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

[bugzilla at redhat.com]

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



--- Comment #31 from Nithya Balachandran <nbalacha at redhat.com> ---
(In reply to Znamensky Pavel from comment #30)
> I'm sorry for the delay. But I can confirm that patched
> https://bugzilla.redhat.com/show_bug.cgi?id=1659432 version with disabled
> readdir-ahead still leaks with near the same speed as v4.1.

Then it is likely to be because the fuse client does not invalidate inodes.
Does your workload access a lot of files? The earlier statedump showed around 3
million inodes in memory. 

pool-name=inode_t
active-count=3050376  <---- Here
sizeof-type=160
padded-sizeof=256
size=780896256
shared-pool=0x5141cc8

Does the statedump without readdir ahead also show a large number of inodes?

https://review.gluster.org/#/c/glusterfs/+/19778/ has a fix to invalidate
inodes but is not targeted for release 5 as yet.

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