sajjad-moradi commented on a change in pull request #7886:
URL: https://github.com/apache/pinot/pull/7886#discussion_r768247751
##########
File path:
pinot-core/src/main/java/org/apache/pinot/core/data/manager/realtime/RealtimeTableDataManager.java
##########
@@ -340,9 +340,18 @@ public void addSegment(String segmentName, TableConfig
tableConfig, IndexLoading
PartitionUpsertMetadataManager partitionUpsertMetadataManager =
_tableUpsertMetadataManager != null ? _tableUpsertMetadataManager
.getOrCreatePartitionManager(partitionGroupId) : null;
- segmentDataManager =
- new LLRealtimeSegmentDataManager(segmentZKMetadata, tableConfig,
this, _indexDir.getAbsolutePath(),
- indexLoadingConfig, schema, llcSegmentName, semaphore,
_serverMetrics, partitionUpsertMetadataManager);
+ try {
+ segmentDataManager =
+ new LLRealtimeSegmentDataManager(segmentZKMetadata, tableConfig,
this, _indexDir.getAbsolutePath(),
+ indexLoadingConfig, schema, llcSegmentName, semaphore,
_serverMetrics, partitionUpsertMetadataManager);
+ } catch (Exception e) {
+ // In case of exception thrown here, segment goes to ERROR state. Then
any attempt to reset the segment from
+ // ERROR -> OFFLINE -> CONSUMING via Helix Admin fails because the
semaphore never gets released and the
+ // semaphore.acquire() is called in constructor of
LLRealtimeSegmentDataManager. Hence releasing the semaphore
+ // here to unblock reset operation via Helix Admin.
+ semaphore.release();
+ throw e;
Review comment:
Semantically the two ways are the same. I didn't want to put a giant try
block which spans more than 100 lines of code, but it's a valid point to
release the semaphore next to the code where it's acquired. Will refactor the
code shortly.
--
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]
