ) In article <918vde$fh7$1@nnrp1.deja.com>,D-   Alan E. Feldman <alan48@my-deja.com> wrote:t  > > > My feeling was that I needed to reformat these drives withC > > $INIT/HEADER=16000, to be safe. Based on above I may be safe toi leaveI > > them as they are?l > Yes.D > Why reduce the number of headers to 16000? You have plenty of freeG > space. And since most of INDEXF.SYS consists of headers, you probablydC > have 5x,000 headers, but only 8000 files. So INDEXF.SYS won't getyF > extended even once until you reach 5x,000. Actually, it's the number of! > file extents that matters here.e   Alan  / My post to you yesterday seems to have bounced.   E I concur with what you are saying. Why reformat to reduce the size ofe indexf.sys.i  C What started this was my colleague tidying up disk space and having B problems copying large files betwen disks. Our initial feeling wasE fragmentation based on indexf.sys apparently not extending, though wehE got on HEADERFULL errors. Based on this info it looks like some other G problem occurred. Just tried copying 100,000 block files around with nosC problems. The only loss is several hours of overtime for myself, ho  hum.   Dave Barlow)     Sent via Deja.comy http://www.deja.com/   ------------------------------  % Date: Fri, 15 Dec 2000 17:28:46 +0100-5 From: Philippe Marmillod <philippe.marmillod@epfl.ch> 9 Subject: Re: It's 20 years since I first logged onto VMS!:' Message-ID: <3A3A46BE.1D2D6E6C@epfl.ch>n   Paul Sture wrote:' > G > If I've got my dates right, on 8-DEC-1980, I logged into my first VMSi  5 In 1980 I was not so modern, my pdp-11 OS was RSX11-M>   Philippe Marmillod   ------------------------------  % Date: Fri, 15 Dec 2000 19:11:57 +1030a% From: Jeremy Begg <jeremy@vsm.com.au>F# Subject: Re: Logging OPCOM messagest* Message-ID: <3A3A6CFD.51B2EFEE@vsm.com.au>    cmccusker@lightbridge.com wrote: > I > This is probably a dumb question, but is there any way of logging OPCOMiG > messages from various different clusters to a single terminal withoutr? >    using something like Polycenter Console Manager? - Thanks.o  F Our "OperCon" product does *exactly* that (and a bit more) and is very0 reasonably priced.  You can read all about it at  #      http://www.vsm.com.au/opercon/s  = Please contact me via email (jeremy@vsm.com.au) to arrange anm evaluation.*   Regards,           Jeremy Begg-  =   +---------------------------------------------------------+i=   |            VSM Software Services Pty. Ltd.              |d=   |                 http://www.vsm.com.au/                  |i=   |       "OpenVMS Systems Management & Programming"        |s=   |---------------------------------------------------------|s=   | P.O.Box 402, Walkerville, |  E-Mail:  jeremy@vsm.com.au |A=   | South Australia 5081      |   Phone:  +61 8 83592155    |F=   |---------------------------|  Mobile:  0414 422 947      |A=   |  A.C.N. 068 409 156       |     FAX:  +61 8 82231777    |I=   +---------------------------------------------------------+P   ------------------------------  # Date: Fri, 15 Dec 2000 14:20:20 GMT  From: pasmith@ppg.com, Subject: Max Users) Message-ID: <91d9av$ss8$1@nnrp1.deja.com>   E I'm running VMS 7.1 on an Alpha 2100.  Is there a way I can tell whateB was the maximum number of users that has been logged in since last> reboot? I know I have had at least 70 users at one time and myF interactive login is set to 112 (The command 'Set login /inter=112' is= the last thing in a script run on reboot).  There has been annE occasional user indicating that they cannot connect.  These users areXC coming in through telnet.  I have run the command 'UCX SHOW SERVICE E TELNET/ALL' and it indicates that the telnet maximum users was 60 out  of a 70 limit.  Any suggestions will be helpful.     Sent via Deja.comu http://www.deja.com/   ------------------------------  # Date: Fri, 15 Dec 2000 15:10:32 GMTn* From: Alan E. Feldman <alan48@my-deja.com> Subject: Re: Max Users) Message-ID: <91dc92$vhi$1@nnrp1.deja.com>T  ) In article <91d9av$ss8$1@nnrp1.deja.com>,e   pasmith@ppg.com wrote:G > I'm running VMS 7.1 on an Alpha 2100.  Is there a way I can tell whatoD > was the maximum number of users that has been logged in since last@ > reboot? I know I have had at least 70 users at one time and myH > interactive login is set to 112 (The command 'Set login /inter=112' is? > the last thing in a script run on reboot).  There has been anIG > occasional user indicating that they cannot connect.  These users aretE > coming in through telnet.  