From: Samuel Just Date: Wed, 22 Jun 2016 16:22:45 +0000 (-0700) Subject: 11429: wait_for_clean between restarting the osds and starting the bench X-Git-Tag: v11.1.1~58^2^2~145^2 X-Git-Url: http://git-server-git.apps.pok.os.sepia.ceph.com/?a=commitdiff_plain;h=c4ad6b3b339cc52be04a218fab31af21617146f7;p=ceph.git 11429: wait_for_clean between restarting the osds and starting the bench ceph.restart now marks the osds down, so the objects are actually being created while slowest of the osds boots. That causes a ton of 1 byte objects to be created in a degraded state and causes the cleanup to take a long time. Also, reduce length of bench since it's only being used to ensure the osds came up correctly. Signed-off-by: Samuel Just --- diff --git a/suites/rados/singleton-nomsgr/all/11429.yaml b/suites/rados/singleton-nomsgr/all/11429.yaml index 755c70f2aaa..51526e51b7b 100644 --- a/suites/rados/singleton-nomsgr/all/11429.yaml +++ b/suites/rados/singleton-nomsgr/all/11429.yaml @@ -114,14 +114,14 @@ tasks: - ceph.restart: [mon.a, mon.b, mon.c, osd.0, osd.1, osd.2] - ceph_manager.wait_for_clean: null - ceph.restart: [osd.0, osd.1, osd.2] + - ceph_manager.wait_for_clean: null - exec: osd.0: - sudo grep -c 'unable to peek at' /var/log/ceph/ceph-osd.0.log - radosbench: clients: [client.0] - time: 30 + time: 5 size: 1 - - ceph_manager.wait_for_clean: null - ceph.restart: [osd.0, osd.1, osd.2] - ceph_manager.wait_for_clean: null - print: '**** done verifying final upgrade'