RepoMirrors/ceph
1
0
Fork 0
mirror of https://github.com/ceph/ceph synced 2024-12-27 22:13:28 +00:00
Code Issues Packages Projects Releases Wiki Activity
5,557 Commits 211 Branches 394 Tags 1.4 GiB
171eadd6a3
Commit Graph

93 Commits

Author SHA1 Message Date
Josh Durgin
c12b3513a7 Merge pull request #1003 from athanatos/wip-15655 
ceph_manager: test [test-]reweight-by-(utilization|pg)

Reviewed-by: Josh Durgin <jdurgin@redhat.com>
 2016-07-22 14:46:21 -07:00
Samuel Just
19854c095b ec_lost_unfound: set min_size to 2 
We changed the default to k+1 instead of k.  Adjust test to compensate.

Fixes: http://tracker.ceph.com/issues/16416
Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-07-05 14:05:12 -07:00
Josh Durgin
274d79ade3 tasks/ceph_manager: make utility_task cluster-aware 
Signed-off-by: Josh Durgin <jdurgin@redhat.com>
 2016-05-09 11:52:00 -07:00
Josh Durgin
256ebf8a12 tasks: move find_remote to util, rename and add helper 
This is a useful for any cluster-aware task.

Signed-off-by: Josh Durgin <jdurgin@redhat.com>
 2016-05-09 11:51:59 -07:00
Josh Durgin
bba323834b tasks/ceph_manager: make Thrasher cluster-aware 
Just a few spots need to know to lookup only osds in this cluster, or
prefix a filename with the cluster. Use CephManager.find_remote() to
avoid a bunch of repetition and look only in the intended cluster.

Signed-off-by: Josh Durgin <jdurgin@redhat.com>
 2016-05-09 11:51:59 -07:00
Josh Durgin
713e717fda tasks/ceph_manager: make mount_osd_data() cluster-aware 
Use a cluster-specific mount point, and address osds by full role,
rather than just id, in the ctx.disk_config structures.

Signed-off-by: Josh Durgin <jdurgin@redhat.com>
 2016-05-09 11:51:59 -07:00
Josh Durgin
524e6d7a5e tasks/ceph_manager: add cluster param to write_conf() 
Only used by cephfs right now, so don't bother changing callers.

Signed-off-by: Josh Durgin <jdurgin@redhat.com>
 2016-05-09 11:51:59 -07:00
Josh Durgin
ff49deb6f0 tasks/ceph_manager: simplify remote lookup, and make it cluster aware 
Re-implement find_remote() using ctx.cluster.only() with a matcher
function that includes the manager's cluster, and use it instead of
miscellaneous ctx.cluster.only() calls elsewhere.

Signed-off-by: Josh Durgin <jdurgin@redhat.com>
 2016-05-09 11:51:59 -07:00
Josh Durgin
141c73d399 tasks/ceph_manager: parameterize CephManager with cluster 
Add --cluster arguments, pass cluster to get_daemon() and
iter_daemons_of_role, replace 'ceph' with cluster in paths, and use
ctx.ceph[cluster] instead of ctx.ceph.

Signed-off-by: Josh Durgin <jdurgin@redhat.com>
 2016-05-09 11:51:58 -07:00
Samuel Just
482a12f348 ceph_manager: test [test-]reweight-by-(utilization|pg) 
Fixes: http://tracker.ceph.com/issues/15655
Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-04-29 15:11:19 -07:00
David Zafman
8da6e97bd4 CephManager: Wait 1 second for pool creation to get far enough along 
Fixes: http://tracker.ceph.com/issues/15673

Signed-off-by: David Zafman <dzafman@redhat.com>
 2016-04-29 14:29:13 -07:00
David Zafman
a595651c54 CephManager: Maximum 2 minutes for raw cluster commands 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2016-04-29 09:43:39 -07:00
David Zafman
447bf873a8 thrasher: Add noscrub_toggle_delay and flip the noscrub osd flags 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2016-04-29 09:43:39 -07:00
David Zafman
7a528763d1 thrasher: Add dump_ops_enable and optrack_toggle_delay options 
Add dump_ops_enable which continuously dumps ops using 3 commands
Add optrack_toggle_delay to alternate op tracking enablement

Signed-off-by: David Zafman <dzafman@redhat.com>
 2016-04-29 09:43:38 -07:00
David Zafman
4ad3b86604 ceph_manager: Add timeout to admin_socket/osd_admin_socket 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2016-04-29 09:43:38 -07:00
Samuel Just
7e53203e80 rados/singleton-nomsgr: add lfn upgrade tests 
Upgrade from hammer/infernalis to x and verify lfn objects are valid
across upgrade.

Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-04-07 15:35:30 -07:00
Samuel Just
93892eb82a ceph_manager: return exit status on do_get, do_put, do_rm 
Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-04-07 15:35:30 -07:00
Samuel Just
269d6002f1 ceph_manager: add do_rm 
Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-04-07 15:35:30 -07:00
Samuel Just
670ca43dfc ceph_manager: extend do_put and do_get to allow a namespace 
Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-04-07 15:35:30 -07:00
Samuel Just
c8f7694d52 ceph_manager: fix do_get to actually do a get 
Currently unused.

Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-04-07 15:35:30 -07:00
John Spray
53f4430d03 tasks/cephfs: further thrasher fixes 
Move the thrasher-specific methods out of CephManager
into MDSThrasher and plumb them into MDSCluster.

Signed-off-by: John Spray <john.spray@redhat.com
 2016-03-11 10:39:37 +00:00
Sage Weil
6deba7c649 tasks/ceph_manager: dump pgs if other peering timeouts expire 
We were doing this for one of the recovery timeouts but not all of them.

Signed-off-by: Sage Weil <sage@redhat.com>
 2016-03-07 12:21:10 -05:00
Samuel Just
8cf25611fb ceph_manager: use time before mon command for timeout 
Slow mon commands can cause a false failure.

Signed-off-by: Samuel Just <sjust@redhat.com>
 2016-02-19 12:28:36 -08:00
John Spray
d8106fa9e1 tasks: add run_ceph_w to CephManager 
Analogous to raw_cluster_command, but instead
of calling blocking CLI command we're invoking
the -w mode.

Signed-off-by: John Spray <john.spray@redhat.com>
 2016-01-05 18:58:00 +00:00
Samuel Just
89dcc0daf3 ceph_manager: do_pg_scrub: keep scrubbing until it's done 
The ceph pg scrub ... command isn't really guarranteed to
start a scrub, keep reissuing it until the scrub actually
happens.

Related: #12746
Signed-off-by: Samuel Just <sjust@redhat.com>
 2015-11-19 15:07:38 -08:00
Sage Weil
f467a98a29 tasks/ceph_manager: %d -> %s 
Signed-off-by: Sage Weil <sage@redhat.com>
 2015-09-30 14:58:32 -04:00
Sage Weil
a53a80b9f0 tasks/ceph_manager: fix logging on failed pool property 
Signed-off-by: Sage Weil <sage@redhat.com>
 2015-09-30 09:24:38 -04:00
Samuel Just
4e9f1df514 rados: add test for 13234.yaml 
Signed-off-by: Samuel Just <sjust@redhat.com>
 2015-09-29 21:19:10 -07:00
Sage Weil
0e2814d81e tasks/ceph_manager: ignore failure getting pg_num 
Otherwise, we may fail while racing with a workload that deletes a pool:

2015-09-23T15:01:52.855 INFO:tasks.workunit.client.1.vpm128.stdout:[ RUN      ] LibRadosTwoPoolsPP.PromoteSnapTrimRace
2015-09-23T15:01:53.892 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get .rgw pg_num'
2015-09-23T15:01:54.206 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get .rgw.gc pg_num'
2015-09-23T15:01:54.462 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get .users.uid pg_num'
2015-09-23T15:01:54.696 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get .users.email pg_num'
2015-09-23T15:01:55.006 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get .users pg_num'
2015-09-23T15:01:55.296 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get .rgw.buckets.index pg_num'
2015-09-23T15:01:55.523 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get .log pg_num'
2015-09-23T15:01:55.752 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get .usage pg_num'
2015-09-23T15:01:56.188 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get .rgw.buckets.extra pg_num'
2015-09-23T15:01:56.625 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get test-rados-api-vpm128-17360-6 pg_num'
2015-09-23T15:01:56.928 INFO:teuthology.orchestra.run.vpm176:Running: 'sudo adjust-ulimits ceph-coverage /home/ubuntu/cephtest/archive/coverage ceph osd pool get test-rados-api-vpm128-17360-13 pg_num'
2015-09-23T15:01:57.193 INFO:teuthology.orchestra.run.vpm176.stderr:Error ENOENT: unrecognized pool 'test-rados-api-vpm128-17360-13'
2015-09-23T15:01:57.206 ERROR:teuthology.parallel:Exception in parallel execution
Traceback (most recent call last):
...

Signed-off-by: Sage Weil <sage@redhat.com>
 2015-09-24 12:19:07 -04:00
Sage Weil
dad981d339 tasks: sudo ceph for cli 
/var/run/ceph is 770.  This is mainly necessary for any
interaction with the daemon sockets, but it is what users do
and it may avoid log noise.

Signed-off-by: Sage Weil <sage@redhat.com>
 2015-09-11 12:15:01 -04:00
Sage Weil
a328e3e60d tasks/ceph_manager: dump pgs when recover times out 
It is really hard to map a stuck recovery back to the pgs that
are stuck.  This will make it easy.

Signed-off-by: Sage Weil <sage@redhat.com>
 2015-09-08 08:59:49 -04:00
Sage Weil
c93fe1f1c6 tasks/ceph_manager: be silent about sending SIGHUPs 
At the default interval this generates tons of log noise.

Signed-off-by: Sage Weil <sage@redhat.com>
 2015-08-04 13:08:04 -04:00
Andrew Schoen
a3c9a763b1 ceph_manager: don't add an osd to live_osds until it's been revived 
also waits to remove it from dead_osds. this fixes an issue where
do_sighup tries to send a signal to an osd that has not been revived
yet.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
 2015-07-28 17:05:31 -05:00
Andrew Schoen
84d24038aa ceph_manager: adds a do_sighup method 
This method runs in a separate greenlet than do_thrash and will pick a
random live osd to send a signal.SIGHUP to. There is a config option,
sighup_delay, which controls how long to delay between sending the
signals.

Signed-off-by: Andrew Schoen <aschoen@redhat.com>
 2015-07-28 14:46:12 -05:00
Andrew Schoen
ed73f67991 ceph_manager: adds a signal_osd method 
Signed-off-by: Andrew Schoen <aschoen@redhat.com>
 2015-07-28 14:13:30 -05:00
David Zafman
b255db820f thrasher: Can't test ceph-objectstore-tool if nodes turned off (powercycle) 
Signed-off-by: David Zafman <dzafman@redhat.com>
 2015-06-09 14:24:47 -07:00
Samuel Just
91b300d12c rados/thrash: add test for radosgw with snaps 
Signed-off-by: Samuel Just <sjust@redhat.com>
 2015-05-28 15:36:39 -07:00
Samuel Just
2a60852a1d squash: ceph_manager: add utility_task doc string 2015-05-04 14:21:31 -07:00
Samuel Just
015ed70f8a suites/rados: add test for 11429 
This patch also adds some convenience facilities for making
some of the ceph_manager methods into tasks usable from a
yaml file.

Signed-off-by: Samuel Just <sjust@redhat.com>
 2015-05-04 11:53:54 -07:00
John Spray
0de712f42a tasks/ceph_manager: DRY in mds_status 
Signed-off-by: John Spray <john.spray@redhat.com>
 2015-04-14 14:13:38 +01:00
John Spray
5c1071b103 ceph_manager: fix bad type assertions 
In python, isinstance(foo, str) will fail if
a unicode string is passed in.  The correct check
is basestring.

Signed-off-by: John Spray <john.spray@redhat.com>
 2015-04-14 14:13:38 +01:00
Yuri Weinstein
581fcf192f Merge pull request #380 from ceph/wip-11204 
Make sure that ulimits are adjusted for ceph-objectstore-tool
 2015-03-27 12:23:37 -07:00
Sage Weil
dcb5e8da9d Merge remote-tracking branch 'gh/hammer' 
Conflicts:
	.gitignore
 2015-03-26 17:09:33 -07:00
David Zafman
e6ce90fdb1 Make sure that ulimits are adjusted for ceph-objectstore-tool 
Fixes: #11204

Signed-off-by: David Zafman <dzafman@redhat.com>
 2015-03-26 15:18:47 -07:00
David Zafman
6c5300552d ceph_manager: Check for exit status 11 from ceph-objectstore-tool import 
Fixes: #11139

Signed-off-by: David Zafman <dzafman@redhat.com>
 2015-03-20 21:25:41 -07:00
Alfredo Deza
4ed442e44c stdin is no longer a kwarg 
Signed-off-by: Alfredo Deza <adeza@redhat.com>
(cherry picked from commit 49a61dc2d2)
 2015-02-26 14:48:40 -05:00
Alfredo Deza
33f7982480 add the log object to ceph_manager 
Signed-off-by: Alfredo Deza <adeza@redhat.com>
(cherry picked from commit f7c1ca4a1e)
 2015-02-26 14:48:30 -05:00
Alfredo Deza
49a61dc2d2 stdin is no longer a kwarg 
Signed-off-by: Alfredo Deza <adeza@redhat.com>
 2015-02-26 11:34:21 -05:00
Alfredo Deza
f7c1ca4a1e add the log object to ceph_manager 
Signed-off-by: Alfredo Deza <adeza@redhat.com>
 2015-02-26 11:33:47 -05:00
Kefu Chai
64de3cd574 Thrasher: log backtrace of thrown exception 
* add a wrapper to log uncaught exception to self.logger, greenlet also
  prints the backtrace and exception to stderr, but teuthology.log does
  not capture stderr. so we need to catch them by ourselves to reveal
  more info to root-cause this issue.
* log uncaught exception thrown by Thrasher.do_thrash() to self.log.

See: #10630
Signed-off-by: Kefu Chai <kchai@redhat.com>
 2015-02-25 16:10:52 +08:00