Repository: aurora Updated Branches: refs/heads/master 19866b516 -> 8fca74501
Modify the watch_secs assertion on scheduler This feature intends to improve reliability and performance of the Aurora scheduler job updater by relying on health check status rather than watch_secs timeout when deciding an individual instance update state. See this epic: https://issues.apache.org/jira/browse/AURORA-894 and the design doc: https://docs.google.com/document/d/1ZdgW8S4xMhvKW7iQUX99xZm10NXSxEWR0a-21FP5d94/edit# for more details and background. Testing Done: ./gradlew build ./gradlew :test --tests "org.apache.aurora.scheduler.updater.JobUpdaterIT" ./build-support/jenkins/build.sh Bugs closed: AURORA-894 Reviewed at https://reviews.apache.org/r/51536/ Project: http://git-wip-us.apache.org/repos/asf/aurora/repo Commit: http://git-wip-us.apache.org/repos/asf/aurora/commit/8fca7450 Tree: http://git-wip-us.apache.org/repos/asf/aurora/tree/8fca7450 Diff: http://git-wip-us.apache.org/repos/asf/aurora/diff/8fca7450 Branch: refs/heads/master Commit: 8fca7450152fc242e632e643f44461279399b7ca Parents: 19866b5 Author: Kai Huang <texasred2...@hotmail.com> Authored: Wed Sep 7 17:06:39 2016 -0700 Committer: Zameer Manji <zma...@apache.org> Committed: Wed Sep 7 17:06:39 2016 -0700 ---------------------------------------------------------------------- RELEASE-NOTES.md | 1 + .../java/org/apache/aurora/scheduler/updater/UpdateFactory.java | 4 ++-- .../java/org/apache/aurora/scheduler/updater/JobUpdaterIT.java | 2 +- 3 files changed, 4 insertions(+), 3 deletions(-) ---------------------------------------------------------------------- http://git-wip-us.apache.org/repos/asf/aurora/blob/8fca7450/RELEASE-NOTES.md ---------------------------------------------------------------------- diff --git a/RELEASE-NOTES.md b/RELEASE-NOTES.md index d79aaad..bbf7198 100644 --- a/RELEASE-NOTES.md +++ b/RELEASE-NOTES.md @@ -34,6 +34,7 @@ scheduler upgrade should be performed by bringing all schedulers down, and then bringing upgraded schedulers up. A rolling upgrade would result in no leading scheduler for the duration of the roll which could be confusing to monitor and debug. +- Add a new MTTS (Median Time To Starting) metric in addition to MTTA and MTTR. ### Deprecations and removals: http://git-wip-us.apache.org/repos/asf/aurora/blob/8fca7450/src/main/java/org/apache/aurora/scheduler/updater/UpdateFactory.java ---------------------------------------------------------------------- diff --git a/src/main/java/org/apache/aurora/scheduler/updater/UpdateFactory.java b/src/main/java/org/apache/aurora/scheduler/updater/UpdateFactory.java index ac8df3e..14c2d2d 100644 --- a/src/main/java/org/apache/aurora/scheduler/updater/UpdateFactory.java +++ b/src/main/java/org/apache/aurora/scheduler/updater/UpdateFactory.java @@ -77,8 +77,8 @@ interface UpdateFactory { requireNonNull(instructions); IJobUpdateSettings settings = instructions.getSettings(); checkArgument( - settings.getMinWaitInInstanceRunningMs() > 0, - "Min wait in running must be positive."); + settings.getMinWaitInInstanceRunningMs() >= 0, + "Min wait in running must be non-negative."); checkArgument( settings.getUpdateGroupSize() > 0, "Update group size must be positive."); http://git-wip-us.apache.org/repos/asf/aurora/blob/8fca7450/src/test/java/org/apache/aurora/scheduler/updater/JobUpdaterIT.java ---------------------------------------------------------------------- diff --git a/src/test/java/org/apache/aurora/scheduler/updater/JobUpdaterIT.java b/src/test/java/org/apache/aurora/scheduler/updater/JobUpdaterIT.java index 04551f1..196df47 100644 --- a/src/test/java/org/apache/aurora/scheduler/updater/JobUpdaterIT.java +++ b/src/test/java/org/apache/aurora/scheduler/updater/JobUpdaterIT.java @@ -1062,7 +1062,7 @@ public class JobUpdaterIT extends EasyMockTest { expectInvalid(update); update = makeJobUpdate().newBuilder(); - update.getInstructions().getSettings().setMinWaitInInstanceRunningMs(0); + update.getInstructions().getSettings().setMinWaitInInstanceRunningMs(-1); expectInvalid(update); }