Getting the Core Dumps for Content Server

gsingh4

Hi,

We have Content Server 16.2 running and seems like either the llservworker process or the llserver process is being restarted randomly. How can we enable the core dumps to pin point the issue?

Core dump at the OS level is alreay enabled, but we don't see any core logs generated at all.

ulimit -a | grep -i core

core file
size          (blocks, -c) 0

Are we missing something?

appuq

Usually the way OT traps that is not using OS Tools. It is helpful for you to look at the thread starts usually when a thread starts you will have thread1.log which is the running thread, it will have almost a lot of technical info as well as finally there should be a line that resembles LLI ready to recive requests and then on it is usually request info. A clue is if you see a lot of thread1_1.out thread1_2 out upto about 10 in a span of say 10 minutes then I would suspect the the threads are restarting. and so on which are actually thread logs saved on restarts or change of log file, if you see a lot of them with very minimal time stamps perhaps 5 mins apart chances are the Llserver is not keeping up with C++/C components and dying and re-starting at which point OT has to be contacted. If you are not running thread logs with INFO please do that. I am saying this from memory about 10 years back but when threads re-started abnormally I think one would see info like LLI thread shutdown received blah blah something like that…

traces happen when an unhandled request happens so if you see trace files that means Oscript code is crashing and it will not usually amount in thread death.

It is better to involve OT support because when it coredumps oscripters like me or no one else can looks at what the problem is it is at the binary(C++) level where we do not have access.