dengziming commented on code in PR #14595:
URL: https://github.com/apache/kafka/pull/14595#discussion_r1446812841
##########
core/src/test/scala/integration/kafka/api/RackAwareAutoTopicCreationTest.scala:
##########
@@ -35,31 +37,37 @@ class RackAwareAutoTopicCreationTest extends
KafkaServerTestHarness with RackAwa
overridingProps.put(KafkaConfig.NumPartitionsProp, numPartitions.toString)
overridingProps.put(KafkaConfig.DefaultReplicationFactorProp,
replicationFactor.toString)
+
def generateConfigs =
(0 until numServers) map { node =>
- TestUtils.createBrokerConfig(node, zkConnect, enableControlledShutdown =
false, rack = Some((node / 2).toString))
+ TestUtils.createBrokerConfig(node, zkConnectOrNull,
enableControlledShutdown = false, rack = Some((node / 2).toString))
} map (KafkaConfig.fromProps(_, overridingProps))
private val topic = "topic"
- @Test
- def testAutoCreateTopic(): Unit = {
+ @ParameterizedTest(name = TestInfoUtils.TestWithParameterizedQuorumName)
+ @ValueSource(strings = Array("zk")) // TODO Partition leader is not evenly
distributed in kraft mode, see KAFKA-15354
Review Comment:
I tried to fix it but we are using a different way to assign leader in kraft
mode, the ideas behind the two algorithms are similar but the implementations
are different.
And the new one, `StripedReplicaPlacer`, is clearer, but it only ensure
leader are distributed evenly across racks, and don't ensure leaders are evenly
distributed across nodes. I wonder whether it is worth evolving it to be
consistent with the older one. cc @cmccabe
--
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
