soarez commented on code in PR #14838:
URL: https://github.com/apache/kafka/pull/14838#discussion_r1406207344
##########
server-common/src/main/java/org/apache/kafka/common/DirectoryId.java:
##########
@@ -121,8 +122,38 @@ public static Map<Integer, Uuid> createAssignmentMap(int[]
replicas, Uuid[] dire
* Create an array with the specified number of entries set to {@link
#UNASSIGNED}.
*/
public static Uuid[] unassignedArray(int length) {
+ return array(length, UNASSIGNED);
+ }
+
+ /**
+ * Create an array with the specified number of entries set to {@link
#MIGRATING}.
+ */
+ public static Uuid[] migratingArray(int length) {
+ return array(length, MIGRATING);
+ }
+
+ /**
+ * Create an array with the specified number of entries set to the
specified value.
+ */
+ private static Uuid[] array(int length, Uuid value) {
Uuid[] array = new Uuid[length];
- Arrays.fill(array, UNASSIGNED);
+ Arrays.fill(array, value);
return array;
}
+
+ /**
+ * Check if a directory is online, given a sorted list of online
directories.
+ * @param dir The directory to check
+ * @param sortedOnlineDirs The sorted list of online directories
+ * @return true if the directory is considered online,
false otherwise
+ */
+ public static boolean isOnline(Uuid dir, List<Uuid> sortedOnlineDirs) {
+ if (UNASSIGNED.equals(dir) || MIGRATING.equals(dir)) {
Review Comment:
Can you clarify what you mean with "migration mode"?
If you mean zk->kraft migration, then I don't think so. Metadata records may
have been persisted in an older version, without directory assignment. That may
happen outside of a ZK->KRaft migration.
When such records are found, the default directory for every replica is
assumed to be `MIGRATING`.
See in #14820 :
* `new PartitionRegistration(PartitionRecord record)`
* `partitionRegistration.merge(PartitionChangeRecord record)`
* `partitionRegistration.toRecord(...)`
* `partitionChangeBuilder.setAssignmentChanges()`
--
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
