Thanks Ryan. I actually already had this in the hbase-site.xml, but somehow missed putting it in the core-site.xml. The cluster seems to be much happier. Dave
-----Original Message----- From: Ryan Rawson [mailto:ryano...@gmail.com] Sent: Monday, August 09, 2010 12:01 PM To: user@hbase.apache.org Subject: Re: hbase error caused by DFS timeout Try this config: <property> <name>dfs.datanode.socket.write.timeout</name> <value>0</value> </property> in both hbase-site.xml and core-site.xml in the hadoop configs. -ryan On Mon, Aug 9, 2010 at 10:02 AM, Buttler, David <buttl...@llnl.gov> wrote: > Hi all, > I seem to get this error far too frequently: > > 2010-08-09 09:54:03,685 INFO org.apache.hadoop.hbase.regionserver.Store: > Started compaction of 1 file(s) in annotations of > doc,293817e024ed1d54a11e9e7c9b836dd837badbbc,1281372823189, > hasReferences=true, into /hbase/doc/compaction.dir/237345967, seqid=1182913218 > 2010-08-09 09:54:03,784 WARN org.apache.hadoop.hdfs.DFSClient: > DFSOutputStream ResponseProcessor exception for block > blk_-4556852958383799371_431518java.net.SocketTimeoutException: 6000 millis > timeout while waiting for channel to be ready for read. ch : > java.nio.channels.SocketChannel[connected local=/10.220.5.35:49924 > remote=/10.220.5.14:50010] > at > org.apache.hadoop.net.SocketIOWithTimeout.doIO(SocketIOWithTimeout.java:164) > at > org.apache.hadoop.net.SocketInputStream.read(SocketInputStream.java:155) > at > org.apache.hadoop.net.SocketInputStream.read(SocketInputStream.java:128) > at java.io.DataInputStream.readFully(DataInputStream.java:178) > at java.io.DataInputStream.readLong(DataInputStream.java:399) > at > org.apache.hadoop.hdfs.protocol.DataTransferProtocol$PipelineAck.readFields(DataTransferProtocol.java:119) > at > org.apache.hadoop.hdfs.DFSClient$DFSOutputStream$ResponseProcessor.run(DFSClient.java:2424) > > This basically is taking one of clusters down consistently. Is there an > obvious thing I can do about this? > I have seen this across three different clusters with radically different > hardware, leading me to believe that I have misconfigured something in either > hbase or hdfs > > Any ideas of where to look? > > Thanks, > Dave >
