[
https://issues.apache.org/jira/browse/KAFKA-21153?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Ken Huang resolved KAFKA-21153.
-------------------------------
Resolution: Fixed
> ClusterInstance#waitTopicDeletion should not wait the file cleanup
> ------------------------------------------------------------------
>
> Key: KAFKA-21153
> URL: https://issues.apache.org/jira/browse/KAFKA-21153
> Project: Kafka
> Issue Type: Improvement
> Reporter: Chia-Ping Tsai
> Assignee: Kuan Po Tseng
> Priority: Minor
>
> {code:java}
> TestUtils.waitForCondition(() -> brokers.stream().allMatch(broker ->
> broker.config().logDirs().stream().allMatch(logDir ->
> Arrays.stream(Objects.requireNonNull(new
> File(logDir).list())).noneMatch(partitionDirectoryName ->
>
> partitionDirectoryName.startsWith(topicPartition.topic() + "-" +
> topicPartition.partition()) &&
>
> partitionDirectoryName.endsWith(UnifiedLog.DELETE_DIR_SUFFIX)))
> ), "Failed to hard-delete the delete directory"); {code}
> This hard-delete check forces callers to set the internal configs
> {{log.initial.task.delay.ms}} and {{log.segment.delete.delay.ms}} to avoid
> timeout. I took a glance at all callers and none of them actually care about
> the file cleanup. Hence, we should remove the wait logic and the configs from
> the callers.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
