artemlivshits commented on code in PR #14705:
URL: https://github.com/apache/kafka/pull/14705#discussion_r1388948636
##########
group-coordinator/src/main/java/org/apache/kafka/coordinator/group/runtime/CoordinatorRuntime.java:
##########
@@ -659,9 +660,21 @@ public TopicPartition key() {
*/
@Override
public void run() {
+ try {
+ runAsync().get();
Review Comment:
To be honest, if we implement proper concurrency granularity for groups
(serialize group updates [not whole partition], keep read "lock" on groups
during commit updates) I'm not sure if we'd get much extra perf gain from
piercing the appendRecords abstraction to implement pipelining. Then we could
get rid of the timeline snapshot structure and hooking into replication
pipeline to listen for HWM updates; we could just do appendRecords and wait for
completion. Then we could completely decouple group coordinator logic from the
storage stack and make it simpler.
--
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]
