[jira] [Comment Edited] (HDFS-14655) [SBN Read] Namenode crashes if one of The JN is down
[ https://issues.apache.org/jira/browse/HDFS-14655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17007828#comment-17007828 ] Chen Liang edited comment on HDFS-14655 at 1/3/20 11:25 PM: Although it's a different message, checked again, does look like HDFS-14934 should fix this too. Thanks for the pointer [~ayushtkn]! was (Author: vagarychen): HDFS-14934 does look like the fix. Thanks for the pointer [~ayushtkn]! > [SBN Read] Namenode crashes if one of The JN is down > > > Key: HDFS-14655 > URL: https://issues.apache.org/jira/browse/HDFS-14655 > Project: Hadoop HDFS > Issue Type: Bug >Affects Versions: 3.3.0 >Reporter: Harshakiran Reddy >Assignee: Ayush Saxena >Priority: Critical > Fix For: 2.10.0, 3.3.0, 3.1.4, 3.2.2 > > Attachments: HDFS-14655-01.patch, HDFS-14655-02.patch, > HDFS-14655-03.patch, HDFS-14655-04.patch, HDFS-14655-05.patch, > HDFS-14655-06.patch, HDFS-14655-07.patch, HDFS-14655-08.patch, > HDFS-14655-branch-2-01.patch, HDFS-14655-branch-2-02.patch, > HDFS-14655.poc.patch > > > {noformat} > 2019-07-04 17:35:54,064 | INFO | Logger channel (from parallel executor) to > XXX/XXX | Retrying connect to server: XXX/XXX. Already tried > 9 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10, > sleepTime=1000 MILLISECONDS) | Client.java:975 > 2019-07-04 17:35:54,087 | FATAL | Edit log tailer | Unknown error encountered > while tailing edits. Shutting down standby NN. | EditLogTailer.java:474 > java.lang.OutOfMemoryError: unable to create new native thread > at java.lang.Thread.start0(Native Method) > at java.lang.Thread.start(Thread.java:717) > at > java.util.concurrent.ThreadPoolExecutor.addWorker(ThreadPoolExecutor.java:957) > at > java.util.concurrent.ThreadPoolExecutor.execute(ThreadPoolExecutor.java:1378) > at > com.google.common.util.concurrent.MoreExecutors$ListeningDecorator.execute(MoreExecutors.java:440) > at > com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56) > at > org.apache.hadoop.hdfs.qjournal.client.IPCLoggerChannel.getJournaledEdits(IPCLoggerChannel.java:565) > at > org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.getJournaledEdits(AsyncLoggerSet.java:272) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectRpcInputStreams(QuorumJournalManager.java:533) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectInputStreams(QuorumJournalManager.java:508) > at > org.apache.hadoop.hdfs.server.namenode.JournalSet.selectInputStreams(JournalSet.java:275) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1681) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1714) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer.doTailEdits(EditLogTailer.java:307) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.doWork(EditLogTailer.java:460) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.access$300(EditLogTailer.java:410) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread$1.run(EditLogTailer.java:427) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:360) > at > org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1709) > at > org.apache.hadoop.security.SecurityUtil.doAsLoginUserOrFatal(SecurityUtil.java:483) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.run(EditLogTailer.java:423) > 2019-07-04 17:35:54,112 | INFO | Edit log tailer | Exiting with status 1: > java.lang.OutOfMemoryError: unable to create new native thread | > ExitUtil.java:210 > {noformat} -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (HDFS-14655) [SBN Read] Namenode crashes if one of The JN is down
[ https://issues.apache.org/jira/browse/HDFS-14655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16935133#comment-16935133 ] Ayush Saxena edited comment on HDFS-14655 at 9/21/19 7:02 PM: -- Thanx [~vinayakumarb] for the review. Have made the suggested changes. Pls Review!! was (Author: ayushtkn): Thanx [~vinayakumarb]for the review. Have made the suggested changes. Pls Review!! > [SBN Read] Namenode crashes if one of The JN is down > > > Key: HDFS-14655 > URL: https://issues.apache.org/jira/browse/HDFS-14655 > Project: Hadoop HDFS > Issue Type: Bug >Affects Versions: 3.3.0 >Reporter: Harshakiran Reddy >Assignee: Ayush Saxena >Priority: Critical > Attachments: