[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: 19990809: LDM 5.0.8 shuts down for no reason!

On Mon, 9 Aug 1999, Unidata Support wrote:

> 
> ------- Forwarded Message
> 
> >To: address@hidden,
> >From: Gilbert Sebenste <address@hidden>
> >Subject: LDM 5.0.8 shuts down for no reason!
> >Organization: NIU
> >Keywords: 199908091422.IAA16410 LDM 5.0.8
> 
> Hello all,
> 
> A perplexing problem has reared it's ugly head. I am running LDM 5.0.8
> under Linux 5.2, kernel 2.036, on a dual PII 450 machine.
> 
> When I feed from UIUC, after a short time, the LDM just kills itself for
> no apparent reason.
> 
> Look:
> 
> Aug 09 14:12:33 weather rpc.ldmd[9386]: Starting Up (built: Aug  4 1999 
> 14:15:41) 
> Aug 09 14:12:33 weather data2[9392]: run_requester: Starting Up: 
> data2.atmos.uiuc.edu 
> Aug 09 14:12:33 weather data2[9392]: run_requester: 19990809141219.158 
> TS_ENDT {{WMO,  ".*"},{MCIDAS,  ".*"},{FSL2,  ".*"}} 
> Aug 09 14:12:33 weather lightning[9393]: run_requester: Starting Up: 
> lightning.alden.com 
> Aug 09 14:12:33 weather striker[9394]: run_requester: Starting Up: 
> striker.atmos.albany.edu 
> Aug 09 14:12:33 weather 162.113.112.7[9391]: run_requester: 
> 19990809141018.456 TS_ENDT {{WSI,  ".*\.gif"},{WSI,  ".*\.m.*g"}} 
> Aug 09 14:12:33 weather data2[9392]: FEEDME(data2.atmos.uiuc.edu): OK 
> Aug 09 14:12:33 weather pqexpire[9387]: Starting Up 
> Aug 09 14:12:33 weather pqbinstats[9388]: Starting Up (9386) 
> Aug 09 14:12:34 weather pqact[9389]: Starting Up 
> Aug 09 14:12:34 weather pqsurf[9390]: Starting Up (9386) 
> Aug 09 14:12:34 weather striker[9394]: run_requester: 19990809140637.776 
> TS_ENDT {{NLDN,  ".*"}} 
> Aug 09 14:12:34 weather pqact[9395]: Starting Up 
> Aug 09 14:12:35 weather striker[9394]: FEEDME(striker.atmos.albany.edu): OK 
> Aug 09 14:12:35 weather localhost[9399]: Connection from localhost 
> Aug 09 14:12:35 weather localhost[9399]: Connection reset by peer 
> Aug 09 14:12:35 weather localhost[9399]: Exiting 
> Aug 09 14:12:35 weather lightning[9393]: run_requester: 19990809140547.333 
> TS_ENDT {{DIFAX,  ".*"}} 
> Aug 09 14:12:36 weather lightning[9393]: FEEDME(lightning.alden.com): OK 
> Aug 09 14:12:42 weather pqexpire[9387]: > Recycled  55283.743 kb/hr (  
> 9882.868 prods per hour) 
> Aug 09 14:12:46 weather weather3[9560]: Connection from 
> weather3.admin.niu.edu 
> Aug 09 14:12:46 weather weather3(feed)[9560]: Starting Up: 19990809141219.158 
> TS_ENDT {{FSL2|UNIDATA,  ".*"}} 
> Aug 09 14:12:46 weather weather3(feed)[9560]: topo:  weather3.admin.niu.edu 
> FSL2|UNIDATA 
> Aug 09 14:12:46 weather www[9565]: Connection from www.umsmed.edu 
> Aug 09 14:12:46 weather www(feed)[9565]: Starting Up: 19990809141157.974 
> TS_ENDT {{DDPLUS,  ".*"}} 
> Aug 09 14:12:46 weather www(feed)[9565]: topo:  www.umsmed.edu DDPLUS 
> Aug 09 14:12:47 weather hpccsun[9567]: Connection from hpccsun.unl.edu 
> Aug 09 14:12:47 weather hpccsun(feed)[9567]: Starting Up: 19990809141157.974 
> TS_ENDT {{DDPLUS,  ".*"}} 
> Aug 09 14:12:47 weather hpccsun(feed)[9567]: topo:  hpccsun.unl.edu DDPLUS 
> Aug 09 14:13:56 weather rpc.ldmd[9386]: child 9389 terminated by signal 11 


Gilbert,

With your experience as a LDM maintainer you should be able to start
diagnosing these type of problems.  The previous line states that :

 child 9389 terminated by signal 11

Then the main ldmd process rpc.ldmd[9386] dies.  I would suspect that
process 9389 would be the culprit, that's pqact.  Did you make any
changes?  If not there could be an error in the pqact with products coming
from UIUC.  There might be different products in the UIUC stream verses
the U-W stream.  Could test this by commenting out the exec pqact entry in
the ldmd.conf line.  I let you solve the rest.

