[Bugs] [Bug 1287517] Memory leak in glusterd

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1287517

Atin Mukherjee <amukherj at redhat.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |amukherj at redhat.com,
                   |                            |kostas.makedos at gmail.com
              Flags|                            |needinfo?(kostas.makedos at gm
                   |                            |ail.com)



--- Comment #1 from Atin Mukherjee <amukherj at redhat.com> ---
(In reply to kmak from comment #0)
> Created attachment 1101371 [details]
> valgrind report
> 
> Description of problem:
> I have a deployment with 3 nodes and 6 bricks.
> Tried to change glusterfs version but problem persists.
What command did you run here? Can you please take statedump of glusterd
process by executing 'kill -USR1 $(pidof glusterd)' and it would generate the
file in /var/run/gluster. Also please attach *glusterd.log & cmd_history.log.

> glusterd process memory consumption is increasing constantly.
> Using top to check its RSS values.
> On a fresh installation started with 27M.
> After 3 hours it was 71M.
> Today is 515M.
> Valgrind file attaches shows some possibly lost records.
> Tried to echo 2 at /proc/sys/vm/drop_caches but no effect in memory
> consumption.
> Consumption seems to increase if self-heal daemon is closed.
> 
> Version-Release number of selected component (if applicable):
> Tried with 3.6.2, 3.7.5 and 3.7.6
> 
> How reproducible:
> Install a 3 node cluster.
> glusterd in storage master is increasing memory consumption.
> 
> Additional info:

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