HDFS-14655-01.patch, HDFS-14655-02.patch, > HDFS-14655-03.patch, HDFS-14655-04.patch, HDFS-14655-05.patch, > HDFS-14655-06.patch, HDFS-14655-07.patch, HDFS-14655-08.patch, > HDFS-14655.poc.patch > > > {noformat} > 2019-07-04 17:35:54,064 | INFO | Logger channel (from parallel executor) to > XXX/XXX | Retrying connect to server: XXX/XXX. Already tried > 9 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10, > sleepTime=1000 MILLISECONDS) | Client.java:975 > 2019-07-04 17:35:54,087 | FATAL | Edit log tailer | Unknown error encountered > while tailing edits. Shutting down standby NN. | EditLogTailer.java:474 > java.lang.OutOfMemoryError: unable to create new native thread > at java.lang.Thread.start0(Native Method) > at java.lang.Thread.start(Thread.java:717) > at > java.util.concurrent.ThreadPoolExecutor.addWorker(ThreadPoolExecutor.java:957) > at > java.util.concurrent.ThreadPoolExecutor.execute(ThreadPoolExecutor.java:1378) > at > com.google.common.util.concurrent.MoreExecutors$ListeningDecorator.execute(MoreExecutors.java:440) > at > com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56) > at > org.apache.hadoop.hdfs.qjournal.client.IPCLoggerChannel.getJournaledEdits(IPCLoggerChannel.java:565) > at > org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.getJournaledEdits(AsyncLoggerSet.java:272) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectRpcInputStreams(QuorumJournalManager.java:533) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectInputStreams(QuorumJournalManager.java:508) > at > org.apache.hadoop.hdfs.server.namenode.JournalSet.selectInputStreams(JournalSet.java:275) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1681) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1714) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer.doTailEdits(EditLogTailer.java:307) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.doWork(EditLogTailer.java:460) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.access$300(EditLogTailer.java:410) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread$1.run(EditLogTailer.java:427) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:360) > at > org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1709) > at > org.apache.hadoop.security.SecurityUtil.doAsLoginUserOrFatal(SecurityUtil.java:483) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.run(EditLogTailer.java:423) > 2019-07-04 17:35:54,112 | INFO | Edit log tailer | Exiting with status 1: > java.lang.OutOfMemoryError: unable to create new native thread | > ExitUtil.java:210 > {noformat} -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (HDFS-14655) [SBN Read] Namenode crashes if one of The JN is down
[ https://issues.apache.org/jira/browse/HDFS-14655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16933585#comment-16933585 ] Ayush Saxena edited comment on HDFS-14655 at 9/19/19 7:48 PM: -- Thanx [~shv] and [~xkrogen]. I too verified by reverting and the test fails for me too without the fix. [~shv] May be you can try once commenting this line too in the test {{TestQuorumJournalManager}} : {code:java} // Don't retry connections - it just slows down the tests. conf.setInt(CommonConfigurationKeysPublic.IPC_CLIENT_CONNECT_MAX_RETRIES_KEY, 0); {code} And may be try increasing the number in the loop too, to some even bigger number: {code:java} for (int i = 0; i < 1000; i++) { {code} May be environment specific. Let me know if you still can't repro, was (Author: ayushtkn): Thanx [~shv] and [~xkrogen]. I too verified by reverting and the test fails for me too without the fix. [~shv] May be you can try once commenting this line too in the test {{TestQuorumJournalManager}} : {code:java} // Don't retry connections - it just slows down the tests. conf.setInt(CommonConfigurationKeysPublic.IPC_CLIENT_CONNECT_MAX_RETRIES_KEY, 0); {code} And may be try increasing the number in the loop too, to some even bigger number: {code:java} for (int i = 0; i < 1000; i++) { {code} And may be try increasing the number in the loop too, to some even bigger number: May be environment specific. Let me know if you still can't repro, > [SBN Read] Namenode crashes if one of The JN is down > > > Key: HDFS-14655 > URL: https://issues.apache.org/jira/browse/HDFS-14655 > Project: Hadoop HDFS > Issue Type: Bug >Affects Versions: 3.3.0 >Reporter: Harshakiran Reddy >Assignee: Ayush Saxena >Priority: Critical > Attachments: HDFS-14655-01.patch, HDFS-14655-02.patch, > HDFS-14655-03.patch, HDFS-14655-04.patch, HDFS-14655-05.patch, > HDFS-14655-06.patch, HDFS-14655-07.patch, HDFS-14655.poc.patch > > > {noformat} > 2019-07-04 17:35:54,064 | INFO | Logger channel (from parallel executor) to > XXX/XXX | Retrying connect to server: XXX/XXX. Already tried > 9 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10, > sleepTime=1000 MILLISECONDS) | Client.java:975 > 2019-07-04 17:35:54,087 | FATAL | Edit log tailer | Unknown error encountered > while tailing edits. Shutting down standby NN. | EditLogTailer.java:474 > java.lang.OutOfMemoryError: unable to create new native thread > at java.lang.Thread.start0(Native Method) > at java.lang.Thread.start(Thread.java:717) > at > java.util.concurrent.ThreadPoolExecutor.addWorker(ThreadPoolExecutor.java:957) > at > java.util.concurrent.ThreadPoolExecutor.execute(ThreadPoolExecutor.java:1378) > at > com.google.common.util.concurrent.MoreExecutors$ListeningDecorator.execute(MoreExecutors.java:440) > at > com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56) > at > org.apache.hadoop.hdfs.qjournal.client.IPCLoggerChannel.getJournaledEdits(IPCLoggerChannel.java:565) > at > org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.getJournaledEdits(AsyncLoggerSet.java:272) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectRpcInputStreams(QuorumJournalManager.java:533) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectInputStreams(QuorumJournalManager.java:508) > at > org.apache.hadoop.hdfs.server.namenode.JournalSet.selectInputStreams(JournalSet.java:275) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1681) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1714) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer.doTailEdits(EditLogTailer.java:307) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.doWork(EditLogTailer.java:460) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.access$300(EditLogTailer.java:410) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread$1.run(EditLogTailer.java:427) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:360) > at > org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1709) > at > org.apache.hadoop.security.SecurityUtil.doAsLoginUserOrFatal(SecurityUtil.java:483) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.run(EditLogTailer.java:423) > 2019-07-04 17:35:54,112 | INFO | Edit log t
[jira] [Comment Edited] (HDFS-14655) [SBN Read] Namenode crashes if one of The JN is down
[ https://issues.apache.org/jira/browse/HDFS-14655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16932900#comment-16932900 ] Konstantin Shvachko edited comment on HDFS-14655 at 9/18/19 11:07 PM: -- Hey guys I could not reproduce the problem with the new test. It passes even when I reverse the patch for QuorumJournalManager, QuorumCall, and IPCLoggerChannel. If we could replace the assert condition with {code} assertEquals(DFSConfigKeys.DFS_HA_TAILEDITS_NUM_THREADS_DEFAULT, num); {code} then it would fail on current code, but {{num ==1}} in the test with the patch. Is this expected? Should we add some failures in {{selectInputStreams()}} to involve use of more threads? was (Author: shv): Hey guys I could not reproduce the problem with the new test. It passes even when I reverse the patch for QuorumJournalManager, QuorumCall, and IPCLoggerChannel. If we could replace the assert condition with {code} assertEquals(DFSConfigKeys.DFS_HA_TAILEDITS_NUM_THREADS_DEFAULT, num); {code} then it would fail on current code, but {{num ==1}} in the test with the patch. Is this expected? > [SBN Read] Namenode crashes if one of The JN is down > > > Key: HDFS-14655 > URL: https://issues.apache.org/jira/browse/HDFS-14655 > Project: Hadoop HDFS > Issue Type: Bug >Affects Versions: 3.3.0 >Reporter: Harshakiran Reddy >Assignee: Ayush Saxena >Priority: Critical > Attachments: HDFS-14655-01.patch, HDFS-14655-02.patch, > HDFS-14655-03.patch, HDFS-14655-04.patch, HDFS-14655-05.patch, > HDFS-14655-06.patch, HDFS-14655-07.patch, HDFS-14655.poc.patch > > > {noformat} > 2019-07-04 17:35:54,064 | INFO | Logger channel (from parallel executor) to > XXX/XXX | Retrying connect to server: XXX/XXX. Already