[Bugs] [Bug 1319992] RFE: Lease support for gluster

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1319992



--- Comment #46 from Poornima G <pgurusid at redhat.com> ---
TODOs in the descending order of priority:

1. Support of leases for EC and Sharded volumes.

2. Lease healing in case of replicated(-like) volumes

3. In case of network disconnect between gfapi client and gluster-server
process, we should replay leases in case if server flushes them. Currently its
been looked as part of Lease healing mentioned in (3). So we shall not have
this support in case of plane distributed volume.

4.  In case of replicated(-like) volumes, we should be able to filter out
duplicate LEASE_RECALL requests sent by them.

5. In case of NFS-Ganesha restart, we shall need support on glusterfs-server to
be able to support reclaim of lost state by NFS clients. But if we failover
only after gluster-server flushes the locks, we should be good.

6. If the client is actively flushing the I/Os, we could extend recall timer.

7. During rebalance/date-tiering, we should migrate lease. But at worst
scenario, we could recall lease before migrating data.

-- 
You are receiving this mail because:
You are on the CC list for the bug.
Unsubscribe from this bug https://bugzilla.redhat.com/token.cgi?t=C8uAH12mXc&a=cc_unsubscribe