[lustre-discuss] lustre 2.5.3 ost not draining
strosahl at jlab.org
Sun Jul 12 15:02:25 PDT 2015
I just thought of a possible solution... to my over all problem.
What if I put every directory in a pool that avoided the 97% full osts, and the osts on the bad oss. Then set everyone back to read write and turn the qos_rr back to 17%. Now no new writes will land on the full / bad osts, but they will still be open for writes. I can then drain down the osts using lfs_migrate and the space should go down accordingly.
It is a hack, but it solves my two major problems (the full osts not draining properly and the qos round robin being broken).
----- Original Message -----
From: "Sean Brisbane" <sean.brisbane at physics.ox.ac.uk>
To: "Kurt Strosahl" <strosahl at jlab.org>
Cc: "Shawn Hall" <shawn.hall at nag.com>, lustre-discuss at lists.lustre.org
Sent: Sunday, July 12, 2015 5:13:07 AM
Subject: RE: [lustre-discuss] lustre 2.5.3 ost not draining
I was following the recommendation that the OST be active to allow the deletion to happen, hence the reactivation followed by mount/unmount is necessary to trigger the cleanup. The OST was therefore active during the mount/unmount.
More information about the lustre-discuss