I have run the command 'UCX SHOW SERVICE,G > TELNET/ALL' and it indicates that the telnet maximum users was 60 out. > of a 70 limit." > Any suggestions will be helpful. >   ) $ @SYS$UPDATE:AUTOGEN SAVPARAMS TESTFILESo    -or-a  ) $ @SYS$UPDATE:AUTOGEN SAVPARAMS GENPARAMSh  B then look at SYS$SYSTEM:AGEN$PARAMS.REPORT. This will tell you theE maximum observed number of processes. If you have been running out of/E process slots, AUTOGEN will suggest new values. Read about AUTOGEN in D the docs and use AUTOGEN to set new system parameters (including theD maximum number of processes) that will allow more processes. Keep an eye on performance though.  G Also look at the output of SHOW MEMORY/SLOTS. If you run out of processt> slots, then a new process cannot be created, regardless of SETE LOGINS/INTER=whatever. Remember that there is not always a one-to-one ? correspondence between number of users and number of processes.i  : *** What message appears when a user cannot "connect"? ***  C Also, are there any messages on the console, or in operator.log? IsaF there a message like "%SYSTEM-F-NOSLOT,  No PCB to create process"? If3 so, you need to increase MAXPROCESSCNT via AUTOGEN.    > Sent via Deja.com- > http://www.deja.com/ >a   --F NOTE: If you wish to e-mail me, please do NOT use the deja address. ItE is broken. Instead, use one of the addresses below, removing the longi wrong part first. Thanks.f   Disclaimer: JMHO Alan E. Feldman  &-)+ w: afeldman@gfigroup.ButItSaidItPrinted.comr5 h: alan48@dellnet.YouCantBelieveEverythingYouRead.come     Sent via Deja.com  http://www.deja.com/   ------------------------------  % Date: Fri, 15 Dec 2000 19:00:00 +10300% From: Jeremy Begg <jeremy@vsm.com.au>f3 Subject: Re: Mouse hang: AlphaStation 1000A VMS 6.2 * Message-ID: <3A3A6A30.214DC969@vsm.com.au>   Alan Frisbie wrote:  > 0 > We have an AlphaServer 1000A running VMS v6.2. > 5 > While tidying up the cables, I unplugged the mouse.Y6 > When I plugged it back in, it did not function until2 > I rebooted.   This was unfortunate, as this is a1 > production system and I had to wait until aftero8 > midnight to reboot.   (I'm just glad it didn't crash!) > > > Is there another way to recover from this without rebooting?  . Not sure if this would work, but you could try  (     $ @ SYS$MANAGER:DECW$STARTUP RESTART   Regards,           Jeremy Begg   =   +---------------------------------------------------------+y=   |            VSM Software Services Pty. Ltd.              |e=   |                 http://www.vsm.com.au/                  | =   |       "OpenVMS Systems Management & Programming"        |o=   |---------------------------------------------------------|a=   | P.O.Box 402, Walkerville, |  E-Mail:  jeremy@vsm.com.au |p=   | South Australia 5081      |   Phone:  +61 8 83592155    |s=   |---------------------------|  Mobile:  0414 422 947      |t=   |  A.C.N. 068 409 156       |     FAX:  +61 8 82231777    |'=   +---------------------------------------------------------+w   ------------------------------    Date: 15 Dec 2000 12:23:19 +0100* From: eplan@kapsch.net (Peter LANGSTOEGER)3 Subject: Re: Mouse hang: AlphaStation 1000A VMS 6.2 ( Message-ID: <3a39ff27@news.kapsch.co.at>  R In article <3A3A6A30.214DC969@vsm.com.au>, Jeremy Begg <jeremy@vsm.com.au> writes: >Alan Frisbie wrote: >> D1 >> We have an AlphaServer 1000A running VMS v6.2.a >> l6 >> While tidying up the cables, I unplugged the mouse.7 >> When I plugged it back in, it did not function untilf3 >> I rebooted.   This was unfortunate, as this is ak2 >> production system and I had to wait until after9 >> midnight to reboot.   (I'm just glad it didn't crash!)d >> b? >> Is there another way to recover from this without rebooting?c > / >Not sure if this would work, but you could trye >g) >    $ @ SYS$MANAGER:DECW$STARTUP RESTART   J This is strange, as we found here that when unplugging mouse or kbd on WNTH you DO have to reboot always to get them working again, and with VMS youJ DO fix - rarely happening - bugs with unplugging kbd or mouse because theyI always get perfectly reinitialized after replug. (One more VMS advantage)T  D Seems VMS and/or new hardware gets more and more PC crap quality :-(   -- u< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888c< <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------  % Date: Fri, 15 Dec 2000 15:09:07 +0100 6 From: "Martin P.J. Zinser" <zinser@sysdev.exchange.de> Subject: Re: Mozilla 0.61 Message-ID: <3A3A2603.3000308@sysdev.exchange.de>g   Hello!   