HP3000-L Archives

November 2004, Week 1

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
"Dave Powell, MMfab" <[log in to unmask]>
Reply To:
Dave Powell, MMfab
Date:
Thu, 4 Nov 2004 12:56:36 -0800
Content-Type:
text/plain
Parts/Attachments:
text/plain (296 lines)
Very interesting.  Maybe its time to revisit a thread about backups from over
a year ago in which it was suggested that diagnostics, predictive & network be
stopped before a full backup, to get everything possible backed up.  Is this
why your system was stopping diagnostics?

Anyway I had some questions & comments I didn't think of till after that
thread died down, so here goes.

Why shut what down?  Its not quite enough to say kill something to get
everything backed up.  How vital are the files involved and how badly screwed
am I if I have to do a reload and they aren't on my full backup ????  I can
live with missing log files (small company, no Sarbanes-Oxley, etc).  Shutting
down the network would be a big issue here because we generate lots of
spoolfiles just before backup and they mostly print to network printers during
the backup.  Killing network printers then would be 'interesting'.

So, what do the experts say about what we really should stop?


After all this, if you are still going to shut down background service 'x', I
have a suggestion on how to do it without leaving it off far too long if your
backup hangs or dies for any reason.
For each service, instead of JUST shutting it down before the backup starts,
try:
STREAM XSTART
SETVAR XSTARTJOBNUM  HPLASTJOB
<issue the command to shut it down>

Then, after the backup is done:
<issue the command to restart that service>
ABORTJOB  XSTARTJOBNUM  (if you like to be tidy and aren't comfy with
redundant start commands).

The XSTART job just has the same old start commands, but  it logs on with a
very high inpri (but not hipri), so it will kick in immediately if the backup
job dies.  If the backup hangs, we have something else that will blow it off
after it has "hung around" too long, so the start job will get to do its thing
then too.


Dave Powell,    MMfab

----- Original Message -----
From: "John Testa" <[log in to unmask]>
To: <[log in to unmask]>
Sent: Thursday, November 04, 2004 11:31
Subject: Re: [HP3000-L] STMXL file in use when installing HP patch OSPM58A and
patch depe ndencies


