]>
git.apps.os.sepia.ceph.com Git - teuthology.git/log
Dan Mick [Tue, 23 Aug 2016 21:02:13 +0000 (14:02 -0700)]
Merge pull request #937 from ceph/wip-console-log
Log IPMI console output for bare metal hosts
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Thu, 18 Aug 2016 21:55:33 +0000 (15:55 -0600)]
Use the new console_log class
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Thu, 18 Aug 2016 21:53:02 +0000 (15:53 -0600)]
Add console_log task
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Mon, 22 Aug 2016 21:43:44 +0000 (15:43 -0600)]
Patch teardown in some test methods
Specifically, test_setup_called() and test_begin_called()
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Mon, 22 Aug 2016 22:59:31 +0000 (16:59 -0600)]
Check ipmi credentials in a separate method
And also call it from spawn_sol_log()
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Thu, 18 Aug 2016 21:47:57 +0000 (15:47 -0600)]
Add PhysicalConsole.spawn_sol_log()
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Mon, 22 Aug 2016 22:55:30 +0000 (16:55 -0600)]
Move ipmi credential storage to PhysicalConsole
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Thu, 18 Aug 2016 23:12:22 +0000 (17:12 -0600)]
PhysicalConsole._pexpect_spawn(): pass logfile arg
... directly when calling pexpect.spawn()
Signed-off-by: Zack Cerza <zack@redhat.com>
Dan Mick [Tue, 23 Aug 2016 18:19:52 +0000 (11:19 -0700)]
Merge pull request #938 from ceph/wip-17084
Kill valgrind.bin processes.
Reviewed-by: Dan Mick <dmick@danceorelse.org>
Zack Cerza [Tue, 23 Aug 2016 15:57:42 +0000 (09:57 -0600)]
Merge pull request #939 from smithfarm/wip-17075
Revert "packages.yaml: reflect python-ceph package split"
Nathan Cutler [Tue, 23 Aug 2016 06:36:36 +0000 (08:36 +0200)]
Revert "packages.yaml: reflect python-ceph package split"
This reverts commit
4543fb64cedffc97c816feff7d4a69cd0456ab15 .
Fixes: http://tracker.ceph.com/issues/17075
Signed-off-by: Nathan Cutler <ncutler@suse.com>
Zack Cerza [Tue, 23 Aug 2016 02:40:31 +0000 (20:40 -0600)]
Kill valgrind.bin processes
http://tracker.ceph.com/issues/17084
Fixes: 17084
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Thu, 18 Aug 2016 21:37:22 +0000 (15:37 -0600)]
PhysicalConsole: rename _exec() to _pexpect_spawn()
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Thu, 18 Aug 2016 21:35:27 +0000 (15:35 -0600)]
PhysicalConsole: split out command generation
Signed-off-by: Zack Cerza <zack@redhat.com>
Dan Mick [Wed, 17 Aug 2016 20:52:20 +0000 (13:52 -0700)]
Merge pull request #936 from ceph/wip-fix-nuke-imports
Fix imports of nuke actions from other modules
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Wed, 17 Aug 2016 20:06:40 +0000 (14:06 -0600)]
Fix imports of nuke actions from other modules
Specifically, ceph-qa-suite.git/tasks/cephfs/filesystem.py apparently
needs the clear_firewall action.
Signed-off-by: Zack Cerza <zack@redhat.com>
Dan Mick [Wed, 17 Aug 2016 17:28:31 +0000 (10:28 -0700)]
Merge pull request #935 from ceph/wip-nuke-improvements
Nuke improvements, and a bonus refactor
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Wed, 17 Aug 2016 15:43:55 +0000 (09:43 -0600)]
Check locks before doing anything else
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 21:09:27 +0000 (15:09 -0600)]
PhysicalConsole.power_cycle(): use timeout=300
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 20:32:49 +0000 (14:32 -0600)]
If console access fails, try SSH
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 20:27:21 +0000 (14:27 -0600)]
Add ConsoleError, and use it
... instead of just a RuntimeError
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 20:18:00 +0000 (14:18 -0600)]
Fix reference to ipmi domain
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 20:14:36 +0000 (14:14 -0600)]
Move most of nuke's actions into their own module
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 20:05:05 +0000 (14:05 -0600)]
Add Remote.console
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 19:52:20 +0000 (13:52 -0600)]
Make teuthology.nuke a subpackage
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 16:16:47 +0000 (10:16 -0600)]
check_console(): attempt a power cycle
If the host is unresponsive and its power is on, reboot it.
