mridulm commented on code in PR #40690:
URL: https://github.com/apache/spark/pull/40690#discussion_r1186666232
##########
core/src/main/scala/org/apache/spark/MapOutputTracker.scala:
##########
@@ -157,22 +164,32 @@ private class ShuffleStatus(
invalidateSerializedMapOutputStatusCache()
}
mapStatuses(mapIndex) = status
+ mapIdToMapIndex(status.mapId) = mapIndex
+ }
+
+ /**
+ * Get the map output that corresponding to a given mapId.
+ */
+ def getMapStatus(mapId: Long): Option[MapStatus] = withReadLock {
+ mapIdToMapIndex.get(mapId).map(mapStatuses(_))
}
/**
* Update the map output location (e.g. during migration).
*/
def updateMapOutput(mapId: Long, bmAddress: BlockManagerId): Unit =
withWriteLock {
try {
- val mapStatusOpt = mapStatuses.find(x => x != null && x.mapId == mapId)
+ val mapIndex = mapIdToMapIndex.get(mapId)
+ val mapStatusOpt = Option(mapIndex.map(mapStatuses(_)).getOrElse(null))
Review Comment:
```suggestion
val mapStatusOpt = mapIndex.map(mapStatuses(_)
```
##########
core/src/main/scala/org/apache/spark/MapOutputTracker.scala:
##########
@@ -157,22 +164,32 @@ private class ShuffleStatus(
invalidateSerializedMapOutputStatusCache()
}
mapStatuses(mapIndex) = status
+ mapIdToMapIndex(status.mapId) = mapIndex
+ }
+
+ /**
+ * Get the map output that corresponding to a given mapId.
+ */
+ def getMapStatus(mapId: Long): Option[MapStatus] = withReadLock {
+ mapIdToMapIndex.get(mapId).map(mapStatuses(_))
}
/**
* Update the map output location (e.g. during migration).
*/
def updateMapOutput(mapId: Long, bmAddress: BlockManagerId): Unit =
withWriteLock {
try {
- val mapStatusOpt = mapStatuses.find(x => x != null && x.mapId == mapId)
+ val mapIndex = mapIdToMapIndex.get(mapId)
+ val mapStatusOpt = Option(mapIndex.map(mapStatuses(_)).getOrElse(null))
mapStatusOpt match {
case Some(mapStatus) =>
logInfo(s"Updating map output for ${mapId} to ${bmAddress}")
mapStatus.updateLocation(bmAddress)
invalidateSerializedMapOutputStatusCache()
case None =>
- val index = mapStatusesDeleted.indexWhere(x => x != null && x.mapId
== mapId)
- if (index >= 0 && mapStatuses(index) == null) {
+ if (mapIndex.nonEmpty && mapStatuses(mapIndex.get) == null &&
+ mapStatusesDeleted(mapIndex.get).mapId == mapId) {
Review Comment:
We dont need to check for `mapStatuses(mapIndex.get) == null` - it should
already be `null` since we are in the `None` case.
```suggestion
if (mapIndex.map(mapStatusesDeleted).exists(_.mapId == mapId) {
```
--
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: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