Vance Haemmerle wrote:  ? > Anyone had a chance to run Mozilla 0.6 on VMS yet?  I had M18 D > running, but 0.6 just seems to hang before displaying any windows.  K For me it does work ok (even a bit better than the previous version). Theres( are still a number of features though...  J Did you check your process/systems settings versus the one recommended for 0.6?   					Greetings, Martin       -- nP Dr. Martin P.J. Zinser                                 zinser@sysdev.exchange.de Deutsche Boerse Systems AGL Neue Boersenstr. 1                                     Tel: +49 69 2101 5634L 60487 Frankfurt                                        FAX: +49 69 2101 3411P Germany                                                Private:  zinser@decus.de   ------------------------------  # Date: Fri, 15 Dec 2000 15:52:25 GMT0+ From: Craig A. Berry <calepine@my-deja.com>O Subject: Re: Mozilla 0.6) Message-ID: <91denl$1q9$1@nnrp1.deja.com>t  2 In article <3A395CC8.D87969B@toyvax.Tucson.AZ.US>,4   Vance Haemmerle <vance@toyvax.Tucson.AZ.US> wrote:? > Anyone had a chance to run Mozilla 0.6 on VMS yet?  I had M18s; > running, but 0.6 just seems to hang before displaying any  windows.  E The installation kit does not set file permissions appropriately.  Ifa= you go to sys$common:[mozilla] after the install and do a setl@ file/prot=(w:re) *.* you should be up and running (at least that worked for me).e     Sent via Deja.coms http://www.deja.com/   ------------------------------    Date: 15 Dec 2000 14:49:28 +0100* From: eplan@kapsch.net (Peter LANGSTOEGER) Subject: Re: Mozilla 0.6* Message-ID: <3a3a2168$1@news.kapsch.co.at>  f In article <3A395CC8.D87969B@toyvax.Tucson.AZ.US>, Vance Haemmerle <vance@toyvax.Tucson.AZ.US> writes:> >Anyone had a chance to run Mozilla 0.6 on VMS yet?  I had M18C >running, but 0.6 just seems to hang before displaying any windows.0  I Yup. It runs (as crappy as M12-M18 before). It's slow and consumes a lot.n  = *) Still no Autoproxy support (which is a show-stopper here).@I *) A lot of "we don't handle eBorderStyle_close yet... please fix me" msgaK *) "Helper Applications" table is still displayed empty, so you still can'tcB    see, what helper is already defined and which one is missing...N *) First Access to the news server (with creating the account) ended in ACCVIO3 	all other accesses couldn't connect to the server.DC *) I still didn't succeeded autoattaching a signature file to mailsy  
 And so on.  G It is the same problem as with MOZILLA on NT, but there is also MSIE...l   -- M< Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888 < <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------  % Date: Fri, 15 Dec 2000 16:19:09 +0100v- From: Jouk Jansen <joukj@hrem.stm.tudelft.nl>. Subject: Re: Mozilla 0.63 Message-ID: <3A3A447D.2997C592@hrem.stm.tudelft.nl>    Peter LANGSTOEGER wrote: > h > In article <3A395CC8.D87969B@toyvax.Tucson.AZ.US>, Vance Haemmerle <vance@toyvax.Tucson.AZ.US> writes:@ > >Anyone had a chance to run Mozilla 0.6 on VMS yet?  I had M18E > >running, but 0.6 just seems to hang before displaying any windows.o > K > Yup. It runs (as crappy as M12-M18 before). It's slow and consumes a lot.t > ? > *) Still no Autoproxy support (which is a show-stopper here).kK > *) A lot of "we don't handle eBorderStyle_close yet... please fix me" msgeM > *) "Helper Applications" table is still displayed empty, so you still can'tuD >    see, what helper is already defined and which one is missing...P > *) First Access to the news server (with creating the account) ended in ACCVIO< >         all other accesses couldn't connect to the server.E > *) I still didn't succeeded autoattaching a signature file to mailsi > D And  no equivalent of the -install option of netscape --> so it eats' up all available colours of my desktop.g                      JoukS --    > Ceterum censeo tertium millennium post Christum natum anno MMI incepturum esse-  P >------------------------------------------------------------------------------<  