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 16:09:15 +0000 (10:09 -0600)]
PhysicalConsole._wait_for_login(): reduce attempts
Instead of trying six times by default, try twice.
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 16:06:47 +0000 (10:06 -0600)]
Move console checking to its own function
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 23:04:38 +0000 (17:04 -0600)]
Remove redundant timeout assignment
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 16 Aug 2016 16:00:10 +0000 (10:00 -0600)]
Move console classes to their own module
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Mon, 15 Aug 2016 23:12:06 +0000 (17:12 -0600)]
Cleanup some timeout logic
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Mon, 15 Aug 2016 23:19:25 +0000 (17:19 -0600)]
Make console checking more readable
Signed-off-by: Zack Cerza <zack@redhat.com>
Kefu Chai [Tue, 16 Aug 2016 08:25:04 +0000 (16:25 +0800)]
Merge pull request #933 from tchaikov/wip-17035
valgrind: add suppression for libunwind in libtcmalloc.so.4.2.6
Reviewed-by: Brad Hubbard <bhubbard@redhat.com>
Kefu Chai [Tue, 16 Aug 2016 08:03:11 +0000 (16:03 +0800)]
valgrind: add suppression for libunwind in libtcmalloc.so.4.2.6
Fixes: http://tracker.ceph.com/issues/17035
Signed-off-by: Kefu Chai <kchai@redhat.com>
Zack Cerza [Mon, 15 Aug 2016 23:02:24 +0000 (17:02 -0600)]
PhysicalConsole._wait_for_login() raise on failure
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Mon, 15 Aug 2016 19:01:53 +0000 (13:01 -0600)]
getRemoteConsole(): look up IPMI credentials
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Mon, 15 Aug 2016 18:41:58 +0000 (12:41 -0600)]
nuke: remove packages before reboot
Just in case daemons coming up after reboot is problematic.
Related: http://tracker.ceph.com/issues/17027
Signed-off-by: Zack Cerza <zack@redhat.com>
Dan Mick [Thu, 11 Aug 2016 21:56:34 +0000 (14:56 -0700)]
Merge pull request #930 from ceph/wip-openstack-regression
Work around regression in openstacksdk 0.9.2
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Thu, 11 Aug 2016 21:30:57 +0000 (15:30 -0600)]
Work around regression in openstacksdk 0.9.2
... that caused `openstack ip floating list -f json` to fail and return
"Connection failure that may be retried."
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Thu, 11 Aug 2016 20:50:44 +0000 (14:50 -0600)]
Merge pull request #922 from ceph/wip-install-py3
install: Allow installing python3 packages universally
Dan Mick [Wed, 10 Aug 2016 19:35:47 +0000 (12:35 -0700)]
Merge pull request #927 from dachary/wip-16972-quiet
openstack: do not display openstack client warnings
Reviewed-by: Dan Mick <dmick@redhat.com>
Dan Mick [Wed, 10 Aug 2016 19:26:20 +0000 (12:26 -0700)]
Merge pull request #928 from ceph/revert-925-wip-ovh-json
Revert "misc.sh(): Don't capture stderr by default"
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Wed, 10 Aug 2016 19:03:50 +0000 (13:03 -0600)]
Revert "misc.sh(): Don't capture stderr by default"
Loic Dachary [Sat, 6 Aug 2016 08:48:14 +0000 (10:48 +0200)]
openstack: do not display openstack client warnings
Because they are concatenated with the STDOUT by the sh function despite
being output on STDERR and that breaks json parsing.
Signed-off-by: Loic Dachary <ldachary@redhat.com>
Dan Mick [Tue, 9 Aug 2016 19:26:36 +0000 (12:26 -0700)]
Merge pull request #926 from ceph/wip-beanstalkd-socketerror
worker: Don't die on a beanstalkc SocketError
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Tue, 9 Aug 2016 18:54:56 +0000 (12:54 -0600)]
Don't die on a beanstalkc SocketError
Signed-off-by: Zack Cerza <zack@redhat.com>
Dan Mick [Mon, 8 Aug 2016 21:13:23 +0000 (14:13 -0700)]
Merge pull request #925 from ceph/wip-ovh-json
misc.sh(): Don't capture stderr by default
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Mon, 8 Aug 2016 19:45:50 +0000 (13:45 -0600)]
misc.sh(): Don't capture stderr by default
If not capturing it, log it.
