A coredump is being created by one of our applications on Solaris server and occupying entire space on the mount, thereby bringing down the application.
While we try to identify the root cause, i tried to limit to limit the size of the core dump.
Executed below command in shell and also updated in .profile file:
ulimit -c 500
When i do a ulimit -a, i see the coredumpsize as 500.
But the application again created a core of size 600 MB. So, the setting is not working.
When i do a plimit <pid>, the coredumpsize is shown as 'unlimited'.
Can someone please help me on restricting the size of coredump.
How is the application run? Daemon? script? what. Does the process that starts the app have privilege? Does the app process itself have privileges?
Answer those and you can fix the problem.
Meanwhile, also consider getting the application 'fixed'. By definition, a production app does not coredump. Because it has been correctly and fully tested, for one good reason. We had a process that dumped core for several years because the manager would not let anyone address the problem. Literally.
Thanks for the reply.
The application is run as a script. I'm not well versed with Solaris OS, so i'm not able to answer your other two questions. Can you please guide me.
Yes we are considering fixing the application, meanwhile thinking of this option to limit coredump size as a workaround