tried > 9 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10, > sleepTime=1000 MILLISECONDS) | Client.java:975 > 2019-07-04 17:35:54,087 | FATAL | Edit log tailer | Unknown error encountered > while tailing edits. Shutting down standby NN. | EditLogTailer.java:474 > java.lang.OutOfMemoryError: unable to create new native thread > at java.lang.Thread.start0(Native Method) > at java.lang.Thread.start(Thread.java:717) > at > java.util.concurrent.ThreadPoolExecutor.addWorker(ThreadPoolExecutor.java:957) > at > java.util.concurrent.ThreadPoolExecutor.execute(ThreadPoolExecutor.java:1378) > at > com.google.common.util.concurrent.MoreExecutors$ListeningDecorator.execute(MoreExecutors.java:440) > at > com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56) > at > org.apache.hadoop.hdfs.qjournal.client.IPCLoggerChannel.getJournaledEdits(IPCLoggerChannel.java:565) > at > org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.getJournaledEdits(AsyncLoggerSet.java:272) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectRpcInputStreams(QuorumJournalManager.java:533) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectInputStreams(QuorumJournalManager.java:508) > at > org.apache.hadoop.hdfs.server.namenode.JournalSet.selectInputStreams(JournalSet.java:275) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1681) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1714) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer.doTailEdits(EditLogTailer.java:307) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.doWork(EditLogTailer.java:460) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.access$300(EditLogTailer.java:410) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread$1.run(EditLogTailer.java:427) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:360) > at > org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1709) > at > org.apache.hadoop.security.SecurityUtil.doAsLoginUserOrFatal(SecurityUtil.java:483) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.run(EditLogTailer.java:423) > 2019-07-04 17:35:54,112 | INFO | Edit log tailer | Exiting with status 1: > java.lang.OutOfMemoryError: unable to create new native thread | > ExitUtil.java:210 > {noformat} -- This message was sent by Atlassian Jira (v8.3.4#803005) - To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org For additional comm
[jira] [Comment Edited] (HDFS-14655) [SBN Read] Namenode crashes if one of The JN is down
[ https://issues.apache.org/jira/browse/HDFS-14655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16927900#comment-16927900 ] Ayush Saxena edited comment on HDFS-14655 at 9/11/19 7:00 PM: -- Yes, As I said above too, It interrupts But doesn’t kill, Then is Client.java it checks whether the thread is interrupted or not, if interrupted then it stops retrying, IIRC So thread can be reused.. was (Author: ayushtkn): Yes, As I said above too, It interrupts But doesn’t kill, Then is Client.java it checks whether the thread is interrupted or not, if interrupted then the thread is killed, IIRC > [SBN Read] Namenode crashes if one of The JN is down > > > Key: HDFS-14655 > URL: https://issues.apache.org/jira/browse/HDFS-14655 > Project: Hadoop HDFS > Issue Type: Bug >Affects Versions: 3.3.0 >Reporter: Harshakiran Reddy >Assignee: Ayush Saxena >Priority: Critical > Attachments: HDFS-14655-01.patch, HDFS-14655-02.patch, > HDFS-14655-03.patch, HDFS-14655-04.patch, HDFS-14655.poc.patch > > > {noformat} > 2019-07-04 17:35:54,064 | INFO | Logger channel (from parallel executor) to > XXX/XXX | Retrying connect to server: XXX/XXX. Already tried > 9 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10, > sleepTime=1000 MILLISECONDS) | Client.java:975 > 2019-07-04 17:35:54,087 | FATAL | Edit log tailer | Unknown error encountered > while tailing edits. Shutting down standby NN. | EditLogTailer.java:474 > java.lang.OutOfMemoryError: unable to create new native thread > at java.lang.Thread.start0(Native Method) > at java.lang.Thread.start(Thread.java:717) > at > java.util.concurrent.ThreadPoolExecutor.addWorker(ThreadPoolExecutor.java:957) > at > java.util.concurrent.ThreadPoolExecutor.execute(ThreadPoolExecutor.java:1378) > at > com.google.common.util.concurrent.MoreExecutors$ListeningDecorator.execute(MoreExecutors.java:440) > at > com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56) > at > org.apache.hadoop.hdfs.qjournal.client.IPCLoggerChannel.getJournaledEdits(IPCLoggerChannel.java:565) > at > org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.getJournaledEdits(AsyncLoggerSet.java:272) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectRpcInputStreams(QuorumJournalManager.java:533) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectInputStreams(QuorumJournalManager.java:508) > at > org.apache.hadoop.hdfs.server.namenode.JournalSet.selectInputStreams(JournalSet.java:275) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1681) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1714) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer.doTailEdits(EditLogTailer.java:307) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.doWork(EditLogTailer.java:460) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.access$300(EditLogTailer.java:410) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread$1.run(EditLogTailer.java:427) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:360) > at > org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1709) > at > org.apache.hadoop.security.SecurityUtil.doAsLoginUserOrFatal(SecurityUtil.java:483) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.run(EditLogTailer.java:423) > 2019-07-04 17:35:54,112 | INFO | Edit log tailer | Exiting with status 1: > java.lang.OutOfMemoryError: unable to create new native thread | > ExitUtil.java:210 > {noformat} -- This message was sent by Atlassian Jira (v8.3.2#803003) - To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (HDFS-14655) [SBN Read] Namenode crashes if one of The JN is down
[ https://issues.apache.org/jira/browse/HDFS-14655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16927900#comment-16927900 ] Ayush Saxena edited comment on HDFS-14655 at 9/11/19 6:59 PM: -- Yes, As I said above too, It interrupts But doesn’t kill, Then is Client.java it checks whether the thread is interrupted or not, if interrupted then the thread is killed, IIRC was (Author: ayushtkn): Yes, As I said above too, It interrupts But doesn’t kill,m > [SBN Read] Namenode crashes if one of The JN is down > > > Key: HDFS-14655 > URL: https://issues.apache.org/jira/browse/HDFS-14655 > Project: Hadoop HDFS > Issue Type: Bug >Affects Versions: 3.3.0 >Reporter: Harshakiran Reddy >Assignee: Ayush Saxena >Priority: Critical > Attachments: HDFS-14655-01.patch, HDFS-14655-02.patch, > HDFS-14655-03.patch, HDFS-14655-04.patch, HDFS-14655.poc.patch > > > {noformat} > 2019-07-04 17:35:54,064 | INFO | Logger channel (from parallel executor) to > XXX/XXX | Retrying connect to server: XXX/XXX. Already tried > 9 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10, > sleepTime=1000 MILLISECONDS) | Client.java:975 > 2019-07-04 17:35:54,087 | FATAL | Edit log tailer | Unknown error encountered > while tailing edits. Shutting down standby NN. | EditLogTailer.java:474 > java.lang.OutOfMemoryError: unable to create new native thread > at java.lang.Thread.start0(Native Method) > at java.lang.Thread.start(Thread.java:717) > at > java.util.concurrent.ThreadPoolExecutor.addWorker(ThreadPoolExecutor.java:957) > at > java.util.concurrent.ThreadPoolExecutor.execute(ThreadPoolExecutor.java:1378) > at > com.google.common.util.concurrent.MoreExecutors$ListeningDecorator.execute(MoreExecutors.java:440) > at > com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56) > at > org.apache.hadoop.hdfs.qjournal.client.IPCLoggerChannel.getJournaledEdits(IPCLoggerChannel.java:565) > at > org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.getJournaledEdits(AsyncLoggerSet.java:272) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectRpcInputStreams(QuorumJournalManager.java:533) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectInputStreams(QuorumJournalManager.java:508) > at > org.apache.hadoop.hdfs.server.namenode.JournalSet.selectInputStreams(JournalSet.java:275) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1681) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1714) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer.doTailEdits(EditLogTailer.java:307) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.doWork(EditLogTailer.java:460) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.access$300(EditLogTailer.java:410) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread$1.run(EditLogTailer.java:427) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:360) > at > org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1709) > at > org.apache.hadoop.security.SecurityUtil.doAsLoginUserOrFatal(SecurityUtil.java:483) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.run(EditLogTailer.java:423) > 2019-07-04 17:35:54,112 | INFO | Edit log tailer | Exiting with status 1: > java.lang.OutOfMemoryError: unable to create new native thread | > ExitUtil.java:210 > {noformat} -- This message was sent by Atlassian Jira (v8.3.2#803003) - To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (HDFS-14655) [SBN Read] Namenode crashes if one of The JN is down
[ https://issues.apache.org/jira/browse/HDFS-14655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16927854#comment-16927854 ] Ayush Saxena edited comment on HDFS-14655 at 9/11/19 5:44 PM: -- Thanx [~xkrogen], Well we can change it to 60 seconds. As it was previously, I guess that should be enough since we are cancelling the threads too? was (Author: ayushtkn): Thanx [~xkrogen], Well we can change it to 60 seconds. I guess that should be enough? > [SBN Read] Namenode crashes if one of The JN is down > > > Key: HDFS-14655 > URL: https://issues.apache.org/jira/browse/HDFS-14655 > Project: Hadoop HDFS > Issue Type: Bug >Affects Versions: 3.3.0 >Reporter: Harshakiran Reddy >Assignee: Ayush Saxena >Priority: Critical > Attachments: HDFS-14655-01.patch, HDFS-14655-02.patch, > HDFS-14655-03.patch, HDFS-14655-04.patch, HDFS-14655.poc.patch > > > {noformat} > 2019-07-04 17:35:54,064 | INFO | Logger channel (from parallel executor) to > XXX/XXX | Retrying connect to server: XXX/XXX. Already tried > 9 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10, > sleepTime=1000 MILLISECONDS) | Client.java:975 > 2019-07-04 17:35:54,087 | FATAL | Edit log tailer | Unknown error encountered > while tailing edits. Shutting down standby NN. | EditLogTailer.java:474 > java.lang.OutOfMemoryError: unable to create new native thread > at java.lang.Thread.start0(Native Method) > at java.lang.Thread.start(Thread.java:717) > at > java.util.concurrent.ThreadPoolExecutor.addWorker(ThreadPoolExecutor.java:957) > at > java.util.concurrent.ThreadPoolExecutor.execute(ThreadPoolExecutor.java:1378) > at > com.google.common.util.concurrent.MoreExecutors$ListeningDecorator.execute(MoreExecutors.java:440) > at > com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56) > at > org.apache.hadoop.hdfs.qjournal.client.IPCLoggerChannel.getJournaledEdits(IPCLoggerChannel.java:565) > at > org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.getJournaledEdits(AsyncLoggerSet.java:272) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectRpcInputStreams(QuorumJournalManager.java:533) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectInputStreams(QuorumJournalManager.java:508) > at > org.apache.hadoop.hdfs.server.namenode.JournalSet.selectInputStreams(JournalSet.java:275) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1681) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1714) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer.doTailEdits(EditLogTailer.java:307) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.doWork(EditLogTailer.java:460) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.access$300(EditLogTailer.java:410) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread$1.run(EditLogTailer.java:427) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:360) > at > org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1709) > at > org.apache.hadoop.security.SecurityUtil.doAsLoginUserOrFatal(SecurityUtil.java:483) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.run(EditLogTailer.java:423) > 2019-07-04 17:35:54,112 | INFO | Edit log tailer | Exiting with status 1: > java.lang.OutOfMemoryError: unable to create new native thread | > ExitUtil.java:210 > {noformat} -- This message was sent by Atlassian Jira (v8.3.2#803003) - To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
[jira] [Comment Edited] (HDFS-14655) [SBN Read] Namenode crashes if one of The JN is down
[ https://issues.apache.org/jira/browse/HDFS-14655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16925848#comment-16925848 ] Erik Krogen edited comment on HDFS-14655 at 9/9/19 3:43 PM: Looks better to me, thanks [~ayushtkn]! I do think we should rename the config and update the description to represent that this config is a _maximum_ thread count; the way it reads now, I would assume that there are always this many threads being used. One thing I noticed, you used a keepalive time of 0: {code} return new HadoopThreadPoolExecutor(1, numThreads, 0L, TimeUnit.MILLISECONDS, new LinkedBlockingQueue(), {code} I feel a longer time would probably be better; if more than 1 thread is needed, it will probably be needed again soon (might represent a slow JN?), so it seems some keepalive would be helpful to avoid the thread creation overhead. Also you can use [diamond-typing|https://docs.oracle.com/javase/tutorial/java/generics/types.html#diamond] here for the {{LinkedBlockingQueue}} instantiation. [~shv], does the current approach address your previous concerns? was (Author: xkrogen): Looks better to me, thanks [~ayushtkn]! I do think we should rename the config and update the description to represent that this config is a _maximum_ thread count; the way it reads now, I would assume that there are always this many threads being used. One thing I noticed, you used a keepalive time of 0: {code} return new HadoopThreadPoolExecutor(1, numThreads, 0L, TimeUnit.MILLISECONDS, new LinkedBlockingQueue(), {code} I feel a longer time would probably be better; if more than 1 thread is needed, it will probably be needed again soon (might represent a slow JN?), so it seems some keepalive would be helpful to avoid the thread creation overhead. Also you can use [diamond-typing|https://docs.oracle.com/javase/tutorial/java/generics/types.html#diamond] here for the {{LinkedBlockingQueue}} instantiation. > [SBN Read] Namenode crashes if one of The JN is down > > > Key: HDFS-14655 > URL: https://issues.apache.org/jira/browse/HDFS-14655 > Project: Hadoop HDFS > Issue Type: Bug >Affects Versions: 3.3.0 >Reporter: Harshakiran Reddy >Assignee: Ayush Saxena >Priority: Critical > Attachments: HDFS-14655-01.patch, HDFS-14655-02.patch, > HDFS-14655-03.patch, HDFS-14655.poc.patch > > > {noformat} > 2019-07-04 17:35:54,064 | INFO | Logger channel (from parallel executor) to > XXX/XXX | Retrying connect to server: XXX/XXX. Already tried > 9 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10, > sleepTime=1000 MILLISECONDS) | Client.java:975 > 2019-07-04 17:35:54,087 | FATAL | Edit log tailer | Unknown error encountered > while tailing edits. Shutting down standby NN. | EditLogTailer.java:474 > java.lang.OutOfMemoryError: unable to create new native thread > at java.lang.Thread.start0(Native Method) > at java.lang.Thread.start(Thread.java:717) > at > java.util.concurrent.ThreadPoolExecutor.addWorker(ThreadPoolExecutor.java:957) > at > java.util.concurrent.ThreadPoolExecutor.execute(ThreadPoolExecutor.java:1378) > at > com.google.common.util.concurrent.MoreExecutors$ListeningDecorator.execute(MoreExecutors.java:440) > at > com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56) > at > org.apache.hadoop.hdfs.qjournal.client.IPCLoggerChannel.getJournaledEdits(IPCLoggerChannel.java:565) > at > org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.getJournaledEdits(AsyncLoggerSet.java:272) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectRpcInputStreams(QuorumJournalManager.java:533) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectInputStreams(QuorumJournalManager.java:508) > at > org.apache.hadoop.hdfs.server.namenode.JournalSet.selectInputStreams(JournalSet.java:275) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1681) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1714) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer.doTailEdits(EditLogTailer.java:307) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.doWork(EditLogTailer.java:460) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.access$300(EditLogTailer.java:410) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread$1.run(EditLogTailer.java:427) > at java.security.AccessController.doPrivileged(Native Method) > at javax.secu
[jira] [Comment Edited] (HDFS-14655) [SBN Read] Namenode crashes if one of The JN is down
