samarthjain commented on a change in pull request #7088: Improve parallelism of
zookeeper based segment change processing
URL: https://github.com/apache/incubator-druid/pull/7088#discussion_r268068243
##########
File path:
server/src/main/java/org/apache/druid/server/coordinator/CuratorLoadQueuePeon.java
##########
@@ -290,128 +307,119 @@ private void processSegmentChangeRequest()
// We do not create the existence watcher first, because then it will
fire when we create the
// node and we'll have the same race when trying to refresh that
watcher.
curator.create().withMode(CreateMode.EPHEMERAL).forPath(path,
noopPayload);
-
- entryRemoved(path);
+ entryRemoved(segmentHolder, path);
}
}
+ catch (KeeperException.NodeExistsException ne) {
+ // This is expected when historicals haven't yet picked up processing
this segment and coordinator
+ // tries reassigning it to the same node.
+ log.warn(ne, "ZK node already exists because segment change request
hasn't yet been processed");
+ failAssign(segmentHolder);
+ }
catch (Exception e) {
- failAssign(e);
+ failAssign(segmentHolder, e);
}
}
- private void actionCompleted()
+ private void actionCompleted(SegmentHolder segmentHolder)
{
- if (currentlyProcessing != null) {
- switch (currentlyProcessing.getType()) {
- case LOAD:
- segmentsToLoad.remove(currentlyProcessing.getSegment());
- queuedSize.addAndGet(-currentlyProcessing.getSegmentSize());
- break;
- case DROP:
- segmentsToDrop.remove(currentlyProcessing.getSegment());
- break;
- default:
- throw new UnsupportedOperationException();
- }
-
- final List<LoadPeonCallback> callbacks =
currentlyProcessing.getCallbacks();
- currentlyProcessing = null;
- callBackExecutor.execute(
- () -> executeCallbacks(callbacks)
- );
+ switch (segmentHolder.getType()) {
+ case LOAD:
+ segmentsToLoad.remove(segmentHolder.getSegment());
+ queuedSize.addAndGet(-segmentHolder.getSegmentSize());
+ break;
+ case DROP:
+ segmentsToDrop.remove(segmentHolder.getSegment());
+ break;
+ default:
+ throw new UnsupportedOperationException();
}
+
+ callBackExecutor.execute(
+ () -> executeCallbacks(segmentHolder)
+ );
}
+
@Override
public void start()
{
- ScheduledExecutors.scheduleAtFixedRate(
- processingExecutor,
- config.getLoadQueuePeonRepeatDelay(),
- config.getLoadQueuePeonRepeatDelay(),
- () -> {
- processSegmentChangeRequest();
-
- if (stopped) {
- return ScheduledExecutors.Signal.STOP;
- } else {
- return ScheduledExecutors.Signal.REPEAT;
- }
- }
- );
+ for (int i = 0; i < numProcessingQueues; i++) {
+ processingExecutor.scheduleAtFixedRate(
+ new SegmentChangeProcessor(segmentProcessingQueues[i]),
+ 0,
+ config.getCuratorCreateZkNodesRepeatDelay().getMillis(),
Review comment:
I took a closer look at the code and I realized that I was making the wrong
assumption that there is only one peon running on the coordinator. Instead,
every historical node gets its own load queue peon on the coordinator.
Considering this, I don't think we need to run multiple tasks per peon for
creating zk nodes. Just having one task should be enough. Made this change in
the latest commit.
----------------------------------------------------------------
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.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
