| b715ea477b64ac4814fea81f7988ee4ef353e9a0 |
|
24-Feb-2015 |
Fang Deng <fdeng@chromium.org> |
Revert "[autotest] Teach the scheduler to recover stuck hosts"
This reverts commit a6c0d253e16cf6ddf94a75ef56e50e376b09411c.
Change-Id: Id430e6d563345d39f55d2335da3ee5f2918da205
Reviewed-on: https://chromium-review.googlesource.com/253020
Reviewed-by: Fang Deng <fdeng@chromium.org>
Commit-Queue: Fang Deng <fdeng@chromium.org>
Tested-by: Fang Deng <fdeng@chromium.org>
/external/autotest/scheduler/monitor_db_cleanup_test.py
|
| a6c0d253e16cf6ddf94a75ef56e50e376b09411c |
|
12-Feb-2015 |
Fang Deng <fdeng@chromium.org> |
[autotest] Teach the scheduler to recover stuck hosts
If we have hosts stuck in a state without a task or hqe the
scheduler should recover the host.
The correctness of this CL relies on the fact that
- the scheduler is single threaded and
- in the epilog of a task agent, deactivation of the
special task/hqe, if occurs, must always occur together with
updating host status to the right status within the same tick,
otherwise we will have race condition on updating the host status.
BUG=chromium:432549
TEST=add unittest. run scheduler, manually corrupt the database,
and observer the host is recovered.
DEPLOY=scheduler
Change-Id: Iab39fd5301e14cfa9623ae09ab80589dac9c7564
Reviewed-on: https://chromium-review.googlesource.com/249010
Tested-by: Fang Deng <fdeng@chromium.org>
Reviewed-by: Mungyung Ryu <mkryu@google.com>
Commit-Queue: Fang Deng <fdeng@chromium.org>
/external/autotest/scheduler/monitor_db_cleanup_test.py
|
| e0493a4af57c1a73376a7bafaed542c01f588196 |
|
15-Nov-2010 |
Eric Li <ericli@chromium.org> |
Merge remote branch 'cros/upstream' into tempbranch
BUG=
TEST=
Review URL: http://codereview.chromium.org/4823005
Change-Id: I5d56f1c10d0fce7f9d7dc3ad727ea52dcb9b2d6c
/external/autotest/scheduler/monitor_db_cleanup_test.py
|
| a3a2841558fddc6653de65c58ff01f799334050b |
|
02-Mar-2010 |
jamesren <jamesren@592f7852-d20e-0410-864c-8624ca9c26a4> |
Adding "executable" property to scheduler unittests
Signed-off-by: James Ren <jamesren@google.com>
git-svn-id: http://test.kernel.org/svn/autotest/trunk@4279 592f7852-d20e-0410-864c-8624ca9c26a4
/external/autotest/scheduler/monitor_db_cleanup_test.py
|
| 8dbd05aa0a62a0b17bf4b19131250a8f6cfccf02 |
|
12-Jan-2010 |
showard <showard@592f7852-d20e-0410-864c-8624ca9c26a4> |
Implement periodic reverification of dead hosts, configurable in global_config. Implemented as part of the periodic cleanup, so the frequency of reverification is bounded by the periodic cleanup interval. I felt this would be acceptable and putting this in the existing cleanup class makes things more nicely organized.
Signed-off-by: Steve Howard <showard@google.com>
git-svn-id: http://test.kernel.org/svn/autotest/trunk@4100 592f7852-d20e-0410-864c-8624ca9c26a4
/external/autotest/scheduler/monitor_db_cleanup_test.py
|