[Bugs] [Bug 1732961] shard file with different gfid in different subvolume causing VM to pause on stale file

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1732961

Olaf Buitelaar <olaf.buitelaar at gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
              Flags|needinfo?(olaf.buitelaar at gm |
                   |ail.com)                    |



--- Comment #6 from Olaf Buitelaar <olaf.buitelaar at gmail.com> ---
Dear Susant, 

Sorry i completely missed your comment. I don't really think it's possible to
update the IO pattern of the App's. It's essentially VM's (via KVM, managed by
oVirt) The vm's are running various apps via docker, probably the most IO
intensive are Mariadb databases.
I've recently updated to gluster 6.4 and are currently running gluster 6.5, and
the issue still exists. however i did notice a few things;
on VM's where this issue did occur before, i ran sparsify
(https://ovirt.org/develop/release-management/features/virt/virt-sparsify.html)
after that i've not observed the issue since.
On 2 VM's the issue still exists, these VM's do have snapshots, and didn't
allow me to run sparsify. They currently only seem to pause when running
xfs_fsr. Except once, while a brick-replace was in action (as we're migrating
to a new environment).

@Nithya I've attached the logs earlier send in the mailing list. If you do
require new logs let me know which exactly you're interested in, since it
shouldn't be hard to reproduce the issue here. I don't think the mount logs
will add much, since the VM's use libgfapi

Thanks Olaf

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