Uh oh!
There was an error while loading. Please reload this page.
TEZ-4488: TaskSchedulerManager might not be initialized when the first DAG comes - #280
Conversation
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
c6f83ec to
0a46582Compare
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
This comment was marked as outdated.
| taskSchedulerManager); | ||
| addIfServiceDependency(taskSchedulerManager, clientRpcServer); | ||
| appMasterReadinessService = createAppMasterReadinessService(); |
There was a problem hiding this comment.
Issue is happening as a corner case, as the taskscheduler could be taking longer to init. In this case, possibly "LlapTaskSchedulerService" is taking time to init due to ZK dependency.
Instead of adding one more dependency, why not throw RuntimeException in TaskSchedulerManager::getTaskSchedulerClassName (i.e until all taskSchedulers are initied w.r.t appContext) ?
abstractdog
commented
Apr 11, 2023
@rbalamohan: I'm afraid we have no other choice than blocking the first DAG here, as throwing an exception from TaskSchedulerManager::getTaskSchedulerClassName would lead to a similar exception that I'm trying to avoid here |
This comment was marked as outdated.
This comment was marked as outdated.
rbalamohan
commented
Apr 20, 2023
Lgtm. +1 on the recent patch with timeout. |
abstractdog
commented
Apr 20, 2023
thanks @rbalamohan , waiting for green tests before merging |
tez-yetus
commented
Apr 20, 2023
🎊 +1 overall
This message was automatically generated. |
tez-yetus
commented
Apr 20, 2023
💔 -1 overall
This message was automatically generated. |
…t DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (cherry picked from commit 249e017)
…t DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (cherry picked from commit 249e017)
…t DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (cherry picked from commit 249e017) (cherry picked from commit 41343dc)
…t DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (cherry picked from commit 249e017) (cherry picked from commit 41343dc)
…t DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (cherry picked from commit 249e017) (cherry picked from commit 41343dc)
…t DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (cherry picked from commit 249e017) (cherry picked from commit 41343dc)
…t DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (cherry picked from commit 249e017) (cherry picked from commit 41343dc)
…t DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (cherry picked from commit 249e017) (cherry picked from commit 41343dc) (cherry picked from commit 6d81766)
…n the first DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (#23) (cherry picked from commit 249e017) (cherry picked from commit 41343dc) (cherry picked from commit 6d81766) Co-authored-by: Bodor Laszlo <bodorlaszlo0202@gmail.com>
…n the first DAG comes (apache#280) (Laszlo Bodor reviewed by Rajesh Balamohan) (#23) (cherry picked from commit 249e017) (cherry picked from commit 41343dc) (cherry picked from commit 6d81766) Co-authored-by: Bodor Laszlo <bodorlaszlo0202@gmail.com>
No description provided.