   Jouk Jansenf 		 s   joukj@hrem.stm.tudelft.nl   E   Technische Universiteit Delft        tttttttttt  uu     uu  dddddddsF   Nationaal centrum voor HREM          tttttttttt  uu     uu  dd    ddG   Rotterdamseweg 137                       tt      uu     uu  dd     ddfG   2628 AL Delft                            tt      uu     uu  dd     ddaF   Nederland                                tt      uu     uu  dd    ddE   tel. 31-15-2781536                       tt       uuuuuuu   ddddddd   P >------------------------------------------------------------------------------<   ------------------------------    Date: 15 Dec 2000 17:20:24 +0100* From: eplan@kapsch.net (Peter LANGSTOEGER) Subject: Re: Mozilla 0.6* Message-ID: <3a3a44c8$1@news.kapsch.co.at>  c In article <3A3A447D.2997C592@hrem.stm.tudelft.nl>, Jouk Jansen <joukj@hrem.stm.tudelft.nl> writes:  >Peter LANGSTOEGER wrote:e >>  i >> In article <3A395CC8.D87969B@toyvax.Tucson.AZ.US>, Vance Haemmerle <vance@toyvax.Tucson.AZ.US> writes:-A >> >Anyone had a chance to run Mozilla 0.6 on VMS yet?  I had M18aF >> >running, but 0.6 just seems to hang before displaying any windows. >> gL >> Yup. It runs (as crappy as M12-M18 before). It's slow and consumes a lot. >> o@ >> *) Still no Autoproxy support (which is a show-stopper here).L >> *) A lot of "we don't handle eBorderStyle_close yet... please fix me" msgN >> *) "Helper Applications" table is still displayed empty, so you still can'tE >>    see, what helper is already defined and which one is missing... Q >> *) First Access to the news server (with creating the account) ended in ACCVIOs= >>         all other accesses couldn't connect to the server.)  # Just now, I had my first success...e  F >> *) I still didn't succeeded autoattaching a signature file to mails >>  E >And  no equivalent of the -install option of netscape --> so it eats-( >up all available colours of my desktop.  K Yup. But I don't like to run -install either. I simply want more colors ;-)i   --  < Peter "EPLAN" LANGSTOEGER           Tel.    +43 1 81111-2651; Network and OpenVMS system manager  Fax.    +43 1 81111-888r< <<< KAPSCH AG  Wagenseilgasse 1     E-mail  eplan@kapsch.netH A-1121 VIENNA  AUSTRIA              "I'm not a pessimist, I'm a realist"   ------------------------------  # Date: Fri, 15 Dec 2000 15:18:07 GMTl1 From: "Mark D. Jilson" <jilly@clarityconnect.com>V. Subject: Re: NI-style cluster upgrade question2 Message-ID: <3A3A3783.41562AF1@clarityconnect.com>   "J. Scott Greig" wrote:t >  > Hello alls > E > We currently have a 3 node LAN cluster (All at OpenVMS7.1 +patches)i" > with the following interconnects > = > Alpha 4100 - INTERCONNECT="NI", NISCS_MAX_PKTSZ=4468 (FDDI)-> > VAX7620   - INTERCONNECT="NICI", NISCS_MAX_PKTSZ=4468 (FDDI)7 > The VAX's twisted pair ethernet card is used for LAT.ZK > Alpha3000 - INTERCONNECT="NI",NISCS_MAX_PKTSZ=1498(Twisted pair ethernet)1 > 7 > The CI on the VAX is used for old HSC95 disk storage.j > ? > We are replacing the VAX7620/HSC95 with a DS20E/RA3000 setup.d > H > My question - the DS20E has a single twisted-pair ethernet port (which= > implies a NISCS_MAX_PKTSZ of 1498) - should the Alpha4100'sm- > NISCS_MAX_PKTSZ be changed to 1498 as well?e  F The packet size for each cluster sysapp connection is negotiated so noD don't change it.  The DSNlink article, [OpenVMS] A Discussion Of How3 VMScluster Members Communicate, discusses all this.l   > F > Unless cluster nodes negociate the maximum packet size, how does our > original configuration work? >  > TIA  >  > Scott    -- uD Jilly	- Working from Home in the Chemung River Valley - Lockwood, NY0 	- jilly@clarityconnect.com			- Brett Bodine fan. 	- Mark.Jilson@Compaq.com			- since 1975 or so, 	- http://www.jilly.baka.com               -   ------------------------------  # Date: Fri, 15 Dec 2000 18:11:46 GMTE- From: goathunter@goatley.com (Hunter Goatley)s) Subject: Re: OpenVMS Hobbyist TCP/IP Helpp/ Message-ID: <3a3a5eba.1443725@swen.process.com>s  , On Wed, 13 Dec 2000 19:22:12 -0500, JF Mezei% <jfmezei.spamnot@videotron.ca> wrote:    >Barry Skidmore wrote:F >> monitor).  I do not see a TCP/IP PAK listed on the Layered ProductsD >> PAK email I received after registering.  Does that mean I will be( >> unable to use FTP or TELNET from DCL? >t" >Look for a product key for "UCX".  # Or run MultiNet or TCPware instead.w  , http://www.process.com/openvms/hobbyist.html   ------------------------------  % Date: Fri, 15 Dec 2000 10:46:06 -0500-# From: "Mar" <kunszt@poczta.onet.pl>l Subject: Pathworks 32-& Message-ID: <91cp66$7h4$1@news.tpi.pl>   Hello,K My PC is runing Pathworks 32 under Windows NT and is assigned a Decnet nodea number.sL I want to copy a file from an OpenVms terminal to this PC (please note, thatL I have to do it using DCL Copy command, not an ftp or Decnet NFT utilities).) Is it possible and if so - how to do it ?tK I have an example taken from some FAQ's on how to do that but it concerns anJ Macintosh client. I've tried to modify and apply the solution to my PC but it doesn't worke Can anybody help me ?a Marian   ------------------------------  % Date: Fri, 15 Dec 2000 12:04:14 -0500 5 From: "Sue Skonetski" <susan.skonetski@compaq.nospam>n Subject: Press Release from ISDB6 Message-ID: <91dir6$c0e$1@mailint03.im.hou.compaq.com>  / Tattoo PW Brings WYSIWYG Label and Form Designs Capabilities to OpenVMS Systems   F Mansfield, MA December 13, 2000: Integrated Software Design (ISD), theC leader in enterprise printing and document management software, has-L announced the availability of Tattoo PW.  