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

bugzilla at redhat.com bugzilla at redhat.com
Thu Oct 26 09:22:20 UTC 2017


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

Sylvain <srn at nexatech.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
              Flags|needinfo?(srn at nexatech.fr)  |



--- Comment #4 from Sylvain <srn at nexatech.fr> ---
(In reply to Milind Changire from comment #3)
> 1. Are you using Gluster for Virtual Machine image storage ?
No, i'm using the "regular" RPM; 

> 2. Are Gluster nodes running on bare metal ?
For this particular test cluster, no; i'm using 30 VMs distributed on 10
physical hosts (3 VMs per host); config of each VMs : 4 CPUs & 8 GB ram.

> 3. Are you using arbiter bricks ?
No; I though it was linked when I was on a replica 2 volume, but now that I am
on replica 3 (where arbiter shouldn't be needed I think) I have the same
problem.


> If possible, could you share TRACE level logs of one of the brick that dies
> after a reboot ?

Here you go.
Bad news if that it didn't change anything for the failing glusterfsd process;
I think the problem happens too early.

You can find attached:
- mnt-dd1-brick_NOK.log (the one which failed);
- mnt-dd2-brick_OK.log (the one which succeeded);

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


More information about the Bugs mailing list