[Gluster-users] mount.glusterfs able to return errors directly to caller rather than default log

[Vijay Bellur]

On Thu, Apr 28, 2016 at 12:12 PM, Saravanakumar Arumugam
<sarumuga at redhat.com> wrote:
>
> On 04/28/2016 07:26 PM, Scott Creeley wrote:
>>
>> Is there a way to set an option to return errors directly to the caller?
>> Looking at the man-page doesn't appear so...but wondering if there is some
>> trick to accomplish this or how hard that would be to implement.
>>
>> For example, current behavior:
>>
>> 1.  mount.glusterfs fails - returns exit code  (exit status 1) and the
>> true errors are logged to log-file parameter (or default)
>>
>>          [2016-04-27 19:25:53.429657] E
>> [socket.c:2332:socket_connect_finish] 0-glusterfs: connection to
>> 192.168.121.222:24007 failed (Connection timed out)
>>          [2016-04-27 19:25:53.429733] E
>> [glusterfsd-mgmt.c:1819:mgmt_rpc_notify] 0-glusterfsd-mgmt: failed to
>> connect with remote-host: 192.168.121.222 (Transport endpoint is not
>> connected)
>>
>> 2.  user or program needs to access the log file to find the root cause of
>> the issue
>>
>>
>> wondering how hard it would be to return the errors directly to the caller
>> with an additional option, something like  verbose or -v
>
> seems work in progress, check this :
> http://review.gluster.org/#/c/11469/
>

It would be nice to have the log message that indicates the reason for
failure more grep friendly. That way users and scripts can easily pick
it up from the verbose output.

Regards,
Vijay