Tattoo PW provides users of ALPHAI OpenVMS with a full featured WYSIWYG label and form design interface, the K ability to include information from enterprise databases and the capabilityA< to print bar codes on multiple types and brands of printers.H Tattoo PW offers companies a vehicle with which to standardize label andI form design throughout their organization.  The versatile capabilities ofaK the software support printing from VMS, OpenVMS, Unix and NT on most laser, ( dot matrix or thermal transfer printers.L Tattoo PW combines the features of two powerful ISD software products into aF total solution for Alpha systems.  Tattoo ID provides flexible WindowsE desktop design features and Print Wheel provides powerful translation K capabilities to allow users to print various size characters, fonts, lines,nJ boxes, pictures and bar codes through their ALPHA OpenVMS systems.  TattooL PW allows users to include information on their labels/forms that comes fromK Enterprise Resource Planning Systems such as SAP, BAAN, Oracle ERP, JDE and K many more.  Tattoo PW lets users print every popular bar code type, such as/J Code 39, Code 128, Codabar, Interleaved 2 of 5, Straight 2 of 5, UPC, EAN,: MSI and POSTNET, without any additional hardware or SIMMs.F Tattoo PW makes it easy to print complex labels through ALPHA OpenVMS.@ First labels are designed through Tattoo IDs Windows interface.E Subsequently, Tattoo PW is used to select and name the file, select aa7 printer driver, configure fonts and translate the file. C Tattoo PW may be used to design and print a vast array of documentsnB including forms, invoices, purchase orders, manufacturing tickets,( documents, letters, and shipping papers.    E Integrated Software Design is the leading supplier of enterprise-widenI printing, tracking and business process automation solutions to over 4500 L U.S. and international businesses.  ISD combines off-the-shelf packages withJ consulting services and custom programming to produce an average of 3 to 6J month full ROI on Business Process Automation projects.  ISD's clients areK in the manufacturing, chemical, retail, educational, financial, health carelG and electronics industries.  ISD's software is installed in over 50% ofnI Fortune 500 companies.  For additional information about ISD products andT@ services, please visit their web site: www.isdweb.com or e-mail: info@isdweb.comi   ------------------------------  % Date: Fri, 15 Dec 2000 12:05:46 -0500r5 From: "Sue Skonetski" <susan.skonetski@compaq.nospam>O' Subject: Press Release from U.S. Designt6 Message-ID: <91diu2$c0m$1@mailint03.im.hou.compaq.com>  5 U.S. Design Announces Support for the Next Generationa0 5.25" (9.1 Gbyte) and DVD (9.4 Gbyte) Technology Under Compaq OpenVMS  J December 14, 2000 Columbia, MD - U.S. Design has announced support for theH new higher capacity optical drives, both 5.25" and DVD, under the CompaqE OpenVMS operating system.  The new hardware will provide both greateroK capacity and performance with all the advantages that continue forward withpK optical technology.  Applications will now be able to take advantage of the J greater capacity, near line capability, and lower cost per GB with the new technology.n  G The 5.25" (WORM or re-writable) solutions will provide standalone drive:C capacity of 9.1 GB per platter.  They are also available in jukeboxBH configurations that provide up to 500 platter capacity, or total storage0 available of 4.55 TB, in a very small footprint.  K The DVD-RAM (re-writeable) solutions will provide standalone drive capacity:I of 9.4 GB per platter.  They are also available in jukebox configurationsmG that provide up to 1840 platter capacity, or total storage available ofi% 17.296 TB, in a very small footprint.e  G Both solutions are available with the choice of the U.S. Design "driveruC based" Optical SuperSTOR software for OpenVMS or the "Java powered"rJ Universal SuperSTOR software.  Both software packages will allow the 5.25"J or DVD optical hardware to be connected to the host system and emulate allJ magnetic disk operations.  