[
https://issues.apache.org/jira/browse/HBASE-8386?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15384613#comment-15384613
]
Sean Busbey commented on HBASE-8386:
------------------------------------
{code}
Tests in error:
org.apache.hadoop.hbase.mapreduce.TestLoadIncrementalHFiles.testRegionCrossingRowColBloom(org.apache.hadoop.hbase.mapreduce.TestLoadIncrementalHFiles)
Run 1:
TestLoadIncrementalHFiles.testRegionCrossingRowColBloom:152->runTest:249->runTest:259->runTest:269->runTest:295
» ScannerTimeout
Run 2:
TestLoadIncrementalHFiles.testRegionCrossingRowColBloom:152->runTest:249->runTest:259->runTest:269->runTest:295
» TableNotFound
Run 3:
TestLoadIncrementalHFiles.testRegionCrossingRowColBloom:152->runTest:249->runTest:259->runTest:269->runTest:295
» TableNotFound
Flaked tests:
org.apache.hadoop.hbase.TestInfoServers.testInfoServersStatusPages(org.apache.hadoop.hbase.TestInfoServers)
Run 1:
TestInfoServers.testInfoServersStatusPages:92->assertContainsContent:113->getUrlContent:129
» SocketTimeout
Run 2: PASS
org.apache.hadoop.hbase.replication.TestMasterReplication.testCyclicReplication1(org.apache.hadoop.hbase.replication.TestMasterReplication)
Run 1:
TestMasterReplication.testCyclicReplication1:160->putAndWait:575->wait:631
Waited too much time for replication. Row:row1. IsDeleteReplication:false
Run 2: PASS
org.apache.hadoop.hbase.replication.TestMasterReplication.testHFileCyclicReplication(org.apache.hadoop.hbase.replication.TestMasterReplication)
Run 1:
TestMasterReplication.testHFileCyclicReplication:193->loadAndValidateHFileReplication:604->wait:614
Waited too much time for bulkloaded data replication. Current count=0,
expected count=200
Run 2: PASS
org.apache.hadoop.hbase.replication.TestMasterReplication.testHFileMultiSlaveReplication(org.apache.hadoop.hbase.replication.TestMasterReplication)
Run 1:
TestMasterReplication.testHFileMultiSlaveReplication:308->loadAndValidateHFileReplication:604->wait:614
Waited too much time for bulkloaded data replication. Current count=0,
expected count=200
Run 2: PASS
org.apache.hadoop.hbase.replication.TestPerTableCFReplication.testPerTableCFReplication(org.apache.hadoop.hbase.replication.TestPerTableCFReplication)
Run 1:
TestPerTableCFReplication.testPerTableCFReplication:419->putAndWaitWithFamily:568
Waited too much time for put replication
Run 2: PASS
{code}
I don't think any of these are related, and they passed when the entire suite
was run at the top of the build tree (labeled 'root' in the report).
> deprecate TableMapReduce.addDependencyJars(Configuration, class<?> ...)
> -----------------------------------------------------------------------
>
> Key: HBASE-8386
> URL: https://issues.apache.org/jira/browse/HBASE-8386
> Project: HBase
> Issue Type: Improvement
> Components: mapreduce
> Reporter: Nick Dimiduk
> Assignee: Sean Busbey
> Fix For: 2.0.0, 1.3.0, 1.4.0
>
> Attachments: HBASE-8386.1.patch
>
>
> We expose two public static methods names {{addDependencyJars}}. One of them,
> {{void addDependencyJars(Job}}, is very helpful -- goes out of its way to
> detect job dependencies as well as shipping all the necessary HBase
> dependencies. The other is shfty and nefarious, {{void
> addDependencyJars(Configuration, Class<?>...)}} -- it only adds exactly what
> the user requests, forcing them to resolve dependencies themselves and giving
> a false sense of security. We should deprecate the latter throw a big giant
> warning when people use that one. The handy functionality of providing help
> when our heuristics fail can be added via a new method signature, something
> like {{void addDependencyJars(Job, Class<?> ...}}. This method would do
> everything {{void addDependencyJars(Job}} does, plus let the user specify
> arbitrary additional classes. That way HBase still can help the user, but
> also gives them super-powers to compensate for when our heuristics fail.
> For reference, this appears to be the reason why HBase + Pig doesn't really
> work out of the box. See
> [HBaseStorage.java|https://github.com/apache/pig/blob/trunk/src/org/apache/pig/backend/hadoop/hbase/HBaseStorage.java#L730]
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
