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

[LDM #YGA-963195]: Future LDM feature request

Gilbert,

> Found another bug. According to Tyler Allison:
> 
> -----Begin forwarded message----
> 
> From: Tyler Allison <address@hidden>
> To: Gilbert Sebenste <address@hidden>
> Subject: found a bug in ldm
> 
> This version of LDM will crash the pqact if it receives a signal 25.
> signal 25 is "file to big" (eg: 2G or more)

Actually, every version of the LDM will crash if a pqact(1) process receives a 
SIGXFSZ.

> It shouldn't crash the entire LDM process in that event..it should
> simply fail on the specific pqact function and move on.
> 
> That's what happened on crunch that caused the metar outage.  It tried
> to write to the river.txt file which was 2G. So as soon as a river
> product came in it crashed the LDM service with a signal 25 (see logs)

What was the specific pqact(1) action being executed? PIPE? EXEC? FILE?...

> -Tyler
> 
> ----End forwarded message----
> 
> The offending log file entry...one of them:
> 
> Jan 01 03:18:44 crunch01 pqact[19455] NOTE: Starting from insertion-time
> 2011-01-01 00:10:00.351522 UTC
> Jan 01 03:18:44 crunch01 10.1.1.15[19461] NOTE: Upstream LDM-6 on
> 10.1.1.15 is willing to be an alternate feeder
> Jan 01 03:18:44 crunch01 10.1.1.15[19458] NOTE: Upstream LDM-6 on
> 10.1.1.15 is willing to be an alternate feeder
> Jan 01 03:18:45 crunch01 ldmd[19453] NOTE: child 19454 terminated by
> signal 25: /home/ldm/bin/pqact -f NIMAGE|WMO|LIGHTNING|EXP|FSL2
> /home/ldm/etc/pqact.conf
> Jan 01 03:18:45 crunch01 ldmd[19453] NOTE: Killing (SIGTERM) process group
> Jan 01 03:18:45 crunch01 pqact[19455] NOTE: Exiting
> Jan 01 03:18:45 crunch01 10.1.1.15[19461] NOTE: Exiting
> Jan 01 03:18:45 crunch01 10.1.1.12[19460] NOTE: Exiting
> Jan 01 03:18:45 crunch01 10.1.1.15[19458] NOTE: Exiting
> Jan 01 03:18:45 crunch01 10.1.1.12[19457] NOTE: Exiting
> Jan 01 03:18:45 crunch01 rtstats[19456] NOTE: Exiting
> Jan 01 03:18:45 crunch01 ldmd[19453] NOTE: Exiting
> Jan 01 03:18:45 crunch01 ldmd[19453] NOTE: Terminating process group
> Jan 01 03:22:31 crunch01 pqcheck[20624] NOTE: Starting Up (20522)
> 
> 
> ----
> *******************************************************************************
> Gilbert Sebenste                                                     ********
> (My opinions only!)                                                  ******
> Staff Meteorologist, Northern Illinois University                      ****
> E-mail: address@hidden                                  ***
> web: http://weather.admin.niu.edu                                      **
> *******************************************************************************

Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: YGA-963195
Department: Support LDM
Priority: Normal
Status: Closed