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Thu, 4 Aug 2016 16:47:44 +0000 (10:47 -0600)]
Merge pull request #924 from smithfarm/wip-lock-centos
teuthology-lock: add centos 7.2 to downburst default table
Zack Cerza [Thu, 4 Aug 2016 14:37:36 +0000 (08:37 -0600)]
Merge pull request #923 from ceph/wip-14794
valgrind: wildcard boost supression
Zack Cerza [Thu, 4 Aug 2016 14:36:33 +0000 (08:36 -0600)]
Merge pull request #916 from smithfarm/wip-python-split
packages.yaml: reflect python-ceph package split
Nathan Cutler [Wed, 3 Aug 2016 12:20:59 +0000 (14:20 +0200)]
teuthology-lock: drop openSUSE 12.2 from downburst default table
AFAIK Downburst does not provide any openSUSE images.
Signed-off-by: Nathan Cutler <ncutler@suse.com>
Nathan Cutler [Wed, 3 Aug 2016 12:18:14 +0000 (14:18 +0200)]
teuthology-lock: add centos 7.2 to downburst default table
When downburst is unavailable, teuthology-lock reverts to default_table and
fails because the only supported CentOS version, 7.2., is missing.
Signed-off-by: Nathan Cutler <ncutler@suse.com>
John Spray [Wed, 3 Aug 2016 11:43:22 +0000 (12:43 +0100)]
valgrind: wildcard boost supression
...so that it will match mangled names.
Fixes: http://tracker.ceph.com/issues/14794
Signed-off-by: John Spray <john.spray@redhat.com>
vasukulkarni [Tue, 2 Aug 2016 03:37:27 +0000 (20:37 -0700)]
Merge pull request #921 from ceph/wip-ansible-dead
CephLab: playbook failures result in status 'dead'
Zack Cerza [Mon, 1 Aug 2016 21:47:49 +0000 (15:47 -0600)]
CephLab: playbook failures result in status 'dead'
Signed-off-by: Zack Cerza <zack@redhat.com>
Dan Mick [Fri, 29 Jul 2016 18:05:18 +0000 (11:05 -0700)]
Merge pull request #914 from ceph/wip-orchestra-logging
orchestra.run logging
Reviewed-by: Dan Mick <dmick@redhat.com>
Dan Mick [Fri, 29 Jul 2016 17:10:49 +0000 (10:10 -0700)]
Merge pull request #920 from ceph/wip-worker-sentinel
Fix an obvious bug in sentinel()
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Fri, 29 Jul 2016 16:53:04 +0000 (10:53 -0600)]
Fix an obvious bug in sentinel()
Signed-off-by: Zack Cerza <zack@redhat.com>
Dan Mick [Wed, 27 Jul 2016 22:22:04 +0000 (15:22 -0700)]
Merge pull request #917 from ceph/wip-dv-broken
Fix regression when using teuthology-suite with --kernel
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Wed, 27 Jul 2016 18:52:20 +0000 (12:52 -0600)]
choose_kernel(): fix call to get_gitbuilder_hash
http://tracker.ceph.com/issues/16838
Fixes: 16838
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Wed, 27 Jul 2016 18:40:25 +0000 (12:40 -0600)]
Remove defaults for get_gitbuilder_hash args
http://tracker.ceph.com/issues/16838
Fixes: 16838
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Wed, 27 Jul 2016 18:38:05 +0000 (12:38 -0600)]
Add reproducer for teuthology-suite regression
http://tracker.ceph.com/issues/16838
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 19 Jul 2016 23:56:58 +0000 (17:56 -0600)]
Update tests to reflect internal changes
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 26 Jul 2016 19:58:34 +0000 (13:58 -0600)]
Instead of StreamHandler, rework copy_file_to()
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 26 Jul 2016 18:29:30 +0000 (12:29 -0600)]
Merge pull request #913 from dmick/wip-os-version
teuthology-suite: add --D/--distroversion
Zack Cerza [Tue, 19 Jul 2016 23:56:58 +0000 (17:56 -0600)]
Drop use of xreadlines()
Signed-off-by: Zack Cerza <zack@redhat.com>
Dan Mick [Tue, 19 Jul 2016 21:12:24 +0000 (14:12 -0700)]
GitbuilderProject: fix/improve warning about multiple refs
warn() tries to warn you when you overspecify a commit (by
giving more than one of ref, tag, branch, or sha1).
