[Gluster-devel] Geo-replication changelog cleanup

[Strahil Nikolov]

Hello All,

As I got no answer in Slack, I hope to find more info here.
Do we have a geo-rep changelog cleanup mechanism ? Some users report running out of inodes on the bricks due to large ammount of changelogs and I couldn't find a clue if Gluster has such mechanism or not.
I know that currently gsyncd is using the /var/lib/misc/gluster directory structure to store processed changelogs list in a tar, yet I don't see a mechanism to combine processed changelogs into a single file inside the brick.
If we don't have such mechanism, I guess I can open an RFE to discuss it as due to never cleaned up changelogs, gluster might run out of inodes and affect production workloads.
Thanks in advance.
Best Regards,Strahil Nikolov
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.gluster.org/pipermail/gluster-devel/attachments/20210922/a6b11524/attachment.html>