kfaraz commented on code in PR #14407:
URL: https://github.com/apache/druid/pull/14407#discussion_r1349610904
##########
indexing-service/src/main/java/org/apache/druid/indexing/common/task/AbstractBatchIndexTask.java:
##########
@@ -443,6 +466,42 @@ protected boolean tryTimeChunkLock(TaskActionClient
client, List<Interval> inter
return true;
}
+ private TaskLockHelper createLockHelper(LockGranularity lockGranularity)
+ {
+ return new TaskLockHelper(
+ lockGranularity == LockGranularity.SEGMENT,
+ determineLockType(lockGranularity)
+ );
+ }
+
+ /**
+ * Determines the type of lock to use with the given lock granularity.
+ */
+ private TaskLockType determineLockType(LockGranularity lockGranularity)
+ {
+ if (lockGranularity == LockGranularity.SEGMENT) {
+ return TaskLockType.EXCLUSIVE;
+ }
+
+ final String contextLockType = getContextValue(Tasks.TASK_LOCK_TYPE);
+ final TaskLockType lockType;
+ if (contextLockType == null) {
+ lockType = getContextValue(Tasks.USE_SHARED_LOCK, false)
+ ? TaskLockType.SHARED : TaskLockType.EXCLUSIVE;
+ } else {
+ lockType = TaskLockType.valueOf(contextLockType);
+ }
+
+ final IngestionMode ingestionMode = getIngestionMode();
+ if ((lockType == TaskLockType.SHARED || lockType == TaskLockType.APPEND)
+ && ingestionMode != IngestionMode.APPEND) {
+ // Lock types SHARED and APPEND are allowed only in APPEND ingestion mode
+ return Tasks.DEFAULT_TASK_LOCK_TYPE;
+ } else {
+ return lockType;
Review Comment:
Yeah, makes sense. And I suppose a validation should be done at the time of
submitting the ingestion spec as well so that an incorrectly configured task
fails fast.
--
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]
