[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference csc32::consolemanager

Title:POLYCENTER Console Manager
Notice:Kits, Scans, Docs on CSC32:: as PCM$KITS:,PCM$DOCS:, PCM$SCANS:
Moderator:CSC32::BUTTERWORTH
Created:Thu Aug 06 1992
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:1541
Total number of notes:6564

422.0. "Control process in loop. CONSOLE MONITOR hangs" by EEMELI::OJUSSILA (Olli Jussila, OMS Finland) Tue Oct 04 1994 07:45

    
    
    We have had more and more problems with PCM V1.5, ECO 1 on
    VMS V6.1 VAX.  
    
    Our PCM has 50 systems. 49 connected with LAT and one pseudo system.
    Pseudo devices is to PCM system itself to get reply messages.
    
    We use PSW V2.2's SNS$FEED_PCM.C to feed some PSW messages
    to PCM.
    
    Lately PCM has hang almost every day. Events get passed but 
    CONSOLE MONITOR command hang. Typically on of the controller
    process is in tight CPU loop. It is not the same controller
    process at each time. Console shutdown can't be used. STOP/ID
    is used to restart PCM. before startup we have to logout
    terminal server ports.
    
    Last night controller wasn't in CPU loop when CONSOLE MONITOR 
    coundn't start. CONSOLE SHUTDOWN stopped all PCM processes 
    except on controller process. It was in LEF state. 
    Process was able to stop with FORCE_EXIT and it also
    terminated LAT connections.
    
    What kind of infformation you need if (when) this happends
    again? Would ypu like to log in to check it.
    
    -Olli
    
T.RTitleUserPersonal
Name
DateLines
422.1OPG::PHILIPAnd through the square window...Tue Oct 04 1994 13:1710
Ollie,

  A full system dump would probably help us out. Just copy it to OPG::

  Out of interest, do you have any empty log files for the nodes you are 
  managing?

Cheers,
Phil

422.2Yes, there is empty logfilesEEMELI::OJUSSILAOlli Jussila, OMS FinlandTue Oct 04 1994 15:1436
    
>  Out of interest, do you have any empty log files for the nodes you are 
>  managing?
    
    Yes there is lot of systems with empty logfiles. Like HSC's.
    We do daily archive.
    
    $ dir console$logfiles:*.log/sel=size=max=0/size=all
    
    Directory CONSOLE$ROOT:[LOG]
    
    ELOSSA.LOG;1               0/0
    HSC03A.LOG;1               0/0
    HSC04A.LOG;1               0/0
    HSC05A.LOG;1               0/0
    HSC14E.LOG;1               0/0
    HSC15A.LOG;1               0/0
    HSC15E.LOG;1               0/0
    HSKSW1.LOG;1               0/0
    KEMMUX.LOG;1               0/0
    KEMTRA.LOG;1               0/0
    LUKKO.LOG;1                0/0
    MANCUS.LOG;1               0/0
    MUKANA.LOG;1               0/0
    NIITTY.LOG;1               0/0
    PROFET.LOG;1               0/0
    SERGEI.LOG;1               0/0
    STRTRA.LOG;1               0/0
    SULKU.LOG;1                0/0
    TALO.LOG;1                 0/0
    TILA.LOG;1                 0/0
    TURVA.LOG;1                0/0
    VARMA.LOG;1                0/0
    
    Total of 22 files, 0/0 blocks.
    
422.3OPG::PHILIPAnd through the square window...Tue Oct 04 1994 15:5315
Ah Ha,

  Do you by chance happen to be getting a line of data with an event on it 
  when you daemon starts to loop?

  If this is the case, then we know about the problem and have a fix. The
  problem is essentially, the first log line of data in the file does not 
  have a timestamp and so if you have an event on that line the code loops 
  forever trying to find a non-existant timestamp!!

  Would you care to try a fixed image to see if your problem goes away? It 
  will mean deleting your log files first.

Cheers,
Phil
422.4Just send a mailEEMELI::OJUSSILAOlli Jussila, OMS FinlandWed Oct 05 1994 11:4714
    
>  Do you by chance happen to be getting a line of data with an event on it 
>  when you daemon starts to loop?
    
    It might be true. Very often it has happend when one system has shut
    down and system manager tries to connect to console. And
    SHUTDOWN is one event,
    
    We are very willing to test new image.
    Just send mail to me or to Veli K�rkk� (EEMELI::KORKKO) and we can
    start to use it right away.
    
    -Olli
    
422.5OPG::SIMONWed Oct 05 1994 12:434
Olli,
     sent the location of a patched imge by mail.

Cheers Simon....
422.6No problems so farVELI::KORKKOGeneral nuisance...Wed Oct 05 1994 23:506
        I have implemented the new image and so far no problems. Of
        course it is TOO EARLY to say anything definitive. But so far no
        hangs and I have been able to perform a few reconfigs without
        getting PCM into "confused" state!
        
        Regards, Veli
422.7EEMELI::OJUSSILAOlli Jussila, OMS FinlandThu Oct 27 1994 22:167
    
    It seems that new image has fixed our problem. 
    When next ECO kit is ready. Would be nice to have it before
    couple of important VCS-> PCM V1.5 upgrade.
    
    	-Olli
    
422.8OPG::PHILIPAnd through the square window...Fri Oct 28 1994 10:2514
Ollie,

  The next ECO will probably be a MUP and wont be ready until
  sometime around the end of November. This is due to the fact
  that we have a couple of problems outstanding that we need to
  fix along with Simon being on holiday until then! (He builds
  the ULTRIX and OSF/1 kits).

  I see no reason why you should let your customers have the
  patched image we have already given you.

Cheers,
Phil