[Bugs] [Bug 1506487] glusterfs / glusterfsd processes may not start properly upon reboot/restart

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1506487



--- Comment #18 from Milind Changire <mchangir at redhat.com> ---
(In reply to Sylvain from comment #17)
> I restarted services two times, then rebooted one time. All my processes
> stayed alive during these 3 tests; so .... I guess you fixed my issue :)
> Thank you very much for the time you spent for me.

I'm glad the effort paid off for you.

> 
> Just a couple of thing I'm wondering:
> 
> - From what I understand, my problem was due to the size of my cluster; is
> there somewhere a document dedicated to large cluster deployment, with
> guidelines, parameters to tune (just like this one), etc ?

Upstream Gluster documentation is available at:
http://docs.gluster.org/en/latest/

All possible cluster-wide and volume-wide options can be listed using:
# gluster volume get all all

You can then set them using the command format:
# gluster volume set <vol> <option> <value

and retrieve a specific value using:
# gluster volume get <vol> <option>

or
# gluster volume get <vol> all

Help for these options can be listed by:
# gluster volume help

> 
> - Is it equivalent to run "gluster volume set mysto transport.listen-backlog
> 100" ?

Yes, that's the way you'd set the option for all the bricks.
But for glusterd, you'll have to edit the /etc/glusterfs/glusterd.vol file
manually. There's no command-line for setting options for glusterd ... other
than the actual command-line in the service file.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
You are the assignee for the bug.