1) fix it to actually work (it was always warning)
2) Improve the warning to indicate not only which is taking
precedence but what was supplied for the others.
Signed-off-by: Dan Mick <dan.mick@redhat.com>
Dan Mick [Tue, 19 Jul 2016 05:42:39 +0000 (22:42 -0700)]
suite: Allow distroversion to contain either codename or version
Signed-off-by: Dan Mick <dan.mick@redhat.com>
Dan Mick [Tue, 19 Jul 2016 05:21:01 +0000 (22:21 -0700)]
suite/tests: add distroversion
Signed-off-by: Dan Mick <dan.mick@redhat.com>
Dan Mick [Tue, 19 Jul 2016 05:20:31 +0000 (22:20 -0700)]
suite: Add distroversion
Allow -D/--distroversion to select version of distro to schedule.
Arg can be either numeric version or codename.
Signed-off-by: Dan Mick <dan.mick@redhat.com>
Dan Mick [Tue, 19 Jul 2016 03:27:16 +0000 (20:27 -0700)]
Add OS class method 'version_codename' to infer one from the other
Signed-off-by: Dan Mick <dan.mick@redhat.com>
Nathan Cutler [Mon, 25 Jul 2016 17:34:03 +0000 (19:34 +0200)]
packages.yaml: reflect python-ceph package split
Signed-off-by: Nathan Cutler <ncutler@suse.com>
Zack Cerza [Thu, 21 Jul 2016 15:12:47 +0000 (09:12 -0600)]
Merge pull request #915 from dachary/wip-14799-valgrind
valgrind: add suppression for tcmalloc in libboost_thread-mt.so.1.53.0
Loic Dachary [Thu, 21 Jul 2016 09:13:04 +0000 (11:13 +0200)]
valgrind: add suppression for tcmalloc in libboost_thread-mt.so.1.53.0
Use regular expression to catch:
tcmalloc::ThreadCache::FetchFromCentralCache(unsigned long, unsigned long)
Fixes: http://tracker.ceph.com/issues/14799
Signed-off-by: Loic Dachary <loic@dachary.org>
Zack Cerza [Tue, 19 Jul 2016 16:25:05 +0000 (10:25 -0600)]
Move some stream management into RemoteProcess
For better organization and less copypasta.
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 19 Jul 2016 16:11:27 +0000 (10:11 -0600)]
Log stderr and stdout as long as they aren't pipes
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Tue, 19 Jul 2016 21:25:48 +0000 (15:25 -0600)]
Rewrite orchestra.run tests using mock
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Wed, 20 Jul 2016 22:05:35 +0000 (16:05 -0600)]
Merge pull request #912 from smithfarm/wip-drop-ceph-devel
packages.yaml: drop ceph-devel from rpm section
Nathan Cutler [Wed, 20 Jul 2016 08:55:43 +0000 (10:55 +0200)]
packages.yaml: drop ceph-devel from rpm section
Upgrade tests on https://github.com/ceph/ceph/9744 (which drops the ceph-devel
package after a long period of deprecation) are currently failing
because teuthology tries to install ceph-devel, even though it is not needed
for any tests.
