ConfX created HDFS-17099:
----------------------------
Summary: Null Pointer Exception when stop namesystem in HDFS
Key: HDFS-17099
URL: https://issues.apache.org/jira/browse/HDFS-17099
Project: Hadoop HDFS
Issue Type: Bug
Reporter: ConfX
Attachments: reproduce.sh
h2. What happend:
Got NullPointerException when stop namesystem in HDFS.
h2. Buggy code:
{code:java}
void stopActiveServices() {
...
if (dir != null && getFSImage() != null) {
if (getFSImage().editLog != null) { // <--- Check whether editLog is
null
getFSImage().editLog.close();
}
// Update the fsimage with the last txid that we wrote
// so that the tailer starts from the right spot.
getFSImage().updateLastAppliedTxIdFromWritten(); // <--- BUG: Even if
editLog is null, this line will still be executed and cause nullpointer
exception
}
...
} public void updateLastAppliedTxIdFromWritten() {
this.lastAppliedTxId = editLog.getLastWrittenTxId(); // <---- This will
cause nullpointer exception if editLog is null
} {code}
h2. StackTrace:
{code:java}
java.lang.NullPointerException
at
org.apache.hadoop.hdfs.server.namenode.FSImage.updateLastAppliedTxIdFromWritten(FSImage.java:1553)
at
org.apache.hadoop.hdfs.server.namenode.FSNamesystem.stopActiveServices(FSNamesystem.java:1463)
at
org.apache.hadoop.hdfs.server.namenode.FSNamesystem.close(FSNamesystem.java:1815)
at
org.apache.hadoop.hdfs.server.namenode.FSNamesystem.<init>(FSNamesystem.java:1017)
at
org.apache.hadoop.hdfs.server.namenode.SecondaryNameNode.initialize(SecondaryNameNode.java:248)
at
org.apache.hadoop.hdfs.server.namenode.SecondaryNameNode.<init>(SecondaryNameNode.java:194)
at
org.apache.hadoop.hdfs.server.namenode.SecondaryNameNode.<init>(SecondaryNameNode.java:181)
{code}
h2. How to reproduce:
(1) Set {{dfs.namenode.top.windows.minutes}} to {{{}37914516,32,0{}}}; or set
{{dfs.namenode.top.window.num.buckets}} to {{{}244111242{}}}.
(2) Run test:
{{org.apache.hadoop.hdfs.server.namenode.TestNameNodeHttpServerXFrame#testSecondaryNameNodeXFrame}}
h2. What's more:
I'm still investigating how the parameter {{dfs.namenode.top.windows.minutes}}
triggered the buggy code.
For an easy reproduction, run the reproduce.sh in the attachment.
We are happy to provide a patch if this issue is confirmed.
{{}}
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-dev-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-dev-h...@hadoop.apache.org
