[CELEBORN-1822] Respond to RegisterShuffle with max epoch PartitionLocation to avoid revive #3135
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
RexXiong
approved these changes
Mar 13, 2025
| @@ -145,6 +145,7 @@ class LifecycleManager(val appUniqueId: String, val conf: CelebornConf) extends | |||
| locations: util.List[PartitionLocation]): Unit = { | |||
| val map = latestPartitionLocation.computeIfAbsent(shuffleId, newMapFunc) | |||
| locations.asScala.foreach(location => map.put(location.getId, location)) | |||
| invalidateLatestMaxLocsCache(shuffleId) | |||
There was a problem hiding this comment.
Since cache invalidate when location updated, so we can consider extending rpcCacheExpireTime to a longer duration, such as 60s
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What changes were proposed in this pull request?
LifecycleManager respond to RegisterShuffle with max epoch PartitionLocation.
Why are the changes needed?
Newly spun up executors in a Spark job will still get the partitionLocations with the minEpoch of the celeborn lost worker.
These executors will fail to connect to the lost worker and then call into revive to get the latest PartitionLocation for a given partitionId in
ChangePartitionManager.getLatestPartition().Return with max epoch can reduce such revive requests.
Does this PR introduce any user-facing change?
No
How was this patch tested?
UT.