See also https://github.com/ceph/ceph-qa-suite/pull/1068
Signed-off-by: Nathan Cutler <ncutler@suse.com>
Dan Mick [Tue, 19 Jul 2016 17:17:46 +0000 (10:17 -0700)]
Merge pull request #905 from lande1234/wip-lan-testing
fix wrong words
Reviewed-by: Dan Mick <dmick@redhat.com>
Dan Mick [Tue, 19 Jul 2016 17:13:46 +0000 (10:13 -0700)]
Merge pull request #910 from ceph/wip-worker-startup
Stop workers on demand
Dan Mick <dmick@redhat.com>
Dan Mick [Tue, 19 Jul 2016 03:34:14 +0000 (20:34 -0700)]
Merge pull request #911 from ceph/wip-queue-status
teuthology-queue: add --status
Zack Cerza [Fri, 15 Jul 2016 22:40:37 +0000 (16:40 -0600)]
teuthology-queue: add --status
Prints queue status, e.g.:
$ teuthology-queue -m smithi --status
{'paused': False, 'count': 2900, 'name': 'smithi'}
Signed-off-by: Zack Cerza <zack@redhat.com>
Zack Cerza [Fri, 15 Jul 2016 16:05:54 +0000 (10:05 -0600)]
Stop workers on demand
Zack Cerza [Fri, 15 Jul 2016 16:03:21 +0000 (10:03 -0600)]
Generalize need_restart()
So that we can also use it to tell the workers to stop
Signed-off-by: Zack Cerza <zack@redhat.com>
John Spray [Fri, 15 Jul 2016 16:14:22 +0000 (17:14 +0100)]
Revert "task/packages: install libcephfs1-dev"
This reverts commit
48be9e89df559dce13731fcdf59153181ffd2646 .
Fixes: http://tracker.ceph.com/issues/16703
Signed-off-by: John Spray <john.spray@redhat.com>
Dan Mick [Fri, 15 Jul 2016 01:45:01 +0000 (18:45 -0700)]
Merge pull request #909 from ceph/wip-worker-startup
Fix slow worker startup
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Fri, 15 Jul 2016 00:19:56 +0000 (18:19 -0600)]
Fix slow worker startup
We can selectively skip fetching and/or bootstrapping if either have
been completed successfully in the last 60s
Signed-off-by: Zack Cerza <zack@redhat.com>
Dan Mick [Thu, 14 Jul 2016 20:56:42 +0000 (13:56 -0700)]
Merge pull request #908 from ceph/wip-sched-fail-ceph-ver
Fix regression in schedule failure emails
Reviewed-by: Dan Mick <dmick@redhat.com>
Zack Cerza [Thu, 14 Jul 2016 18:08:25 +0000 (12:08 -0600)]
Fix regression in schedule failure emails
Emails weren't getting sent when builds were broken
Signed-off-by: Zack Cerza <zack@redhat.com>
Oleh Prypin [Thu, 14 Jul 2016 16:54:15 +0000 (19:54 +0300)]
install: Allow installing python3 packages universally
by replacing "python3-*" with "python34-*" on CentOS
Signed-off-by: Oleh Prypin <oleh@pryp.in>
Zack Cerza [Wed, 13 Jul 2016 15:11:32 +0000 (09:11 -0600)]
Merge pull request #904 from dmick/wip-version
teuthology: log version number
Zack Cerza [Wed, 13 Jul 2016 15:10:35 +0000 (09:10 -0600)]
Merge pull request #907 from ceph/wip-16640
task/packages: install libcephfs1-dev
John Spray [Wed, 13 Jul 2016 11:55:50 +0000 (12:55 +0100)]
task/packages: install libcephfs1-dev
The way we currently build the cephfs java
bindings depends on the unversioned .so file
which is in the -dev package, not the
main one.
Fixes: http://tracker.ceph.com/issues/16640
Signed-off-by: John Spray <john.spray@redhat.com>
Dan Mick [Thu, 7 Jul 2016 17:50:48 +0000 (10:50 -0700)]
teuthology: log version number
Round 2, this time with testing
Signed-off-by: Dan Mick <dan.mick@redhat.com>
Zack Cerza [Tue, 12 Jul 2016 22:08:23 +0000 (16:08 -0600)]
Merge pull request #902 from dmick/wip-schedule-suite-test
Add tests for schedule_suite(); fix some schedule_fail arguments
Zack Cerza [Mon, 11 Jul 2016 19:47:04 +0000 (13:47 -0600)]
Merge pull request #906 from dmick/wip-email
Resume paying attention to --email
Dan Mick [Mon, 11 Jul 2016 19:25:07 +0000 (12:25 -0700)]
Resume paying attention to --email
Change in
61350db caused email arg to be lost. Fix, add test
Signed-off-by: Dan Mick <dan.mick@redhat.com>
lande1234 [Fri, 8 Jul 2016 05:53:16 +0000 (13:53 +0800)]
fix wrong words
Signed-off-by: lande1234 <lan.de3@zte.com.cn>