> Beware of doing this when you have pending IOs on a tape drive!
>
> This is a known issue with MPE that I do not believe has a patch.  If you
> stop diagnostics (or predictive, I believe) when a drive has a pending IO
> (waiting), the drive may possibly hang.
>
> We had a process that was stopping and restarting diagnostics every night
> (some ancient thing someone else had written -- not sure why it was in
> there) and were having problems with tape drives not AVRing.  Turned out
> this is why.
>
> According to HP, the only reason to stop diagnostics manually is when you
> have to issue a netcontrol stop command (not sure about nscontrol, but
> don't think so).
>
> I realize this is a one-time thing, but just thought I'd throw this out
> there for people to consider.  It's a very odd problem, but it exists.
>
> John Testa
> SLPS Homewood
> 708-647-4346
> [log in to unmask]
>
>
>
>                       Denis St-Amand
>                       <[log in to unmask]        To:
[log in to unmask]
>                       moveit.com>                  cc:
>                       Sent by: HP-3000             Subject:  Re: [HP3000-L]
STMXL file in use when installing HP patch OSPM58A
>                       Systems Discussion            and patch
depe              ndencies
>                       <[log in to unmask]
>                       DU>
>
>
>                       11/04/2004 12:02 PM
>                       Please respond to
>                       Denis St-Amand
>
>
>
>
>
>
>  Pascal,
>
> Go into Posix Shell (XEQ SH.HPBIN.SYS -L), then use following commands:
>
>     Shell/iX>ps -ef | grep stm <= It will show you remaining stm processes
>     Shell/iX>kill -n pid   <= replace pid by diagmond process number
>     Shell/iX>ps -ef | grep stm  <= Should no longer show stm processese
>
> By the way, cstm daemons are automatically lauched at system restart (when
> you see message "successfully launched diagnostic monitor" message at
> console so using START NORECOVERY NOSYSSTART will not forbid those starting
> !
>
>
> Hope this will help!
>
> --
> ==============
> Denis St-Amand
> (remove removeit from email address)
> "Pascal Luyckx" <[log in to unmask]> wrote in message
> news:4u2dnQiSdPVyyBfcRVn-tA@fidnet.com...
> > Craig,
> >
> > This is also the first time I have this problem.
> > I did a START NORECOVERY NOSYSSTART and I also removed the Netbase shell.
> >
> > I had a problem with predictive (error 200) and this patch should fix
> this
> > error.
> >
> > Regards,
> >
> > Pascal
> >
> > -----Original Message-----
> > From: Craig Lalley [mailto:[log in to unmask]]
> > Sent: Thursday, November 04, 2004 5:12 PM
> > To: Pascal Luyckx
> > Subject: Re: STMXL file in use when installing HP patch OSPM58A and patch
> > depe ndencies
> >
> > Pascal,
> >
> > Believe it or not, I can honestly say, I have never had this problem.
> >
> > Are you doing a START NORECOVERY NOSYSSTART?
> >
> > What does the OSPM58A patch fix?
> >
> > -Craig
> >
> >
> > --- Pascal Luyckx <[log in to unmask]> wrote:
> >
> > > Also tried STMSHUT, but diagmond process keeps running.
> > >
> > > Pascal
> > >
> > > -----Original Message-----
> > > From: Craig Lalley [mailto:[log in to unmask]]
> > > Sent: Thursday, November 04, 2004 5:05 PM
> > > To: Pascal Luyckx
> > > Subject: Re: STMXL file in use when installing HP patch OSPM58A and
> > > patch depe ndencies
> > >
> > > STMSHUT.DIAG.SYS  ?
> > >
> > > -Craig
> > >
> > >
> > > --- Pascal Luyckx <[log in to unmask]> wrote:
> > >
> > > > Hi Craig,
> > > >
> > > > No, I did not start the network before running Patch/iX the second
> time.
> > > > I just rebooted the system, did a startspool and streams on the
> > > > console and launched patchix for the second time on the console.
> > > >
> > > > Any other ideas?
> > > >
> > > > Regards,
> > > >
> > > > Pascal
> > > >
> > > >
> > > > -----Original Message-----
> > > > From: Craig Lalley [mailto:[log in to unmask]]
> > > > Sent: Thursday, November 04, 2004 3:41 PM
> > > > To: Pascal Luyckx
> > > > Subject: Re: STMXL file in use when installing HP patch OSPM58A and
> > > > patch depe ndencies
> > > >
> > > > Pascal,
> > > >
> > > > Did you start the network before running Patch/iX a second time?
> > > >
> > > > You should only
> > > >
> > > > Logon on (console)
> > > >
> > > > :startspool lp
> > > > :streams 10
> > > >
> > > > :run patchix
> > > >
> > > > That should work
> > > >
> > > > -Craig
> > > >
> > > > --- Pascal Luyckx <[log in to unmask]> wrote:
> > > >
> > > > > Hi HP3000-L list,
> > > > >
> > > > > I downloaded patch OSPMX58A and its patch dependencies for MPE/iX
> > > > > 7.0 from the it resource center and installed it with Patch/iX. I
> > > > > updated the system from the CSLT tape and started the second phase
> > > > > of Patch/iX. Patch/iX then streams 3 installation jobs, but I
> > > > > encounter a problem with the installation of patch ODIMXQ2C. The
> > > > > following message is displayed on the
> > > > > console:
> > > > >
> > > > > *********************************************
> > > > > * STMXL.DIAG.SYS library is in use.         *
> > > > > * Critical diagnostics files will not be    *
> > > > > * able to be replaced if they are in use.   *
> > > > > * ODIMXQ2 C is unable to continue.          *
> > > > > *********************************************
> > > > >
> > > > > I check who is accessing the STMXL file :
> > > > >
> > > > > :listf stmxl.diag.sys,9
> > > > > ********************
> > > > > FILE: STMXL.DIAG.SYS
> > > > > 1 Accessor(O:1,P:1,L:0,W:0,R:1),Share
> > > > > system  MANAGER.SYS                P:1,L:0,W:1,R:1      LDEV:
> > > > > #P52    (/usr/sbin/stm/uut/bin/sys/diagmond)
> > > > > ACCESS: R,X,L-gmulti      REC#: 0       FNUM: 22
> > > > > LOCKS: none
> > > > >
> > > > > I try to stop the diagmond process by executing STMSHUT.DIAG.SYS,
> > > > > but the diagmond process keeps on running. I then try to kill the
> > > > > process with the ABORTPROC command, but then my session is aborted.
> > > > >
> > > > > Is there any other way of killing the diagmond process?
> > > > >
> > > > > Regards,
> > > > >
> > > > > Pascal Luyckx
> > > > > Cheops technology nv
> > > > >
> > > > > * To join/leave the list, search archives, change list settings, *
> > > > > * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
> > > > >
> > > >
> > > >
> > > >
> > > >
> > > > __________________________________
> > > > Do you Yahoo!?
> > > > Check out the new Yahoo! Front Page.
> > > > www.yahoo.com
> > > >
> > > > * To join/leave the list, search archives, change list settings, *
> > > > * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
> > > >
> > >
> > >
> > >
> > >
> > > __________________________________
> > > Do you Yahoo!?
> > > Check out the new Yahoo! Front Page.
> > > www.yahoo.com
> > >
> > > * To join/leave the list, search archives, change list settings, *
> > > * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
> > >
> >
> >
> > __________________________________________________
> > Do You Yahoo!?
> > Tired of spam?  Yahoo! Mail has the best spam protection around
> > http://mail.yahoo.com
> >
> > * To join/leave the list, search archives, change list settings, *
> > * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
> >
>
> * To join/leave the list, search archives, change list settings, *
> * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
>
> * To join/leave the list, search archives, change list settings, *
> * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
>

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

ATOM RSS1 RSS2