[Bugs] [Bug 1214644] Upcall: Migrate state during rebalance/tiering
bugzilla at redhat.com
bugzilla at redhat.com
Tue Apr 28 02:27:01 UTC 2015
https://bugzilla.redhat.com/show_bug.cgi?id=1214644
--- Comment #2 from Soumya Koduri <skoduri at redhat.com> ---
Another approach suggested by Shyam -
The solution outline is for the rebalance process to migrate the locks, with
some additional coordination with the locks/lease/upcall xlators.
The problem however is _mapping_ all of the lock information across the 2
different storage node brick processes. (i.e client_t information)
>>>>>>>>
More details in the below threads.
http://thread.gmane.org/gmane.comp.file-systems.gluster.devel/9079
http://www.gluster.org/pipermail/gluster-devel/2014-December/043284.html
We may primarily need to check if we can make client_t mapping uniform across
the bricks and the issues with it.
--
You are receiving this mail because:
You are the assignee for the bug.
More information about the Bugs
mailing list