Gregory Farnum
|
6d2e4c9b7b
|
Merge pull request #19973 from liewegas/wip-peering-fast-dispatch
osd: fast dispatch of peering events and pg_map + osd sharded wq refactor
Reviewed-by: Greg Farnum <gfarnum@redhat.com>
|
2018-04-06 11:48:11 -07:00 |
|
Sage Weil
|
3b3c32f643
|
qa/suites/rados/singleton/all/recovery_preemption: whitelist SLOW_OPS
Recovery and peering can be slow enough with all the logging enabled to
trigger a slow ops warning.
Signed-off-by: Sage Weil <sage@redhat.com>
|
2018-04-04 16:24:31 -05:00 |
|
Sage Weil
|
29a885c915
|
qa/suites/rados/singleton/all/recovery_preemption: make test more reliable
A 30 second run did only 7000 ops, which means ~50 log entires per pg...
not enough to trigger backfill.
Signed-off-by: Sage Weil <sage@redhat.com>
|
2018-04-04 08:26:57 -05:00 |
|
Sage Weil
|
76d84ac194
|
qa/suites/rados/singleton/all/recover-preemption: handle slow starting osd
The OSD may not be marked up yet; set the config via the admin
socket.
Signed-off-by: Sage Weil <sage@redhat.com>
|
2017-10-02 07:20:57 -05:00 |
|
Sage Weil
|
d7b29acb19
|
qa/suites/rados/singleton/all/recovery-preemption: add test
This mirrors what I was testing locally.
Signed-off-by: Sage Weil <sage@redhat.com>
|
2017-09-28 13:48:14 -04:00 |
|