]> git.apps.os.sepia.ceph.com Git - ceph.git/commit
projects / ceph.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: d944e7e) | patch
lost_unfound: do peer after, until wait_for_clean propagates last_epoch_started
authorSage Weil <sage@newdream.net>
Sun, 26 Feb 2012 05:35:31 +0000 (21:35 -0800)
committerSage Weil <sage@newdream.net>
Sun, 26 Feb 2012 05:35:31 +0000 (21:35 -0800)
commit6295578f165db7e07e2c205aebb19aa00b184844
treec5e749393b6ac0f87b60bf9fd65afbbe64adb094tree | snapshot
parentd944e7ee1e4ba9bc8e495d6e6c9ad2b653be4562commit | diff
lost_unfound: do peer after, until wait_for_clean propagates last_epoch_started

The peer task does wait_for_clean, and then lost_unfound immediately marks
something down.  But the PGs become clean before the replica last_epoch_started
is moved forward in time, which means they block waiting for the now down
OSD.  Needlessly.

Until we fix this, just do the peer test after.
suites/regression/singleton/all/lost-unfound.yaml diff | blob | history
Unnamed repository; edit this file 'description' to name the repository.