hadoop-common-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sameer Alsakran" <sam...@imeem.com>
Subject Problem with edits.new rolling over
Date Sun, 09 Sep 2007 00:05:20 GMT
Hello -

 

I have been running into problems with one or two reduce tasks stalling
out. This was occuring regularly in 0.13, and remains an issue with
0.14.1

 

In both cases the fsimage file in the namenode's current image is empty,
with the it, as well as the edits file not having been touched in either
a day (with 0.14.1) or 3 weeks (in the case of the old 0.13
installation). The edits.new file is continuing to grow. Shutting down
the namenode when this occurred resulted in a state where the namenode
would not restart.

 

In the namenode log,  there are multiple exceptions of the form listed
below.

********************

2007-09-08 16:58:04,182 INFO org.apache.hadoop.fs.FSNamesystem: Roll
Edit Log fr

om 10.33.0.56

2007-09-08 16:58:04,182 INFO org.apache.hadoop.ipc.Server: IPC Server
handler 34

 on 50001, call rollEditLog() from 10.33.0.56:41291: error:

java.io.IOException: Attempt to roll edit log but edits.new exists

java.io.IOException: Attempt to roll edit log but edits.new exists

        at
org.apache.hadoop.dfs.FSEditLog.rollEditLog(FSEditLog.java:582)

        at
org.apache.hadoop.dfs.FSNamesystem.rollEditLog(FSNamesystem.java:3510

)

        at org.apache.hadoop.dfs.NameNode.rollEditLog(NameNode.java:553)

        at sun.reflect.GeneratedMethodAccessor11.invoke(Unknown Source)

        at
sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAcces

sorImpl.java:25)

        at java.lang.reflect.Method.invoke(Method.java:597)

        at org.apache.hadoop.ipc.RPC$Server.call(RPC.java:340)

        at org.apache.hadoop.ipc.Server$Handler.run(Server.java:566)

 

**************

 

Has anyone run into this? 

 

Thanks

-Sameer


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message