In other words, all applications that currentlyH support magnetic disk will operate with the new optical storage hardware? using either Optical SuperSTOR or Universal SuperSTOR software.y   - more -  K "The only problem that OpenVMS customers will have is deciding which of the/J powerful optical software packages from U.S. Design is the appropriate oneL for their application," stated Mr. Harry Garonzik, President and CEO of U.S.I Design Corporation.  "Optical SuperSTOR remains the industry standard for E OpenVMS customers with features such as Files 11 and cluster support.-E Universal SuperSTOR, our newest software, takes advantage of the Java E programming language to offer a solution that not only operates under-K OpenVMS, but also Tru64 Unix, Linux, and Windows NT/2000.  Now the customera# has a choice, stated Mr. Garonzik.   F Founded in 1978, U.S. Design is a leading supplier and manufacturer ofF high-end optical storage solutions.  The companys customer base spansK Fortune 500 companies throughout the world.  Clients such as Alabama Power,aI SMS, Georgia Pacific, Compaq and Bayer Chemical continue to purchase U.S. I Design solutions for applications such as data archival, imaging, backup,gK multimedia and data distribution. For the latest company information, visitn5 U.S. Design on the World Wide Web [www.usdesign.com].a  K Note to Editors: All product names mentioned herein are trademarks of theirb respective owners.   ------------------------------  # Date: Fri, 15 Dec 2000 13:46:04 GMTl$ From: mark@NOSPAMtechop.co.uk (Mark)) Subject: Strange access violation problems/ Message-ID: <3a3a0728.10220766@news.force9.net>g   Hi,o   Environment:C OpenVMS Alpha V7.1-1H2, C Compiler V6.2-003, C++ Compiler V6.2-035,u POSIX Threads.  D I am porting an application from UNIX to VMS and I am getting a veryC strange access violation error.  I can reproduce this problem undereB the debugger and the problem occurs as I step into a function. TheC function has already been called several times before successfully.   @ The only difference with the call that fails is that it is being= called from the context of a different thread.  The code doesV@ use mutexes to protect critical regions and it does work on UNIX< (Solaris & VMS).  The code in question is written in 'C' and= is always called from the same code (written in C++).  If thee; function is then called from the main thread then it always@ works.  < I also notice that examining C++ member variables within the: debugger can sometimes cause the debugger itself to become< unstable.  Is this a sign that there is problems in our code# or could it just be a debugger bug?t  D Has anyone any suggestions on how I can find why this is happenning?   TIAg
 Mark Williams    ------------------------------  % Date: Fri, 15 Dec 2000 16:11:36 -0000 - From: wspencer@ap.nospam.org (Warren Spencer)P- Subject: Re: Strange access violation probleme/ Message-ID: <t3kglo2hksmtf2@news.supernews.com>   : mark@NOSPAMtechop.co.uk (Mark) wrote in <3a3a0728.10220766 @news.force9.net>:   >Hi, > 
 >Environment: D >OpenVMS Alpha V7.1-1H2, C Compiler V6.2-003, C++ Compiler V6.2-035, >POSIX Threads.  > E >I am porting an application from UNIX to VMS and I am getting a very D >strange access violation error.  I can reproduce this problem under  
 << snip >>  C I've bumped into that type of failure before.  Usually points to a  K corrupted stack.  In my case I was writing too many characters to a string t9 buffer allocated on the stack (ie "automatic" variable). z   ws   -- M3 << What if there were no hypothetical questions? >>h   Warren Spencer Senior Software Engineer The Associated Press  ? ** My employer does not necessarily agree with my statements **[   ------------------------------    Date: 15 Dec 2000 02:19:47 -0500/ From: jordan@lisa.gemair.com (Jordan Henderson)m Subject: Re: Sun Cluster* Message-ID: <91cgmj$nhr$1@lisa.gemair.com>  ; In article <Jah_5.5358$1M.1139918@typhoon.ne.mediaone.net>,@3 Terry C. Shannon <terryshannon@mediaone.net> wrote:0 >  >[snip]D >3J >I'm not saying that DECpaq couldn't execute the balancing act better (I'dK >argue for faster, better, sooner) but it's easy for me to tell those folksnE >what to do with their money and their chips. (Of course, using 20-20eM >hindsight, a larger investment in Alpha--to hurry things up--might have been82 >money better spent than the CMGI investment, no?) >   @ Of course, hindsight is 20/20, but remember that Compaq receivedA FROM CMGI $2Bil at the time of the investment for the purchase ofp