[ https://issues.apache.org/jira/browse/HDFS-14655?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16923651#comment-16923651 ] Erik Krogen edited comment on HDFS-14655 at 9/5/19 5:49 PM: -[~ayushtkn] isn't 1 thread insufficient to perform the calls in parallel? Shouldn't the max thread pool size be equal to the number of JNs by default?- Nevermind, I reread my previous comment and reminded myself that the {{parallelExecutor}} is scoped to a single JN. Given that the intent of this executor (IIUC) is to allow multiple requests to execute in parallel, it seems we should allow _some_ parallelism at least, but maybe scope it to some reasonable limit of threads (10?). [~shv], regarding your previous comment, if I understand correctly this pool of threads is already re-used due to the use of the {{ThreadPoolExecutor}}. Let me know if I'm missing anything. was (Author: xkrogen): [~ayushtkn] isn't 1 thread insufficient to perform the calls in parallel? Shouldn't the max thread pool size be equal to the number of JNs by default? [~shv], regarding your previous comment, if I understand correctly this pool of threads is already re-used due to the use of the {{ThreadPoolExecutor}}. Let me know if I'm missing anything. > [SBN Read] Namenode crashes if one of The JN is down > > > Key: HDFS-14655 > URL: https://issues.apache.org/jira/browse/HDFS-14655 > Project: Hadoop HDFS > Issue Type: Bug >Reporter: Harshakiran Reddy >Assignee: Ayush Saxena >Priority: Major > Attachments: HDFS-14655-01.patch, HDFS-14655.poc.patch > > > {noformat} > 2019-07-04 17:35:54,064 | INFO | Logger channel (from parallel executor) to > XXX/XXX | Retrying connect to server: XXX/XXX. Already tried > 9 time(s); retry policy is RetryUpToMaximumCountWithFixedSleep(maxRetries=10, > sleepTime=1000 MILLISECONDS) | Client.java:975 > 2019-07-04 17:35:54,087 | FATAL | Edit log tailer | Unknown error encountered > while tailing edits. Shutting down standby NN. | EditLogTailer.java:474 > java.lang.OutOfMemoryError: unable to create new native thread > at java.lang.Thread.start0(Native Method) > at java.lang.Thread.start(Thread.java:717) > at > java.util.concurrent.ThreadPoolExecutor.addWorker(ThreadPoolExecutor.java:957) > at > java.util.concurrent.ThreadPoolExecutor.execute(ThreadPoolExecutor.java:1378) > at > com.google.common.util.concurrent.MoreExecutors$ListeningDecorator.execute(MoreExecutors.java:440) > at > com.google.common.util.concurrent.AbstractListeningExecutorService.submit(AbstractListeningExecutorService.java:56) > at > org.apache.hadoop.hdfs.qjournal.client.IPCLoggerChannel.getJournaledEdits(IPCLoggerChannel.java:565) > at > org.apache.hadoop.hdfs.qjournal.client.AsyncLoggerSet.getJournaledEdits(AsyncLoggerSet.java:272) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectRpcInputStreams(QuorumJournalManager.java:533) > at > org.apache.hadoop.hdfs.qjournal.client.QuorumJournalManager.selectInputStreams(QuorumJournalManager.java:508) > at > org.apache.hadoop.hdfs.server.namenode.JournalSet.selectInputStreams(JournalSet.java:275) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1681) > at > org.apache.hadoop.hdfs.server.namenode.FSEditLog.selectInputStreams(FSEditLog.java:1714) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer.doTailEdits(EditLogTailer.java:307) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.doWork(EditLogTailer.java:460) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.access$300(EditLogTailer.java:410) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread$1.run(EditLogTailer.java:427) > at java.security.AccessController.doPrivileged(Native Method) > at javax.security.auth.Subject.doAs(Subject.java:360) > at > org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1709) > at > org.apache.hadoop.security.SecurityUtil.doAsLoginUserOrFatal(SecurityUtil.java:483) > at > org.apache.hadoop.hdfs.server.namenode.ha.EditLogTailer$EditLogTailerThread.run(EditLogTailer.java:423) > 2019-07-04 17:35:54,112 | INFO | Edit log tailer | Exiting with status 1: > java.lang.OutOfMemoryError: unable to create new native thread | > ExitUtil.java:210 > {noformat} -- This message was sent by Atlassian Jira (v8.3.2#803003) - To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: hdfs-issues-h...@hadoo