Hi Binh, Could you try upgrading Flink's Java runtime? It was previously reported that upgrading to jdk1.8.0_251 was solving the problem.
Piotrek pt., 9 paź 2020 o 19:41 Binh Nguyen Van <binhn...@gmail.com> napisał(a): > Hi, > > Thank you for helping me! > The code is compiled on > > java version "1.8.0_161" > Java(TM) SE Runtime Environment (build 1.8.0_161-b12) > Java HotSpot(TM) 64-Bit Server VM (build 25.161-b12, mixed mode) > > But I just checked our Hadoop and its Java version is > > java version "1.8.0_77" > Java(TM) SE Runtime Environment (build 1.8.0_77-b03) > Java HotSpot(TM) 64-Bit Server VM (build 25.77-b03, mixed mode) > > Thanks > -Binh > > On Fri, Oct 9, 2020 at 10:23 AM Piotr Nowojski <pnowoj...@apache.org> > wrote: > >> Hi, >> >> One more thing. It looks like it's not a Flink issue, but some JDK bug. >> Others reported that upgrading JDK version (for example to jdk1.8.0_251) >> seemed to be solving this problem. What JDK version are you using? >> >> Piotrek >> >> pt., 9 paź 2020 o 17:59 Piotr Nowojski <pnowoj...@apache.org> napisał(a): >> >>> Hi, >>> >>> Thanks for reporting the problem. I think this is a known issue [1] on >>> which we are working to fix. >>> >>> Piotrek >>> >>> [1] https://issues.apache.org/jira/browse/FLINK-18196 >>> >>> pon., 5 paź 2020 o 08:54 Binh Nguyen Van <binhn...@gmail.com> >>> napisał(a): >>> >>>> Hi, >>>> >>>> I have a streaming job that is written in Apache Beam and uses Flink as >>>> its runner. The job is working as expected for about 15 hours and then it >>>> started to have checkpointing error. The error message looks like this >>>> >>>> java.lang.Exception: Could not perform checkpoint 910 for operator Source: >>>> <source-name> (8/60). >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTask.triggerCheckpoint(StreamTask.java:785) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTask.lambda$triggerCheckpointAsync$3(StreamTask.java:760) >>>> at java.util.concurrent.FutureTask.run(FutureTask.java:266) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTaskActionExecutor$SynchronizedStreamTaskActionExecutor.run(StreamTaskActionExecutor.java:87) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.mailbox.Mail.run(Mail.java:78) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.mailbox.MailboxProcessor.processMail(MailboxProcessor.java:261) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.mailbox.MailboxProcessor.runMailboxLoop(MailboxProcessor.java:186) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTask.runMailboxLoop(StreamTask.java:485) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTask.invoke(StreamTask.java:469) >>>> at org.apache.flink.runtime.taskmanager.Task.doRun(Task.java:708) >>>> at org.apache.flink.runtime.taskmanager.Task.run(Task.java:533) >>>> at java.lang.Thread.run(Thread.java:745) >>>> Caused by: java.lang.NullPointerException >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTask$CheckpointingOperation.executeCheckpointing(StreamTask.java:1394) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTask.checkpointState(StreamTask.java:974) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTask.lambda$performCheckpoint$5(StreamTask.java:870) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTaskActionExecutor$SynchronizedStreamTaskActionExecutor.runThrowing(StreamTaskActionExecutor.java:94) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTask.performCheckpoint(StreamTask.java:843) >>>> at >>>> org.apache.flink.streaming.runtime.tasks.StreamTask.triggerCheckpoint(StreamTask.java:776) >>>> ... 11 more >>>> >>>> When this happened, I have to stop the job and then start it again, and >>>> then 15 hours later the issue happens again. >>>> >>>> Here are some additional information >>>> >>>> - Flink version is 1.10.1 >>>> - Job reads data from Kafka, transform, and then writes to Kafka >>>> - There are 6 tasks with the parallelism of 60 each (each task >>>> reads from 1 Kafka topic) >>>> - The job is deployed to run on YARN with 60 task managers and each >>>> task manager has 1 slot >>>> - The State backend is filesystem and HDFS is the storage (Doesn’t >>>> seem to related to the type of state backend since the issue also >>>> happened >>>> when I use memory as the state backend) >>>> - The checkpointing interval is 60 seconds (The longest duration of >>>> the normal checkpoint as shown in Flink UI is 14 seconds) >>>> - The minimum pause between checkpoints is 30 seconds >>>> - Hadoop cluster is Kerberized but Kafka is not. Keytab and >>>> principal are set in the Flink configuration file >>>> >>>> Can someone please help? >>>> >>>> Thanks >>>> -Binh >>>> >>>
