https://tracker.ceph.com/https://tracker.ceph.com/favicon.ico2016-12-03T07:51:59ZCeph Ceph - Backport #18132: hammer: ReplicatedBackend::build_push_op: add a second config to limit omap entries/chunk independently of object datahttps://tracker.ceph.com/issues/18132?journal_id=824752016-12-03T07:51:59ZAlexey Sheplyakovasheplyakov@mirantis.com
<ul><li><strong>Copied from</strong> <i><a class="issue tracker-2 status-3 priority-4 priority-default closed" href="/issues/16128">Feature #16128</a>: ReplicatedBackend::build_push_op: add a second config to limit omap entries/chunk independently of object data</i> added</li></ul> Ceph - Backport #18132: hammer: ReplicatedBackend::build_push_op: add a second config to limit omap entries/chunk independently of object datahttps://tracker.ceph.com/issues/18132?journal_id=824802016-12-03T12:52:29ZNathan Cutlerncutler@suse.cz
<ul><li><strong>Status</strong> changed from <i>Pending Backport</i> to <i>New</i></li></ul> Ceph - Backport #18132: hammer: ReplicatedBackend::build_push_op: add a second config to limit omap entries/chunk independently of object datahttps://tracker.ceph.com/issues/18132?journal_id=824812016-12-03T12:52:48ZNathan Cutlerncutler@suse.cz
<ul><li><strong>Description</strong> updated (<a title="View differences" href="/journals/82481/diff?detail_id=79536">diff</a>)</li></ul><a name="original-description"></a>
<h3 >original description<a href="#original-description" class="wiki-anchor">¶</a></h3>
<p>In ceph deployments with large numbers of objects (typically generated by use of radosgw for object storage)<br />it's quite possible for OSDs recovering data to hit their suicide timeout and shutdown because of the number<br />of objects each was trying to recover in a single chunk between heartbeats.</p>
<p><a class="external" href="https://bugs.launchpad.net/ubuntu/+source/ceph/+bug/1628750/comments/0">https://bugs.launchpad.net/ubuntu/+source/ceph/+bug/1628750/comments/0</a></p>
<p>build_push_op assumes 8MB of omap entries is about as much work to read as 8MB of object data. This is probably false.<br />Add a config (osd_recovery_max_omap_entries_per_chunk ?) with a sane default (50k?) and change build_push_op to use it.</p> Ceph - Backport #18132: hammer: ReplicatedBackend::build_push_op: add a second config to limit omap entries/chunk independently of object datahttps://tracker.ceph.com/issues/18132?journal_id=828612016-12-09T13:45:06ZNathan Cutlerncutler@suse.cz
<ul><li><strong>Description</strong> updated (<a title="View differences" href="/journals/82861/diff?detail_id=79884">diff</a>)</li><li><strong>Status</strong> changed from <i>New</i> to <i>In Progress</i></li><li><strong>Assignee</strong> changed from <i>Alexey Sheplyakov</i> to <i>Nathan Cutler</i></li></ul> Ceph - Backport #18132: hammer: ReplicatedBackend::build_push_op: add a second config to limit omap entries/chunk independently of object datahttps://tracker.ceph.com/issues/18132?journal_id=837942017-01-05T18:39:42ZNathan Cutlerncutler@suse.cz
<ul><li><strong>Status</strong> changed from <i>In Progress</i> to <i>Resolved</i></li><li><strong>Target version</strong> set to <i>v0.94.10</i></li></ul>