回复：Flink 1.6 Job fails with IllegalStateException: Buffer pool is destroyed.

Hi, I think the problem in the attched image is not the root cause of your job failure. It must exist other task or TaskManager failures, then all the related tasks will be cancelled by job manager, and the problem in attched image is just caused by task cancelled. You can review the log of jo

Re: Flink 1.6 Job fails with IllegalStateException: Buffer pool is destroyed.

Hi Bill, Can you provide more information, such as whether Checkpoint is enabled and whether exact-once is specified, and whether there is back pressure generated in the Flink web UI. Here is a ticket that also gives feedback to this question. [1] Stackoverflow has also been asked the same questio

Re: After OutOfMemoryError State can not be readed

Hi, what I can say is that any failures like OOMs should not corrupt checkpoint files, because only successfully completed checkpoints are used for recovery by the job manager. Just to get a bit more info, are you using full or incremental checkpoints? Unfortunately, it is a bit hard to say fro

Re: How to customize schedule mode and result partition type?

Sorry to attach this message here, but if someone see this email please reply an ack. Without any reply I wonder if this email has been received by someone in the mail list. Best, tison.

Re: Operator metrics do not get unregistered after job finishes

Hi Vino, The log shows no problems. The problem can be reproduced easily. I created https://issues.apache.org/jira/browse/FLINK-10300 . Best, Helmut > On 18. Aug 2018, at 04:53, vino yang wrote: > > Hi Helmut, > > Is the metrics of all the

Re: Flink 1.6 Job fails with IllegalStateException: Buffer pool is destroyed.

Thank you for you advice. I had not noticed that the log level was set to WARN. INFO logs suggest that the job fails because of akka timeout and the root cause is long gc pause. On Fri, Sep 7, 2018 at 5:43 PM Zhijiang(wangzhijiang999) < wangzhijiang...@aliyun.com> wrote: > You may need to config

Re: Setting Flink Monitoring API Port on YARN Cluster

Hi Gary, Thank you so much for the detailed explanation and links. Extremely helpful. For all others interested, this is also available through the YARN CLI command `yarn application -status {appId}`. Once again, thanks for your help! Austin On Fri, Sep 7, 2018, 2:24 AM Gary Yao wrote: > Hi A

Re: Behaviour of Process Window Function

Hi Harshvardhan, *> 1) Does the state in the process window function qualify as KeyedState or OperatorState? * KeyedState *> We want to be able to rehydrate the guava cache at the beginning of each window by making an external rest call and clear the cache at the end of that respective window. Ho

Flink Failing to Connect to Kafka org.apache.kafka.common.protocol.types.SchemaException: Error computing size for field 'topics': java.lang.NullPointerException

Hi, I have a Flink 1.4.0 cluster running on OpenShift with a job that connects to a Kafka 0.11.0.1 cluster in the same Openshift project. The job reads from one topic and writes to two others. The job deploys OK but when it starts up it immediately crashes with the following exception org.apache

maxOutOfOrderness

Hello, Does maxOutOfOrderness affect performance? Setting it to Long.MAX_VALUE doesn't affect performance, so either flink is really fast in my simple pipeline or my understandi

Re: maxOutOfOrderness

Hi Nicos, Setting it to Long.MAX_VALUE makes watermark always smaller than timestamp. In this case, the event time window will never be triggered. It is meaningless. Best, Hequn On Fri, Sep 7, 2018 at 11:17 PM Nicos Maris wrote: > Hello, > > > Does maxOutOfOrderness >

RE: Flink Failing to Connect to Kafka org.apache.kafka.common.protocol.types.SchemaException: Error computing size for field 'topics': java.lang.NullPointerException

Sorry, there was a code issue, where I was creating a kafka 10 consumer. Problem solved. From: Oliver Buckley-Salmon Sent: 07 September 2018 15:04 To: user@flink.apache.org Subject: Flink Failing to Connect to Kafka org.apache.kafka.common.protocol.types.SchemaException: Error computing size for

Re: After OutOfMemoryError State can not be readed

Hi Stefan, Vino, Thanks for your answers. We are using full checkpointing, not incremental. We are using custom serializers for the operators state classes, The serializers perform encryption before writing and decrypt when reading. The serializer is stateless. We register the Serializers by using