[Gluster-users] How do we resolve the "cannot self-heal" problem?

andro at skybroadband.com.ph andro at skybroadband.com.ph
Thu Aug 11 00:10:58 UTC 2011

Hello from the Philippines!

Apparently we're the first company here to use Gluster and Enomaly, and
we've snagged on the same well-documented race condition lock-up. I read
that Gluster 3.2.2 does not suffer from this problem, so we upgraded.

Catch: We lost 1 of 4 storage nodes a week before the upgrade.
Issue: It seems that bringing that 1 guy back online is interfering with
self-heal on a number of files (VMs, really).

How do we resolve the "cannot self-heal" problem?

Andro Mauricio

More information about the Gluster-users mailing list