Commit Graph

2769 Commits

Author SHA1 Message Date
Zack Cerza
d32890de42 Merge pull request #310 from ceph/wip-valgrind
misc: pass premable through when no valgrind
2014-08-12 10:24:58 -06:00
Sage Weil
a5bb9d9552 misc: pass premable through when no valgrind
Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-11 17:31:57 -07:00
Zack Cerza
293c2bd49c Be more informative if owner info is missing
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-11 14:55:49 -06:00
Zack Cerza
edfeffe3dd Fix format string
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-11 14:20:33 -06:00
Zack Cerza
2e7cf59395 Don't redefine 'remote'
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-11 14:18:50 -06:00
Zack Cerza
1720eee33d Don't import from inside functions
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-11 14:14:45 -06:00
Zack Cerza
e09a6ebdce Attempt fetches on init, before running jobs
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-08 13:20:25 -06:00
Zack Cerza
32485e7503 Also rmtree on RuntimeError
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-08 13:18:44 -06:00
Zack Cerza
35b82675a5 Disable pyc removal for now
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-08 13:11:52 -06:00
Zack Cerza
7b9db2f41b Bump all log levels to info
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-08 11:43:46 -06:00
Zack Cerza
28535846d0 Default to using the master branch of teuthology
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-07 12:11:13 -06:00
Zack Cerza
1f957a6a9e For non-scheduled jobs, check out the suite repo.
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-07 11:13:42 -06:00
Zack Cerza
7912fd5742 After resetting a repo, remove all pyc files
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-07 11:13:42 -06:00
Zack Cerza
ba7a2dad39 Make teuthology-{suite,worker} share repo code
Repo-fetching code, that is.

Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-07 11:13:42 -06:00
Zack Cerza
f7d882c24e Reimplement the file locker as a context manager
Also give it a 'noop' flag to enable its callers to optionally not use
locking.

Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-07 11:13:42 -06:00
Zack Cerza
14015af475 Move exception into a new exceptions module
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-07 11:13:42 -06:00
Zack Cerza
6807a10d27 Move functions into repo_utils
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-07 11:13:41 -06:00
Zack Cerza
76a93fd3e1 Fix NameError
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-07 11:13:23 -06:00
Zack Cerza
a677b4f2a9 Correctly find both internal and external tasks
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-06 10:06:35 -06:00
Zack Cerza
5b5f02ecd2 Remove most ceph-specific tasks. They are in ceph-qa-suite now.
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-06 10:06:34 -06:00
Zack Cerza
d5f26aaf6a Merge pull request #306 from ceph/wip-9024
ignore errors if 'ps aux' fails
2014-08-06 09:42:49 -06:00
Alfredo Deza
f9aa9c1e64 ignore errors if 'ps aux' fails
Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
2014-08-06 10:53:21 -04:00
Zack Cerza
b93abaf5d3 Merge pull request #305 from ceph/wip-8640
implement `--version` and improve `setup.py`
2014-08-04 14:38:29 -06:00
Alfredo Deza
fe6f46114a create a --version flag for teuthology
Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
2014-08-04 15:12:06 -04:00
Alfredo Deza
7d0a072da1 pull the version and description in setup.py
Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
2014-08-04 15:11:47 -04:00
Alfredo Deza
a2bb186fe3 add an initial version to teuthology
Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
2014-08-04 15:11:26 -04:00
Zack Cerza
0d1fe79b70 Tweak usage statement
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-04 11:50:22 -06:00
Zack Cerza
ef21080605 Merge pull request #304 from ceph/wip-fix-build-matrix
suite: fix build_matrix when non-yaml present in % dir
2014-08-04 10:10:02 -06:00
Sage Weil
21e27637d3 suite: fix build_matrix when non-yaml present in % dir
First, if we get a non-yaml file, return an empty list.

Second, if we are in a % mode, exclude an empty raw result from the
combination.  Otherwise, we get no resulting jobs.

Signed-off-by: Sage Weil <sage@redhat.com>
2014-08-04 09:07:08 -07:00
David Zafman
9f7107b3a9 Merge pull request #303 from ceph/wip-8930
Tasks are failing since using "data" pool no longer part of default install

Reviewed-by: Sage Weil <sage@redhat.com>
2014-08-01 18:28:05 -07:00
David Zafman
54af8b2be3 Tasks are failing since using "data" pool no longer part of default install
Create a pool specifically for each task

Fixes: 8930

Signed-off-by: David Zafman <david.zafman@inktank.com>
2014-08-01 16:06:39 -07:00
Zack Cerza
3440d664e7 Expand '~' in ssh key paths
... since paramiko doesn't bother to do this

Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-08-01 14:27:59 -06:00
Zack Cerza
f4c953c1fa Merge pull request #301 from ceph/wip-nuke-vpm
nuke: allow nuking vpm hosts
2014-08-01 13:25:18 -06:00
Sage Weil
f84458bd93 ceph_manager: increase osd revival timeout from 75 -> 150 seconds
Saw this fail on a firefly run just bc of valgrind slowness.  Make the
timeout higher!

