Home | About | Sematext search-lucene.com search-hadoop.com
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB
 Search Hadoop and all its subprojects:

Switch to Threaded View
Flume >> mail # dev >> Measuring Flume perf


Copy link to this message
-
Re: Measuring Flume perf
Are all of your HDFS sinks writing to the same directory with the same file prefix? Looks like it is - and they are stepping on each other’s toes.  
Thanks,
Hari
On Sunday, December 15, 2013 at 7:30 PM, Roshan Naik wrote:

> Increasing the batch size to 10k didnt make any diff: 24.6k event/sec
> Subsequently increased the number of sources from 1 to 8 which improved it
> a bit .. 33k e/s
>  
> Yes I will try to take a deeper look using a profiler.
>  
> Here is another issue that comes up occasionally with HDFS sink ... any
> thoughts ?
>  
> 15 Dec 2013 11:13:26,689 ERROR
> [SinkRunner-PollingRunner-DefaultSinkProcessor]
> (org.apache.flume.sink.hdfs.AbstractHDFSWriter.isUnderReplicated:82) -
> Unexpected error while checking replication factor
>  
> java.lang.reflect.InvocationTargetException
>  
> at sun.reflect.GeneratedMethodAccessor1.invoke(Unknown Source)
>  
> at
> sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25)
>  
> at java.lang.reflect.Method.invoke(Method.java:597)
>  
> at
> org.apache.flume.sink.hdfs.AbstractHDFSWriter.getNumCurrentReplicas(AbstractHDFSWriter.java:147)
>  
> at
> org.apache.flume.sink.hdfs.AbstractHDFSWriter.isUnderReplicated(AbstractHDFSWriter.java:68)
>  
> at
> org.apache.flume.sink.hdfs.BucketWriter.shouldRotate(BucketWriter.java:452)
>  
> at
> org.apache.flume.sink.hdfs.BucketWriter.append(BucketWriter.java:387)
>  
> at
> org.apache.flume.sink.hdfs.HDFSEventSink.process(HDFSEventSink.java:392)
>  
> at
> org.apache.flume.sink.DefaultSinkProcessor.process(DefaultSinkProcessor.java:68)
>  
> at
> org.apache.flume.SinkRunner$PollingRunner.run(SinkRunner.java:147)
>  
> at java.lang.Thread.run(Thread.java:662)
>  
> Caused by:
> org.apache.hadoop.ipc.RemoteException(org.apache.hadoop.hdfs.server.namenode.LeaseExpiredException):
> Lease mismatch on /flume/bidder/_flume_tx_agent01_sink01_web06-
> east.manage.com_2013121422.1387093314310.snappy.tmp owned by
> DFSClient_NONMAPREDUCE_-1742177356_32 but is accessed by
> DFSClient_NONMAPREDUCE_643393114_25
>  
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.checkLease(FSNamesystem.java:2770)
>  
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.analyzeFileState(FSNamesystem.java:2567)
>  
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getAdditionalBlock(FSNamesystem.java:2480)
>  
> at
> org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer.addBlock(NameNodeRpcServer.java:555)
>  
> at
> org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolServerSideTranslatorPB.addBlock(ClientNamenodeProtocolServerSideTranslatorPB.java:387)
>  
> --  
> CONFIDENTIALITY NOTICE
> NOTICE: This message is intended for the use of the individual or entity to  
> which it is addressed and may contain information that is confidential,  
> privileged and exempt from disclosure under applicable law. If the reader  
> of this message is not the intended recipient, you are hereby notified that  
> any printing, copying, dissemination, distribution, disclosure or  
> forwarding of this communication is strictly prohibited. If you have  
> received this communication in error, please contact the sender immediately  
> and delete it from your system. Thank You.
>  
>  
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB