[Labs-l] Ongoing Labs outage, must manually fix glusterfs

[Ryan Lane]

Gluster is turning out to be a major problem. After virt3 died the
other day, the glusterfs volume somehow went crazy and has basically
worked itself into an unrecoverable state.

The data is still good, though, so I'll be manually moving it into a
new volume, and bringing things back up. I'll also be upgrading
glusterfs to 3.3, as recommended in #gluster, as apparently 3.2 has
issues with self-healing large files (like VM images).

- Ryan