[GitHub] [incubator-celeborn] waitinfuture commented on a diff in pull request #1588: [CELEBORN-656] Batch revive RPCs in client to avoid too many requests

["waitinfuture (via GitHub)" <gi...@apache.org>]

waitinfuture commented on code in PR #1588:
URL: https://github.com/apache/incubator-celeborn/pull/1588#discussion_r1231746550


##########
client/src/main/scala/org/apache/celeborn/client/LifecycleManager.scala:
##########
@@ -497,43 +492,57 @@ class LifecycleManager(appId: String, val conf: CelebornConf) extends RpcEndpoin
       context: RpcCallContext,
       applicationId: String,
       shuffleId: Int,
-      mapId: Int,
-      attemptId: Int,
-      partitionId: Int,
-      oldEpoch: Int,
-      oldPartition: PartitionLocation,
-      cause: StatusCode): Unit = {
+      mapIds: util.List[Integer],
+      partitionIds: util.List[Integer],
+      oldEpochs: util.List[Integer],
+      oldPartitions: util.List[PartitionLocation],
+      causes: util.List[StatusCode]): Unit = {
+    val contextWrapper =
+      ChangeLocationsCallContext(context, partitionIds.size())
     // If shuffle not registered, reply ShuffleNotRegistered and return
     if (!registeredShuffle.contains(shuffleId)) {
       logError(s"[handleRevive] shuffle $shuffleId not registered!")
-      context.reply(ChangeLocationResponse(StatusCode.SHUFFLE_NOT_REGISTERED, None))
+      0 until partitionIds.size() foreach (idx => {

Review Comment:
   What does "bypass method" mean?



-- 
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: issues-unsubscribe@celeborn.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org