You are viewing a plain text version of this content. The canonical link for it is here.
Posted to reviews@aurora.apache.org by David McLaughlin <da...@dmclaughlin.com> on 2018/02/14 05:27:34 UTC
Review Request 65648: Do not reschedule a PARTITIONED task if it was
in KILLING state
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/65648/
-----------------------------------------------------------
Review request for Aurora, Jordan Ly and Santhosh Kumar Shanmugham.
Repository: aurora
Description
-------
In previous behavior KILLING -> LOST never resulted in a reschedule. In the new code, KILLING -> PARTITIONED -> LOST would. This adds a check to PARTITIONED -> LOST to only reschedule when KILLING isn't one of the previous states.
Diffs
-----
src/main/java/org/apache/aurora/scheduler/state/TaskStateMachine.java f325bf46350bda49c3aaca151b76aad9649cb91a
src/test/java/org/apache/aurora/scheduler/state/TaskStateMachineTest.java 050a46ad8f06bbeeb88079a2583296348b6f8359
Diff: https://reviews.apache.org/r/65648/diff/1/
Testing
-------
./gradlew test
Thanks,
David McLaughlin
Re: Review Request 65648: Do not reschedule a PARTITIONED task if it
was in KILLING state
Posted by Jordan Ly <jo...@gmail.com>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/65648/#review197574
-----------------------------------------------------------
Ship it!
Ship It!
- Jordan Ly
On Feb. 14, 2018, 5:27 a.m., David McLaughlin wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/65648/
> -----------------------------------------------------------
>
> (Updated Feb. 14, 2018, 5:27 a.m.)
>
>
> Review request for Aurora, Jordan Ly and Santhosh Kumar Shanmugham.
>
>
> Repository: aurora
>
>
> Description
> -------
>
> In previous behavior KILLING -> LOST never resulted in a reschedule. In the new code, KILLING -> PARTITIONED -> LOST would. This adds a check to PARTITIONED -> LOST to only reschedule when KILLING isn't one of the previous states.
>
>
> Diffs
> -----
>
> src/main/java/org/apache/aurora/scheduler/state/TaskStateMachine.java f325bf46350bda49c3aaca151b76aad9649cb91a
> src/test/java/org/apache/aurora/scheduler/state/TaskStateMachineTest.java 050a46ad8f06bbeeb88079a2583296348b6f8359
>
>
> Diff: https://reviews.apache.org/r/65648/diff/1/
>
>
> Testing
> -------
>
> ./gradlew test
>
>
> Thanks,
>
> David McLaughlin
>
>
Re: Review Request 65648: Do not reschedule a PARTITIONED task if it
was in KILLING state
Posted by Aurora ReviewBot <wf...@apache.org>.
-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/65648/#review197479
-----------------------------------------------------------
Master (cb0faf8) is red with this patch.
./build-support/jenkins/build.sh
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
WARN] Reached consecutive failure limit.
--------------- Captured log call ----------------
health_checker.py 167 INFO INFO] Reached consecutive success limit.
health_checker.py 143 WARNING WARN] Health check failure: failure-2
health_checker.py 184 WARNING WARN] Ignoring failure of attempt: 2
health_checker.py 143 WARNING WARN] Health check failure: failure-3
health_checker.py 159 WARNING WARN] Reached consecutive failure limit.
3 failed, 797 passed, 6 skipped in 347.64 seconds
FAILURE
Waiting for background workers to finish.
05:47:32 06:26 [complete]
FAILURE
I will refresh this build result if you post a review containing "@ReviewBot retry"
- Aurora ReviewBot
On Feb. 14, 2018, 5:27 a.m., David McLaughlin wrote:
>
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/65648/
> -----------------------------------------------------------
>
> (Updated Feb. 14, 2018, 5:27 a.m.)
>
>
> Review request for Aurora, Jordan Ly and Santhosh Kumar Shanmugham.
>
>
> Repository: aurora
>
>
> Description
> -------
>
> In previous behavior KILLING -> LOST never resulted in a reschedule. In the new code, KILLING -> PARTITIONED -> LOST would. This adds a check to PARTITIONED -> LOST to only reschedule when KILLING isn't one of the previous states.
>
>
> Diffs
> -----
>
> src/main/java/org/apache/aurora/scheduler/state/TaskStateMachine.java f325bf46350bda49c3aaca151b76aad9649cb91a
> src/test/java/org/apache/aurora/scheduler/state/TaskStateMachineTest.java 050a46ad8f06bbeeb88079a2583296348b6f8359
>
>
> Diff: https://reviews.apache.org/r/65648/diff/1/
>
>
> Testing
> -------
>
> ./gradlew test
>
>
> Thanks,
>
> David McLaughlin
>
>