CentOS 5.5 question - failing app

SRTMCSE

We run our own in house software for trading (HFT firm) and our developers are in the process of converting our windows apps to Linux (CentOS). They're claiming that the issue is NOT their code but the process is sporadically failing, but their own logging is not saying why. I'm 99.98% positive it's their code but I need to know what is the general method to troubleshoot failing processes/applications on a Linux server? This way I not only have some proof it's not the server just being difficult but also know because I'm a linux novice and have only started supporting Linux in the last 6 months due to this new initiative.

Forsaken_GA

well the first thing is to check and find out if it's happening sporadically, or if it's at regular intervals (ie, does it occur at the same time every day or within the same time interval?)

If you can nail that down, then it's a little easier. Just strace the process before it's scheduled to die and it should give you some idea at the system level of why the process is dying, ie, is it receiving a SIG(TERM|KILL|whatever)? Is it running out of memory? Is it segfaulting?

dmesg and syslog might also be able to provide some info if the vendor's logging provisions aren't