question hdfs journalnode fail, can not start in Support Questions

hdfs journalnode fail, can not start

BaiTing — Thu, 14 Apr 2016 10:14:36 GMT

repeat this log: 4/14, 16:53:28.739 WARN org.apache.hadoop.hdfs.server.namenode.FSImage Caught exception after scanning through 0 ops from /data01/dfs/jn/pt-nameservice/current/edits_inprogress_0000000000182289435 while determining its valid length. Position was 987136 java.io.IOException: Can't scan a pre-transactional edit log. at org.apache.hadoop.hdfs.server.namenode.FSEditLogOp$LegacyReader.scanOp(FSEditLogOp.java:4592) at org.apache.hadoop.hdfs.server.namenode.EditLogFileInputStream.scanNextOp(EditLogFileInputStream.java:245) at org.apache.hadoop.hdfs.server.namenode.EditLogFileInputStream.scanEditLog(EditLogFileInputStream.java:355) at org.apache.hadoop.hdfs.server.namenode.FileJournalManager$EditLogFile.scanLog(FileJournalManager.java:551) at org.apache.hadoop.hdfs.qjournal.server.Journal.scanStorageForLatestEdits(Journal.java:193) at org.apache.hadoop.hdfs.qjournal.server.Journal.(Journal.java:153) at org.apache.hadoop.hdfs.qjournal.server.JournalNode.getOrCreateJournal(JournalNode.java:93) at org.apache.hadoop.hdfs.qjournal.server.JournalNode.getOrCreateJournal(JournalNode.java:102) at org.apache.hadoop.hdfs.qjournal.server.JournalNodeRpcServer.newEpoch(JournalNodeRpcServer.java:136) at org.apache.hadoop.hdfs.qjournal.protocolPB.QJournalProtocolServerSideTranslatorPB.newEpoch(QJournalProtocolServerSideTranslatorPB.java:133) at org.apache.hadoop.hdfs.qjournal.protocol.QJournalProtocolProtos$QJournalProtocolService$2.callBlockingMethod(QJournalProtocolProtos.java:25417) at org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ProtoBufRpcInvoker.call(ProtobufRpcEngine.java:617) at org.apache.hadoop.ipc.RPC$Server.call(RPC.java:1060) at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2086) at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2082) at java.security.AccessController.doPrivileged(Native Method) at javax.security.auth.Subject.doAs(Subject.java:422) at org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1671) at org.apache.hadoop.ipc.Server$Handler.run(Server.java:2080)

Re: hdfs journalnode fail, can not start

ben.hemphill — Fri, 15 Apr 2016 04:24:11 GMT

This check seems to be added from https://issues.apache.org/jira/browse/HDFS-8965

/data01/dfs/jn/pt-nameservice/current/edits_inprogress_0000000000182289435 is either corrupted, or perhaps It is a log from before an upgrade?

Either way, To get it back I would suggest backing up the /data01/dfs/jn/pt-nameservice/current/ directory somewhere else and copy the journalnode data from one of your other journalnodes to that location.

Re: hdfs journalnode fail, can not start

dpetrovan — Mon, 18 Apr 2016 15:50:31 GMT

Hi Ben,

 I have configure a cluster HDFS HA using the Quorum Journal Manager with 5 Journal Node, on one of them I receive the same error as BaiTing, If I copy the journal files from a working Journal Node should resolve the issue?

Thanks in advance!

Re: hdfs journalnode fail, can not start

BaiTing — Mon, 18 Apr 2016 16:08:43 GMT

Hi,ben

it does work for me,thanks!

Re: hdfs journalnode fail, can not start

ben.hemphill — Mon, 18 Apr 2016 16:12:45 GMT

Yes, correct, moving out the old, corrupted files and copying in the files from a working Journal Node should allow you to start the journalnode.

Re: hdfs journalnode fail, can not start

ben.hemphill — Mon, 18 Apr 2016 16:13:21 GMT

Glad to hear it BaiTing!

Re: hdfs journalnode fail, can not start

dpetrovan — Mon, 18 Apr 2016 17:43:47 GMT

It works thanks!

Re: hdfs journalnode fail, can not start

ussama — Mon, 08 Aug 2016 10:48:39 GMT

I am also facing the same issue but have a question: Do we need to stop HDFS ervice or put it to maintenance/read-only mode before copying the data? Otherwise, there may be data being written to the health journalnode (from which data will be copied).

Re: hdfs journalnode fail, can not start

ben.hemphill — Thu, 11 Aug 2016 19:00:06 GMT

you shouldn't have to, as when the journalnode starts up, it should recognize that it is behind and then sync up with the other journalnodes. (similar things happen when you stop a journalnode and then restart it later)

Re: hdfs journalnode fail, can not start

rafaelpecin — Thu, 23 Feb 2017 20:23:22 GMT

You don't have to stop any instance nor service on the cluster.
The error messages stop after 30~60 seconds.

Re: hdfs journalnode fail, can not start

banga — Wed, 25 Nov 2020 06:08:51 GMT

it works thanks