[ https://issues.apache.org/jira/browse/FLINK-5048?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15662984#comment-15662984 ]
ASF GitHub Bot commented on FLINK-5048: --------------------------------------- Github user tzulitai commented on a diff in the pull request: https://github.com/apache/flink/pull/2789#discussion_r87744744 --- Diff: flink-streaming-connectors/flink-connector-kafka-0.9/src/test/java/org/apache/flink/streaming/connectors/kafka/Kafka09FetcherTest.java --- @@ -323,8 +329,154 @@ else if (partition.topic().equals("another")) { // check that there were no errors in the fetcher final Throwable caughtError = error.get(); - if (caughtError != null) { + if (caughtError != null && !(caughtError instanceof Handover.ClosedException)) { --- End diff -- Perhaps we should be suppressing the fetcher of throwing `Handover.ClosedException`, as it doesn't really make sense to the main thread. Please see my above comments. > Kafka Consumer (0.9/0.10) threading model leads problematic cancellation > behavior > --------------------------------------------------------------------------------- > > Key: FLINK-5048 > URL: https://issues.apache.org/jira/browse/FLINK-5048 > Project: Flink > Issue Type: Bug > Components: Kafka Connector > Affects Versions: 1.1.3 > Reporter: Stephan Ewen > Assignee: Stephan Ewen > Fix For: 1.2.0 > > > The {{FLinkKafkaConsumer}} (0.9 / 0.10) spawns a separate thread that > operates the KafkaConsumer. That thread is shielded from interrupts, because > the Kafka Consumer has not been handling thread interrupts well. > Since that thread is also the thread that emits records, it may block in the > network stack (backpressure) or in chained operators. The later case leads to > situations where cancellations get very slow unless that thread would be > interrupted (which it cannot be). > I propose to change the thread model as follows: > - A spawned consumer thread pull from the KafkaConsumer and pushes its > pulled batch of records into a blocking queue (size one) > - The main thread of the task will pull the record batches from the > blocking queue and emit the records. > This allows actually for some additional I/O overlay while limiting the > additional memory consumption - only two batches are ever held, one being > fetched and one being emitted. -- This message was sent by Atlassian JIRA (v6.3.4#6332)