Robb...




> Aug 09 14:13:56 weather rpc.ldmd[9386]: Killing (SIGINT) process group 
> Aug 09 14:13:56 weather rpc.ldmd[9386]: Interrupt 
> Aug 09 14:13:56 weather rpc.ldmd[9386]: Exiting 
> Aug 09 14:13:56 weather hpccsun(feed)[9567]: Interrupt 
> Aug 09 14:13:56 weather hpccsun(feed)[9567]: Exiting 
> Aug 09 14:13:56 weather rpc.ldmd[9386]: Terminating process group 
> Aug 09 14:13:56 weather data2[9392]: Interrupt 
> Aug 09 14:13:56 weather data2[9392]: Exiting 
> Aug 09 14:13:56 weather lightning[9393]: Interrupt 
> Aug 09 14:13:56 weather lightning[9393]: Exiting 
> Aug 09 14:13:56 weather striker[9394]: Interrupt 
> Aug 09 14:13:56 weather striker[9394]: Exiting 
> Aug 09 14:13:56 weather weather3(feed)[9560]: Interrupt 
> Aug 09 14:13:56 weather weather3(feed)[9560]: Exiting 
> Aug 09 14:13:56 weather www(feed)[9565]: Interrupt 
> Aug 09 14:13:56 weather www(feed)[9565]: Exiting 
> Aug 09 14:13:56 weather pqexpire[9387]: Interrupt 
> Aug 09 14:13:56 weather pqexpire[9387]: Exiting 
> Aug 09 14:13:56 weather pqexpire[9387]: > Up since:      19990809141233.508 
> Aug 09 14:13:56 weather pqexpire[9387]: > Queue usage (bytes):250003448 
> Aug 09 14:13:56 weather pqexpire[9387]: >          (nregions):   39493 
> Aug 09 14:13:56 weather pqexpire[9387]: > nbytes recycle:      2136600 ( 
> 55283.743 kb/hr) 
> Aug 09 14:13:56 weather pqexpire[9387]: > nprods deleted:          373 (  
> 9882.868 per hour) 
> Aug 09 14:13:56 weather pqexpire[9387]: > First deleted: 19990809130516.704 
> Aug 09 14:13:56 weather pqexpire[9387]: > Last  deleted: 19990809130732.576 
> Aug 09 14:13:56 weather pqbinstats[9388]: Interrupt 
> Aug 09 14:13:56 weather pqbinstats[9388]: Exiting 
> Aug 09 14:13:56 weather pqsurf[9390]: Exiting 
> Aug 09 14:13:56 weather pqsurf[9390]:   Queue usage (bytes): 4435592 
> Aug 09 14:13:56 weather pqsurf[9390]:            (nregions):   25195 
> Aug 09 14:13:56 weather pqsurf[9390]: Number of products 20 
> Aug 09 14:13:56 weather pqsurf[9390]: Number of observations 50 
> Aug 09 14:13:56 weather pqsurf[9390]: Number of dups 8 
> Aug 09 14:13:56 weather pqact[9395]: Interrupt 
> Aug 09 14:13:56 weather pqact[9395]: Exiting 
> 
> When I feed off of any other site, I do not have this problem. I don't
> know what the problem is...I am now getting the exact same feeds from U-W
> Madison, and it's stable. I tried deleting the queues, but to no avil.
> After about 20 seconds, it just dies.
> 
> ?????
> 
> Gilbert
> 
> *******************************************************************************
> Gilbert Sebenste                                                     ********
> Internet: address@hidden    (My opinions only!)                     ******
> Staff Meteorologist, Northern Illinois University                      ****
> Work phone: 815-753-5492                                                ***
> *******************************************************************************
>  
> 
> 
> 
> ------- End of Forwarded Message
> 

===============================================================================
Robb Kambic                                Unidata Program Center
Software Engineer III                      Univ. Corp for Atmospheric Research
address@hidden             WWW: http://www.unidata.ucar.edu/
===============================================================================