[Bugs] [Bug 1623107] FUSE client's memory leak

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1623107

Nithya Balachandran <nbalacha at redhat.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |nbalacha at redhat.com
              Flags|                            |needinfo?(kompastver at gmail.
                   |                            |com)



--- Comment #24 from Nithya Balachandran <nbalacha at redhat.com> ---
(In reply to Yan from comment #23)
> The issue was observed in v5.0.1 as well. Anyone can help with a resolution?

Hi,

Please provide the following on the v5.0.1 client:

1. gluster volume info (I want to check if anything has changed since
comment#13)
2. Statedump of the client process which is consuming a lot of memory (one that
is not controlled by cgroups)
 - make sure /var/run/gluster exists on the client machine
 - kill -SIGUSR1 <gluster-fuse-process-pid>
Do this twice with an interval of maybe 30mins in between.

Please send me the glusterdump.* files generated in /var/run/gluster.


After collecting the above, if not already done, disable readdir-ahead on the
volume.

gluster v set <volname> performance.readdir-ahead off

and see if the memory continues rising.

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