9.28.2012

syslogd cores

 HP-UX syslog crashes with a core file in /var/adm/syslog.   file /var/adm/syslog/core:  ELF-32 core file - IA64 from 'syslogd' - received SIGXFSZ
 
if none of your files in /var/adm/syslog are over 2GB, then change the LOG_SIZE to NOLIMIT in /etc/default/syslogd  like this:
 
query value:
ch_rc -l -p LOG_SIZE /etc/default/syslogd
 
set to NOLIMIT:
ch_rc -a -p LOG_SIZE=NOLOMIT /etc/default/syslogd
 
what happened is that OVO (HPOM) process opcle had syslog.log open and it hit 2GB so when syslog was recycled, the opcle process didn't release and the system still thinks that syslog.log is over 2GB.  That's the theory anyway.  I guess if that's the case you probably could just restart OVO/HPOM and that might fix it.