Core file under /tmp

Hi Guys

Recently we have a issue where the core file under /tmp grows and there by /tmp becomes 100% full in our HP unix server.

Any ideas or suggestions about how we can resolve this.

Find out why there are core dumps in /tmp . Consider enlarging the /tmp filesystem if you need to debug those core dumps..

If you know the crashing program, and you do not want to send its core file to the application vendor, then you can limit or disable core dump size in the environment, e.g. in a start shell:

#!/bin/sh
ulimit -c 2048 # limit core dump size to 1MB, enough for "file"
exec /path/to/the/real/program "$@"

The real program inherits the limit set by the start shell.

I suggest you take a look to coreadm command

With that command I centralice all cores into a single filesystem /var/adm/crash that if it becomes full it has no impact :wink:

I hope it helps

1 Like