 AltaVista.  < So, Compaq got to make $2Bil in 1999 and write off $1Bil in ; their investment in CMGI in 2000.  Seems like a pretty good  deal to me, in retrospect.   >r >[snip]e >n >h   -Jordan HendersonC jordan@greenapple.come   ------------------------------    Date: 15 Dec 2000 11:33:27 +0100G From: Jan Vorbrueggen <jan@mailhost.neuroinformatik.ruhr-uni-bochum.de>m Subject: Re: Sun ClusterH Message-ID: <y4zohy9doo.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>  , kuhrt@eisner.decus.org (Marty Kuhrt) writes:  A > It has been my experience that Digital/Compaq only announce new 0 > products when they are getting ready to ship.   M Except for the 6/833-based systems, which lead to the withdrawal, by SPEC, of-F the corresponding CPU2000 results and the issue of new ones with a newL availability date of January 2001. No explanation of what caused this delay.% One can only hope it happens again...E   	Jan   ------------------------------  % Date: Fri, 15 Dec 2000 12:09:40 -0500e- From: JF Mezei <jfmezei.spamnot@videotron.ca> $ Subject: SYS$CANCEL and CMU-IP stack, Message-ID: <3A3A5053.DC97B85F@videotron.ca>  M I have an old CMU-IP application which picks POP emails from a remote server.M  N The way I have implented the IO is that for a read, a SYS$SETIMR followed by a: QIO, both of which set their own event flag when complete.  N When the IO completes first, I SYS$CANTIM, and when the timer completes first,! I do a SYS$CANCEL on the channel.t  K In this situation where my program was expecting to read more data than the.K server was willing to give, the timeout situation cropped up. Subsequent toCI the SYS$CANCEL, *all* I/O would fail with the SYSTEM-F-ABORT (decimal 44)s error code.u  S I dug up the CMU-IP manual and there is no mention of SYS$CANCEL that I could find.l  G My experience is that SYS$CANCEL should only cancel all pending IOs butgM nothing else. Are there precedents where SYS$CANCEL would actually disconnectr/ a link or render a channel no longer unusable ?-  N Secondly, if the CMU IP stack does not support SYS$CANCEL, is there some otherN way to cause the outstanding io to be cancelled ? If I don't cancel it, when II next send a command to the remote host, the results will go to the buffer N specifified in the uncancelled IO instead of the buffer where the resulkts are4 expected (and thus everything will be out of phase).   Any comments, suggestions ?   J Does the TCPIP V5 stack allow SYS$CANCEL ? That program will eventually beK ported to TCPIP V5 so I don't want to spend too much time with CMU stuff ifd, the TCPIP stack supports the SYS$CANCEL etc.  W And why isn't there SS$_TIMEOUT on TCP stacks ? It woudl make programs so much cleaner.c   ------------------------------  # Date: Fri, 15 Dec 2000 09:22:45 GMT:$ From: Dave Barlow <uzko@my-deja.com>% Subject: Re: TCP Bug - MULTINET + UCXA) Message-ID: <91cnt2$g6f$1@nnrp1.deja.com>J  ) In article <91bth8$rv0$1@nnrp1.deja.com>,P*   moi_is_me <moi_is_me@my-deja.com> wrote:  E > Found what i consider a serious bug. It seems to be present in bothuE > UCX (TCP/IP Services) 5.0 and Multinet 4.2 on a ALPHA running 7.2-1   D Try posting to news://vmsnet.protocols.tcp-ip.multinet and .ucx, the3 developers read those groups and may be interested.    Dave Barlow.     Sent via Deja.com  http://www.deja.com/   ------------------------------  % Date: Fri, 15 Dec 2000 08:08:22 -0500e1 From: Jeff Schreiber <SCHREIBER@Eisner.DECUS.org>m% Subject: Re: TCP Bug - MULTINET + UCXl1 Message-ID: <01JXQ5IB79S200DW6Q@Eisner.DECUS.org>   = On Fri, 15 Dec 2000, moi_is_me <moi_is_me@my-deja.com> wrote:  >>E >Compare running totals in the Server and Client - More bytes arrivednD >at Server than what the Client is aware of,  ie when Cancelling theC >Write within the Client, neither UCX or Multinet let you know thatcE >some bytes were actually transmitted, the iosb.nbytes always says 0.n >f  F   Sounds a lot like some changes we just made recently.  The followingE   changes are being tested by the customer that had requested it, andp+   is not yet available for general release.N  D       - For $QIO(W) calls to INETDRIVER and EXOSDRIVER, if a call toE         SYS$CANCEL is made to cancel an i/o request is made, the IOSB E         would not reflect the count of data accepted into the kernel.n         This has been changed.  A         The previous behaviour can be restored with the followingM         command:  :                 $ MULTINET SET/KERNEL CANCEL_RETURN_BCNT 0  
