jsun98 commented on a change in pull request #6431: Add Kinesis Indexing
Service to core Druid
URL: https://github.com/apache/incubator-druid/pull/6431#discussion_r238004628
##########
File path:
extensions-core/kinesis-indexing-service/src/main/java/org/apache/druid/indexing/kinesis/KinesisRecordSupplier.java
##########
@@ -213,39 +231,55 @@ private Runnable getRecordRunnable()
}
catch (ProvisionedThroughputExceededException e) {
long retryMs = Math.max(PROVISIONED_THROUGHPUT_EXCEEDED_BACKOFF_MS,
fetchDelayMillis);
- log.warn("Exceeded provisioned throughput, retrying in [%,dms]",
retryMs);
rescheduleRunnable(retryMs);
}
catch (Throwable e) {
log.error(e, "getRecordRunnable exception, retrying in [%,dms]",
EXCEPTION_RETRY_DELAY_MS);
rescheduleRunnable(EXCEPTION_RETRY_DELAY_MS);
}
+
};
}
private void rescheduleRunnable(long delayMillis)
{
if (started && !stopRequested) {
- scheduledExec.schedule(getRecordRunnable(), delayMillis,
TimeUnit.MILLISECONDS);
+ try {
+ scheduledExec.schedule(getRecordRunnable(), delayMillis,
TimeUnit.MILLISECONDS);
+ }
+ catch (RejectedExecutionException e) {
+ log.info(
Review comment:
this happens whenever a user does a seek/seekToEarliest/seekToLatest, in
which case the current implementation will stop all scheduledExec jobs and
filter out any records in the buffer that has the same shardId as the shards
being seeked to. So it does happens quite often.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on 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]
