[
https://issues.apache.org/jira/browse/HIVE-13278?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15744366#comment-15744366
]
Rui Li commented on HIVE-13278:
-------------------------------
Hi [~xuefuz], the conclusion is we somehow try to read reduce.xml for map-only
job, and yes it happens to MR as well. The call path is
{{HiveOutputFormatImpl.checkOutputSpecs -> Utilities.getMapRedWork}}. The
reason why HiveOutputFormatImpl needs to get the MapRedWork is it needs to do
some check on all the FS operators. Since FS only exists at the end of a job,
my suggestion is we firstly try to get MapWork. If the MapWork has an FS in it,
it means this is a map-only job so we don't have to look for ReduceWork. But
[~stakiar] found that some map-only job may not have FS in the MapWork, e.g.
{{ANALYZE TABLE}}. To have a complete fix, we'll need some flag in the JobConf
indicating if this is map-only. Or we can use my solution, which solves the
issue for most cases.
Some special handling for HoS may be needed. For HoS, each map.xml and
reduce.xml resides in a different path. We can use {{mapred.task.is.map}} to
determine whether the JobConf is for MapWork or ReduceWork. And then call
getMapWork or getReduceWork respectively.
> Many redundant 'File not found' messages appeared in container log during
> query execution with Hive on Spark
> ------------------------------------------------------------------------------------------------------------
>
> Key: HIVE-13278
> URL: https://issues.apache.org/jira/browse/HIVE-13278
> Project: Hive
> Issue Type: Bug
> Environment: Hive on Spark engine
> Found based on :
> Apache Hive 2.0.0
> Apache Spark 1.6.0
> Reporter: Xin Hao
> Assignee: Sahil Takiar
> Priority: Minor
>
> Many redundant 'File not found' messages appeared in container log during
> query execution with Hive on Spark.
> Certainly, it doesn't prevent the query from running successfully. So mark it
> as Minor currently.
> Error message example:
> {noformat}
> 16/03/14 01:45:06 INFO exec.Utilities: File not found: File does not exist:
> /tmp/hive/hadoop/2d378538-f5d3-493c-9276-c62dd6634fb4/hive_2016-03-14_01-44-16_835_623058724409492515-6/-mr-10010/0a6d0cae-1eb3-448c-883b-590b3b198a73/reduce.xml
> at
> org.apache.hadoop.hdfs.server.namenode.INodeFile.valueOf(INodeFile.java:66)
> at
> org.apache.hadoop.hdfs.server.namenode.INodeFile.valueOf(INodeFile.java:56)
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocationsUpdateTimes(FSNamesystem.java:1932)
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocationsInt(FSNamesystem.java:1873)
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocations(FSNamesystem.java:1853)
> at
> org.apache.hadoop.hdfs.server.namenode.FSNamesystem.getBlockLocations(FSNamesystem.java:1825)
> at
> org.apache.hadoop.hdfs.server.namenode.NameNodeRpcServer.getBlockLocations(NameNodeRpcServer.java:565)
> at
> org.apache.hadoop.hdfs.server.namenode.AuthorizationProviderProxyClientProtocol.getBlockLocations(AuthorizationProviderProxyClientProtocol.java:87)
> at
> org.apache.hadoop.hdfs.protocolPB.ClientNamenodeProtocolServerSideTranslatorPB.getBlockLocations(ClientNamenodeProtocolServerSideTranslatorPB.java:363)
> at
> org.apache.hadoop.hdfs.protocol.proto.ClientNamenodeProtocolProtos$ClientNamenodeProtocol$2.callBlockingMethod(ClientNamenodeProtocolProtos.java)
> at
> org.apache.hadoop.ipc.ProtobufRpcEngine$Server$ProtoBufRpcInvoker.call(ProtobufRpcEngine.java:617)
> at org.apache.hadoop.ipc.RPC$Server.call(RPC.java:1060)
> at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2086)
> at org.apache.hadoop.ipc.Server$Handler$1.run(Server.java:2082)
> at java.security.AccessController.doPrivileged(Native Method)
> at javax.security.auth.Subject.doAs(Subject.java:415)
> at
> org.apache.hadoop.security.UserGroupInformation.doAs(UserGroupInformation.java:1671)
> at org.apache.hadoop.ipc.Server$Handler.run(Server.java:2080)
> {noformat}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