 					-Jeff --/ Jeff Schreiber,            Process Software LLCc1 schreiber@mx.process.com   http://www.process.coms+      TCPware & MultiNet: Stronger than Everr   ------------------------------  % Date: Fri, 15 Dec 2000 18:26:23 +0000l% From: Alan Greig <a.greig@virgin.net>_) Subject: Re: Virtual I/O Cache and OracleE* Message-ID: <3A3A624F.59B8F89A@virgin.net>   Dan Sugalski wrote:h  L > This is generally a bad idea with Oracle. The DB accesses don't go throughE > the VIOC, so it's just memory wasted as far as Oracle is concerned.9 >i  S Is that true? We do have an Oracle 8 database on the system but it really just acts N as the back end for a very old archived data repository. I assumed Oracle diskL reads passed through the VIOC but it isn't heavily used enough for anyone toQ notice. Now Oracle DBMS certainly passes through VIOC (unless you set buffer size R greater than 35). The production systems also handles a lot of DCL, login/logoffs,O some Pathworks, web etc. I find a 200MB VIOC gets around an 80%  read hit rate.nS Overnight runs bump up the DBMS buffers to around 600MB per process (any higher and O DBMS frequently bugchecks). I have no intention of giving up my 200MB VIOC justVP because there's a lightly used Oracle app on the system! I can agree with Arne'sO statement that if all your machine does is run Oracle and nothing else then you-S might gain by turning off VIOC but I can't go along with "generally a bad idea witht Oracle". --
 Alan Greig   ------------------------------  % Date: Fri, 15 Dec 2000 10:40:44 +0000b4 From: John Laird <john@laird-towers.freeserve.co.uk> Subject: Re: Weird I/O problem8 Message-ID: <hjsj3tg94rf0pal0g7hs22d9pd5mdbr9v6@4ax.com>  4 On Thu, 14 Dec 2000 18:37:55 +0100, "Steven Driesen"& <NOSPAMsteven.driesen@tijd.com> wrote:  K >I have the following weird problem on a VAX/VMS system running OpenVMS 6.2o >CK >I have a program, written in Fortran, which runs every hour scheduled by ae
 >batch queue.wH >It does some file I/O when it runs. At start it creates a new temporary5 >file, then writes and reads from the file and at thenK >end closes the file and finally deletes it. This process is repeated everyt >hour with a new file. >rH >This is what sometimes happens for a couple of runs and then spontanous >disappears:2 >- the open of the file is allways ok (IOSTAT = 0)M >- a read from the same file fails with IOSTAT=29 (FILE NOT FOUND !!!). ThereT >is no close statement between> >  the open and the read. I checked, and the file does exist!! >n? >Another thing that also sometimes happens and then disappears:t >- open of file is oksM >- write to file fails with IOSTAT = 66 (RECORD OVERFLOW). I checked the data-/ >that is being written to the file and it looks0 >  OK for the format statement.e >i >s, >Especially the first problem is very weird. >lK >Does anybody know what might be happening. I am starting to think it a VMSt	 >problem.g  F Both of these problems would occur if your program was somehow closingF the file in question, *or* if the unit number had unexpectedly changedF (is it a variable or a hardcode value?).  The next READ or WRITE wouldE cause an implicit OPEN on FORnnn.DAT (nnn being your unit number), or E wherever the logical name  FORnnn pointed (if it existed), and if thehE file did not exist, the READ would fail, where the WRITE would createpH the file, but using default OPEN parameters which may well not specify aC record length long enough for your WRITE, hence the overflow error.d  E If you can eliminate the unintentional CLOSE, then look for your unitoC number variable being overwritten.  Try the debugger and SET WATCH.o   	John  --  
 John Laird   ------------------------------  % Date: Fri, 15 Dec 2000 16:01:12 -0000S- From: wspencer@ap.nospam.org (Warren Spencer)A Subject: Re: Weird I/O problem/ Message-ID: <t3kg28k2151e09@news.supernews.com>r  7 NOSPAMsteven.driesen@tijd.com (Steven Driesen) wrote inl- <2q8_5.765$yz3.27601@iguano.antw.online.be>: t   >Hi all, >e >pG >I have the following weird problem on a VAX/VMS system running OpenVMSl >6.2 r >oI >I have a program, written in Fortran, which runs every hour scheduled byu  
 << snip >>  L Does the program use native Fortran I/O calls or call $QIO?  If $QIO, is it  $QIO or $QIOW?   ws   --  3 << What if there were no hypothetical questions? >>e   Warren Spencer Senior Software Engineer The Associated Press  ? ** My employer does not necessarily agree with my statements **m   ------------------------------   End of INFO-VAX 2000.699 ************************