CLONE - Issue with HDFSEventSink for append support for HDFSCompressedDataStream
--------------------------------------------------------------------------------
Key: FLUME-1112
URL: https://issues.apache.org/jira/browse/FLUME-1112
Project: Flume
Issue Type: Bug
Reporter: Inder SIngh
Assignee: Inder SIngh
Fix For: v1.2.0
Enabled append feature in HDFS and flume conf
Hit the following exceptions -
2012-03-29 13:27:03,284 (SinkRunner-PollingRunner-DefaultSinkProcessor) [WARN -
org.apache.flume.sink.hdfs.HDFSEventSink.process(HDFSEventSink.java:370)] HDFS
IO error
java.io.FileNotFoundException: java.io.FileNotFoundException: failed to append
to non-existent file /flume.0.tmp on client 127.0.0.1
org.apache.hadoop.hdfs.DFSClient.append(DFSClient.java:586)
at
org.apache.hadoop.hdfs.DistributedFileSystem.append(DistributedFileSystem.java:209)
at org.apache.hadoop.fs.FileSystem.append(FileSystem.java:630)
at
org.apache.flume.sink.hdfs.HDFSDataStream.open(HDFSDataStream.java:42)
at org.apache.flume.sink.hdfs.BucketWriter.open(BucketWriter.java:82)
at org.apache.flume.sink.hdfs.BucketWriter.open(BucketWriter.java:102)
at
org.apache.flume.sink.hdfs.HDFSEventSink.process(HDFSEventSink.java:342)
at
org.apache.flume.sink.DefaultSinkProcessor.process(DefaultSinkProcessor.java:65)
at org.apache.flume.SinkRunner$PollingRunner.run(SinkRunner.java:148)
at java.lang.Thread.run(Thread.java:680)
Caused by: org.apache.hadoop.ipc.RemoteException:
java.io.FileNotFoundException: failed to append to non-existent file
/flume.0.tmp on client 127.0.0.1
at
org.apache.hadoop.hdfs.server.namenode.FSNamesystem.startFileInternal(FSNamesystem.java:1187)
at
org.apache.hadoop.hdfs.server.namenode.FSNamesystem.appendFile(FSNamesystem.java:1357)
at
org.apache.hadoop.hdfs.server.namenode.NameNode.append(NameNode.java:600)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
Looking at the code and debugging further looks like the case of first write
when append="true" isn't handled.
Here is the code in picture -> HDFSDataStream.open
if (conf.getBoolean("hdfs.append.support", false) == true) {
outStream = hdfs.append(dstPath);
} else {
outStream = hdfs.create(dstPath);
}
it should be something like this -
if (conf.getBoolean("hdfs.append.support", false) == true &&
hdfs.isFile(dstPath)) {
outStream = hdfs.append(dstPath);
} else {
outStream = hdfs.create(dstPath);
}
Refer -
http://hadoop.apache.org/common/docs/current/api/org/apache/hadoop/fs/FileSystem.html
- append works on an existent file.
Pardon my ignorance and correct me if am missing something here
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
