Mailing-List: contact user-help@hbase.apache.org; run by ezmlm
Precedence: bulk
Reply-To: user@hbase.apache.org
Received-SPF: pass (nike.apache.org: domain of brian.jeltema@digitalenvoy.net
 designates 68.64.43.136 as permitted sender)
From: Brian Jeltema <brian.jeltema@digitalenvoy.net>
Content-Type: multipart/alternative;
 boundary="Apple-Mail=_AB818AE6-C6A3-405D-9577-27AC9313CA0C"
Subject: ExportSnapshot webhdfs problems
Message-Id: <DF4011BD-F6CB-4710-8560-89F4191111EA@digitalenvoy.net>
Date: Tue, 30 Sep 2014 07:33:17 -0400
To: user@hbase.apache.org
Mime-Version: 1.0 (Mac OS X Mail 7.2 \(1874\))

--Apple-Mail=_AB818AE6-C6A3-405D-9577-27AC9313CA0C
Content-Transfer-Encoding: quoted-printable
Content-Type: text/plain;
	charset=windows-1252

I=92m trying to use ExportSnapshot to copy a snapshot from a Hadoop 1 to =
a Hadoop 2 cluster using the webhdfs protocol.
I=92ve done this successfully before, though there are always mapper =
failures and retries in the job log. However, I=92m not
having success with a rather large table due to an excessive number of =
failures. The exceptions in the job log are always:

14/09/29 20:28:11 INFO mapred.JobClient: Task Id : =
attempt_201409241055_0024_m_000005_1, Status : FAILED
org.apache.hadoop.ipc.RemoteException
	at =
org.apache.hadoop.hdfs.web.JsonUtil.toRemoteException(JsonUtil.java:114)
	at =
org.apache.hadoop.hdfs.web.WebHdfsFileSystem.validateResponse(WebHdfsFileS=
ystem.java:290)
	at =
org.apache.hadoop.hdfs.web.WebHdfsFileSystem.access$500(WebHdfsFileSystem.=
java:98)
	at =
org.apache.hadoop.hdfs.web.WebHdfsFileSystem$2.close(WebHdfsFileSystem.jav=
a:653)
	at =
org.apache.hadoop.hbase.snapshot.ExportSnapshot$ExportMapper.copyFile(Expo=
rtSnapshot.java:204)
	at =
org.apache.hadoop.hbase.snapshot.ExportSnapshot$ExportMapper.map(ExportSna=
pshot.java:146)
        =85
        =85
        =85


So I presume the real exception is taking place on the target system. =
However, examining the namenode logs and a handful of
the datanode logs has not revealed any exceptions that correlate with =
those in the job log. Is there some other log I should be looking at?
I reduced the number of mappers to 6 and the target cluster has 10 =
datanodes, so it=92s hard to believe its a capacity problem.

Thanks
Brian=

--Apple-Mail=_AB818AE6-C6A3-405D-9577-27AC9313CA0C--