Signed-off-by: Sage Weil <sage@redhat.com>
2014-07-30 21:34:45 -07:00
Sage Weil
7c3fc404d6 valgrind: ignore ec plugin factory leaks
Signed-off-by: Sage Weil <sage@redhat.com>
2014-07-30 18:14:08 -07:00
John Spray
9e2c82824b Merge pull request #300 from ceph/wip-mpi-vers
task/mpi: Explicit check for version mismatch
2014-07-30 23:37:39 +01:00
Zack Cerza
094cd5b3ba Merge pull request #302 from ceph/wip-8850
tell us if you are really not running any more
2014-07-30 14:31:27 -06:00
Alfredo Deza
4686115c48 add another call to use "ps aux"
Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
2014-07-30 16:28:02 -04:00
Alfredo Deza
ec342d8146 tell us if you are really not running any more
Signed-off-by: Alfredo Deza <alfredo.deza@inktank.com>
2014-07-30 16:17:01 -04:00
Sage Weil
1c7c2eb4b9 nuke: allow nuking vpm hosts
Skip the console check, though.

Signed-off-by: Sage Weil <sage@redhat.com>
2014-07-30 11:12:09 -07:00
John Spray
491817961a task/mpi: Explicit check for version mismatch
Instead of proceeding to have mpiexec fail or hang,
do an up-front check that the MPI version is the same
on all of the nodes in the test.

Signed-off-by: John Spray <john.spray@redhat.com>
2014-07-30 14:29:11 +01:00
Zack Cerza
b8687a54aa Fix package_version_for_hash() call
Signed-off-by: Zack Cerza <zack.cerza@inktank.com>
2014-07-29 14:43:41 -06:00
John Spray
0147c74671 task/mds_journal_migration: fix cleanup
Previously this was remounting in the background
which meant that if a workunit was executed immediately
afterwards it could put files in the mount dir before
the FS was mounted, causing the mount to fail.

Signed-off-by: John Spray <john.spray@redhat.com>
2014-07-29 13:11:12 +01:00
John Spray
cc7c9b2e8b task/workunit: add sudo to dir delete
This failure to delete used to be hidden in
a catch-all exception handler that was
removed in 396dc073.

This was causing failures in kcephfs suite
like:
Command failed on plana96 with status 1: 'rm -rf -- /home/ubuntu/cephtest/mnt.0/client.0'

Signed-off-by: John Spray <john.spray@redhat.com>
2014-07-27 12:04:23 +01:00
Zack Cerza
2999a4d492 Merge pull request #296 from ceph/wip-standby-mds
Wip standby mds
2014-07-25 16:44:34 -06:00
Sage Weil
883f2a76f4 task/osd_recovery: use rbd instead of metadata pool
Signed-off-by: Sage Weil <sage@redhat.com>
2014-07-25 11:39:42 -07:00
Sage Weil
4e56521ebe task/osd_backfill: use 'rbd' instead of 'data' pool
Signed-off-by: Sage Weil <sage@redhat.com>
2014-07-25 11:33:25 -07:00
John Spray
2ed1131fc9 task/workunit: general tidy
* PEP8-ize whitespace
* specialize a catch-all except:
* don't alias 'misc' to 'teuthology'

Signed-off-by: John Spray <john.spray@redhat.com>
2014-07-25 13:33:31 +01:00
John Spray
396dc073cb task/workunit: clean up dir deletion
This was always trying to remove the mountpoint,
and then swallowing the exception and printing
it to the log.

Instead:
 * Ensure it's only trying to delete mnt if it created it
 * Don't swallow those exceptions: if deleting the dirs
   fails then something has gone wrong and it should bubble
   up.

Signed-off-by: John Spray <john.spray@redhat.com>
2014-07-25 13:33:31 +01:00
John Spray
a96a7b027c task/cephfs: generalise Filesystem for multi-MDS
This enables tasks like mds_journal_migration to be
run in an environment with standby-replay MDSs present.

Signed-off-by: John Spray <john.spray@redhat.com>
2014-07-25 13:33:31 +01:00