[GitHub] [kafka] showuon commented on a diff in pull request #12596: KAFKA-14203 Don't make snapshots on broker after metadata errors

[GitBox <gi...@apache.org>]

showuon commented on code in PR #12596:
URL: https://github.com/apache/kafka/pull/12596#discussion_r964347667


##########
core/src/main/scala/kafka/server/metadata/BrokerMetadataListener.scala:
##########
@@ -41,8 +43,19 @@ class BrokerMetadataListener(
   val maxBytesBetweenSnapshots: Long,
   val snapshotter: Option[MetadataSnapshotter],
   brokerMetrics: BrokerServerMetrics,
-  metadataLoadingFaultHandler: FaultHandler
+  _metadataLoadingFaultHandler: FaultHandler
 ) extends RaftClient.Listener[ApiMessageAndVersion] with KafkaMetricsGroup {
+
+  private val metadataFaultOccurred = new AtomicBoolean(false)
+  private val metadataLoadingFaultHandler: FaultHandler = new FaultHandler() {
+    override def handleFault(failureMessage: String, cause: Throwable): RuntimeException = {
+      if (metadataFaultOccurred.compareAndSet(false, true)) {
+        error("Disabling metadata snapshots until this broker is restarted.")
+      }
+      _metadataLoadingFaultHandler.handleFault(failureMessage, cause)
+    }

Review Comment:
   Agree with @jsancio , or maybe we can just shutdown the broker with error, since the metadata cache in this broker is already wrong?



-- 
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-unsubscribe@kafka.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org