You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@druid.apache.org by "cryptoe (via GitHub)" <gi...@apache.org> on 2023/04/06 16:43:31 UTC
[GitHub] [druid] cryptoe opened a new pull request, #14038: Fixing overlord unable to become a leader when syncing the lock from metadata store
cryptoe opened a new pull request, #14038:
URL: https://github.com/apache/druid/pull/14038
Overlord fails to become the leaders if its encounters an exception while getting the taskLocks from metadatastore.
I saw this problem when one of the clusters was running a MSQ job before this patch: https://github.com/apache/druid/pull/13282 which changes task priority.
During upgrade, the overlord failed to start due to:
```
2023-03-28T15:59:53,571 ERROR [LeaderSelector[/druid/overlord/_OVERLORD]] org.apache.druid.curator.discovery.CuratorDruidLeaderSelector - listener becomeLeader() failed. Unable to become leader: {class=org.apache.druid.curator.discovery.CuratorDruidLeaderSelector, exceptionType=class java.lang.RuntimeException, exceptionMessage=java.lang.reflect.InvocationTargetException}
java.lang.RuntimeException: java.lang.reflect.InvocationTargetException
at org.apache.druid.indexing.overlord.TaskMaster$1.becomeLeader(TaskMaster.java:179)
at org.apache.druid.curator.discovery.CuratorDruidLeaderSelector$1.isLeader(CuratorDruidLeaderSelector.java:98)
at org.apache.curator.framework.listen.MappingListenerManager.lambda$forEach$0(MappingListenerManager.java:92)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1128) ~[?:?]
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:628) ~[?:?]
at java.lang.Thread.run(Thread.java:829) ~[?:?]
Caused by: java.lang.reflect.InvocationTargetException
at jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[?:?]
at jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) ~[?:?]
at jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:?]
at java.lang.reflect.Method.invoke(Method.java:566) ~[?:?]
at org.apache.druid.java.util.common.lifecycle.Lifecycle$AnnotationBasedHandler.start(Lifecycle.java:446)
at org.apache.druid.java.util.common.lifecycle.Lifecycle.start(Lifecycle.java:341)
at org.apache.druid.indexing.overlord.TaskMaster$1.becomeLeader(TaskMaster.java:176)
Caused by: java.lang.IllegalArgumentException: lock priority[0] is different from task priority[50]
at com.google.common.base.Preconditions.checkArgument(Preconditions.java:148) ~[guava-16.0.1.jar:?]
at org.apache.druid.indexing.overlord.TaskLockbox.verifyAndCreateOrFindLockPosse(TaskLockbox.java:260)
at org.apache.druid.indexing.overlord.TaskLockbox.syncFromStorage(TaskLockbox.java:169)
at org.apache.druid.indexing.overlord.TaskQueue.start(TaskQueue.java:179)
at jdk.internal.reflect.NativeMethodAccessorImpl.invoke0(Native Method) ~[?:?]
at jdk.internal.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:62) ~[?:?]
at jdk.internal.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:43) ~[?:?]
at java.lang.reflect.Method.invoke(Method.java:566) ~[?:?]
at org.apache.druid.java.util.common.lifecycle.Lifecycle$AnnotationBasedHandler.start(Lifecycle.java:446)
at org.apache.druid.java.util.common.lifecycle.Lifecycle.start(Lifecycle.java:341)
at org.apache.druid.indexing.overlord.TaskMaster$1.becomeLeader(TaskMaster.java:176)
... 5 more
```
Went ahead and fixed the db sync such that any task groups whose task locks are unable to be synced with the db are first identified.
Post that all the tasks in that task group are killed by the overlord piggy backing on the logic introduced in this patch : https://github.com/apache/druid/pull/13172
##### Key changed/added classes in this PR
* `TaskLockBox`
<hr>
<!-- Check the items by putting "x" in the brackets for the done things. Not all of these items apply to every PR. Remove the items which are not done or not relevant to the PR. None of the items from the checklist below are strictly necessary, but it would be very helpful if you at least self-review the PR. -->
This PR has:
- [x] been self-reviewed.
- [x] added unit tests or modified existing tests to cover new code paths, ensuring the threshold for [code coverage](https://github.com/apache/druid/blob/master/dev/code-review/code-coverage.md) is met.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@druid.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@druid.apache.org
For additional commands, e-mail: commits-help@druid.apache.org
[GitHub] [druid] cryptoe merged pull request #14038: Fixing overlord unable to become a leader when syncing the lock from metadata store
Posted by "cryptoe (via GitHub)" <gi...@apache.org>.
cryptoe merged PR #14038:
URL: https://github.com/apache/druid/pull/14038
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: commits-unsubscribe@druid.apache.org
For queries about this service, please contact Infrastructure at:
users@infra.apache.org
---------------------------------------------------------------------
To unsubscribe, e-mail: commits-unsubscribe@druid.apache.org
For additional commands, e-mail: commits-help@druid.apache.org