fix(entrypoint): dump cores on persistent location

[vishnuitta]

This is continuation to #282

In this PR, on enabling coredumps, coredumps gets collected onto hostpath "/var/openebs/sparse" which is persistent storage.

Without any directory set in core pattern, core dumps gets collected to the CWD of the shell from which child process got started.

Tested with command like: docker run --env ENABLE_COREDUMP=1 --mount 'type=bind,src=/home/vitta/Desktop,dst=/var/openebs/sparse' openebs/cstor-pool:ulimit5

Signed-off-by: Vitta vitta@mayadata.io

[mynktl]

We should return non-zero value if sparse directory doesn't exist.

[pawanpraka1]

$ help exit
exit: exit [n]
Exit the shell.

Exits the shell with a status of N.  If N is omitted, the exit status
is that of the last command executed.

[vishnuitta]

it returns error 2 in this case 'no such file or directory'

vitta@vitta-laptop:~$ sudo docker run --env ENABLE_COREDUMP=1 --mount 'type=bind,src=/home/vitta/Desktop,dst=/home/core' openebs/cstor-pool:ulimit5
Enabling coredumps
at call_exit..
exit code: 0
reference:  /usr/local/bin/entrypoint-poolimage.sh
/usr/local/bin/entrypoint-poolimage.sh: 25: cd: can't cd to /var/openebs/sparse
vitta@vitta-laptop:~$ echo $?
2

[pawanpraka1]

looks good.

[vishnuitta]

Environment variable ENABLE_COREDUMP has been added to cStor pool container to control whether cores need to be dumped in case of process crashes. By default, dumping cores will be disabled.
Make sure this environment variable is not enabled if mountPoint of SparseDir has been changed in CAS Templates.