AndrewJSchofield commented on code in PR #19443:
URL: https://github.com/apache/kafka/pull/19443#discussion_r2047031680
##########
share-coordinator/src/main/java/org/apache/kafka/coordinator/share/ShareCoordinatorShard.java:
##########
@@ -574,6 +575,46 @@ public
CoordinatorResult<InitializeShareGroupStateResponseData, CoordinatorRecor
return new CoordinatorResult<>(List.of(record), responseData);
}
+ /**
+ * Iterates over the soft state to determine the share partitions whose
last snapshot is
+ * older than the allowed time interval. The candidate share partitions
are force snapshotted.
+ *
+ * @return A result containing snapshot records, if any, and a void
response.
+ */
+ public CoordinatorResult<Void, CoordinatorRecord> snapshotColdPartitions()
{
+ long coldSnapshottedPartitionsCount = shareStateMap.values().stream()
+ .filter(shareGroupOffset -> shareGroupOffset.createTimestamp() -
shareGroupOffset.writeTimestamp() != 0)
+ .count();
+
+ // If all share partitions are snapshotted, it means that
+ // system is quiet and cold snapshotting will not help much.
+ if (coldSnapshottedPartitionsCount != 0 &&
coldSnapshottedPartitionsCount == shareStateMap.size()) {
Review Comment:
I'm uncertain about this test. If the share state map is empty, then this if
condition will be false, and we will proceed to do zero snapshots. Surely, this
is a situation which should skip the snapshotting. What's wrong with just
`coldSnapshottedPartitionsCount == shareStateMap.size()`? I know the code
previously did that, but it seems a better test than what is there now.
--
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: jira-unsubscr...@kafka.apache.org
For queries about this service, please contact Infrastructure at:
us...@infra.apache.org
