https://tracker.ceph.com/https://tracker.ceph.com/favicon.ico2011-02-03T10:24:57ZCeph Ceph - Bug #759: osd: pgs spend a long time peering when marking osds outhttps://tracker.ceph.com/issues/759?journal_id=23432011-02-03T10:24:57ZSage Weilsage@newdream.net
<ul><li><strong>Status</strong> changed from <i>New</i> to <i>In Progress</i></li></ul> Ceph - Bug #759: osd: pgs spend a long time peering when marking osds outhttps://tracker.ceph.com/issues/759?journal_id=23502011-02-03T12:02:44ZSage Weilsage@newdream.net
<ul></ul><p>this appears to be scrubbing related:</p>
<p>- we get a new osdmap. handle_osd_map tries to pause the op threadpool.<br />- a long running scrub op takes forever to complete<br />- handle_osd_map finally continues.</p>
<p>during that whole time the main dispatch thread is blocked up, and peering gets backed up as a result.</p> Ceph - Bug #759: osd: pgs spend a long time peering when marking osds outhttps://tracker.ceph.com/issues/759?journal_id=23512011-02-03T12:14:08ZSage Weilsage@newdream.net
<ul><li><strong>Assignee</strong> changed from <i>Sage Weil</i> to <i>Samuel Just</i></li></ul><p>the replica scrub needs to go in a different work queue (not op_wq). scrub_wq, or something else that's assigned to the disk threadpool disk_tp.</p> Ceph - Bug #759: osd: pgs spend a long time peering when marking osds outhttps://tracker.ceph.com/issues/759?journal_id=24002011-02-09T10:18:36ZSamuel Justsjust@redhat.com
<ul></ul><p>1a01e5ee1b88a217547873296e0371858be13f37 merged in a branch moving replica scrubbing to rep_scrub_wq with a new non-osdop message for initiating a replica scrub. Scrub still blocks in the disk_tp while waiting for replicas to scrub, though, working on that now.</p> Ceph - Bug #759: osd: pgs spend a long time peering when marking osds outhttps://tracker.ceph.com/issues/759?journal_id=24012011-02-09T10:21:30ZSage Weilsage@newdream.net
<ul><li><strong>Status</strong> changed from <i>In Progress</i> to <i>Resolved</i></li></ul>