[ 
https://issues.apache.org/jira/browse/KAFKA-7136?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16534201#comment-16534201
 ] 

Ismael Juma commented on KAFKA-7136:
------------------------------------

I think the important question is whether custom metrics reporters are likely 
to do the same as PushHttpMetricsReporter. If so, that could be very bad. We 
have to be very careful about not introducing regressions in bug fix releases 
so it's a tough one.

> PushHttpMetricsReporter may deadlock when processing metrics changes
> --------------------------------------------------------------------
>
>                 Key: KAFKA-7136
>                 URL: https://issues.apache.org/jira/browse/KAFKA-7136
>             Project: Kafka
>          Issue Type: Bug
>          Components: metrics
>    Affects Versions: 1.1.0, 2.0.0
>            Reporter: Rajini Sivaram
>            Assignee: Rajini Sivaram
>            Priority: Blocker
>             Fix For: 2.0.0
>
>
> We noticed a deadlock in {{PushHttpMetricsReporter}}. Locking for metrics was 
> changed under KAFKA-6765 to avoid {{NullPointerException}} in metrics 
> reporters due to concurrent read and updates. {{PushHttpMetricsReporter}} 
> requires a lock to process metrics registration that is invoked while holding 
> the sensor lock. It also reads metrics attempting to acquire sensor lock 
> while holding its lock (inverse order). This resulted in the deadlock below.
> {quote}Found one Java-level deadlock:
>  Java stack information for the threads listed above:
>  ===================================================
>  "StreamThread-7":
>  at 
> org.apache.kafka.tools.PushHttpMetricsReporter.metricChange(PushHttpMetricsReporter.java:144)
>  - waiting to lock <0x0000000655a54310> (a java.lang.Object)
>  at org.apache.kafka.common.metrics.Metrics.registerMetric(Metrics.java:563)
>  - locked <0x0000000655a44a28> (a org.apache.kafka.common.metrics.Metrics)
>  at org.apache.kafka.common.metrics.Sensor.add(Sensor.java:236)
>  - locked <0x000000065629c170> (a org.apache.kafka.common.metrics.Sensor)
>  at org.apache.kafka.common.metrics.Sensor.add(Sensor.java:217)
>  at 
> org.apache.kafka.common.network.Selector$SelectorMetrics.maybeRegisterConnectionMetrics(Selector.java:1016)
>  at 
> org.apache.kafka.common.network.Selector.pollSelectionKeys(Selector.java:462)
>  at org.apache.kafka.common.network.Selector.poll(Selector.java:425)
>  at org.apache.kafka.clients.NetworkClient.poll(NetworkClient.java:510)
>  at 
> org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.poll(ConsumerNetworkClient.java:271)
>  at 
> org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.poll(ConsumerNetworkClient.java:242)
>  at 
> org.apache.kafka.clients.consumer.internals.ConsumerNetworkClient.poll(ConsumerNetworkClient.java:218)
>  at 
> org.apache.kafka.clients.consumer.internals.Fetcher.getTopicMetadata(Fetcher.java:274)
>  at 
> org.apache.kafka.clients.consumer.internals.Fetcher.getAllTopicMetadata(Fetcher.java:254)
>  at 
> org.apache.kafka.clients.consumer.KafkaConsumer.listTopics(KafkaConsumer.java:1820)
>  at 
> org.apache.kafka.clients.consumer.KafkaConsumer.listTopics(KafkaConsumer.java:1798)
>  at 
> org.apache.kafka.streams.processor.internals.StoreChangelogReader.refreshChangelogInfo(StoreChangelogReader.java:224)
>  at 
> org.apache.kafka.streams.processor.internals.StoreChangelogReader.initialize(StoreChangelogReader.java:121)
>  at 
> org.apache.kafka.streams.processor.internals.StoreChangelogReader.restore(StoreChangelogReader.java:74)
>  at 
> org.apache.kafka.streams.processor.internals.TaskManager.updateNewAndRestoringTasks(TaskManager.java:317)
>  at 
> org.apache.kafka.streams.processor.internals.StreamThread.runOnce(StreamThread.java:824)
>  at 
> org.apache.kafka.streams.processor.internals.StreamThread.runLoop(StreamThread.java:767)
>  at 
> org.apache.kafka.streams.processor.internals.StreamThread.run(StreamThread.java:736)
>  "pool-17-thread-1":
>  at 
> org.apache.kafka.common.metrics.KafkaMetric.measurableValue(KafkaMetric.java:82)
>  - waiting to lock <0x000000065629c170> (a 
> org.apache.kafka.common.metrics.Sensor)
>  at org.apache.kafka.common.metrics.KafkaMetric.value(KafkaMetric.java:58)
>  at 
> org.apache.kafka.tools.PushHttpMetricsReporter$HttpReporter.run(PushHttpMetricsReporter.java:177)
>  - locked <0x0000000655a54310> (a java.lang.Object)
>  at java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:511)
>  at java.util.concurrent.FutureTask.runAndReset(FutureTask.java:308)
>  at 
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.access$301(ScheduledThreadPoolExecutor.java:180)
>  at 
> java.util.concurrent.ScheduledThreadPoolExecutor$ScheduledFutureTask.run(ScheduledThreadPoolExecutor.java:294)
>  at 
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
>  at 
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
>  at java.lang.Thread.run(Thread.java:748)
> Found 1 deadlock.
> {quote}



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)

Reply via email to