[
https://issues.apache.org/jira/browse/HBASE-8258?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13623739#comment-13623739
]
Jonathan Hsieh commented on HBASE-8258:
---------------------------------------
bq. As I said above, 0.95 needs a stable hadoop 2 release (2.0.4-alpha).
2.0.4 is not released yet http://hadoop.apache.org/releases.html. When it is,
we'll deal with it.
bq. SNAPSHOT artifacts would allow us to find the problem much sooner.
-1 to building hbase on SNAPSHOT default.
bq. One such config param is YarnConfiguration.IS_MINI_YARN_CLUSTER which must
be set true in order for MiniMRCluster to work.
>From my experiments I'm not sure if that is relevant to the current problems.
I've narrowed it down to only needing to set these two yarn configs (because
something overrides their defaults). All the rest of the yarn config values
are not needed.
{code}
yarn.resourcemanager.scheduler.address : localhost:49649
yarn.resourcemanager.address : localhost:48993
{code}
yarn.rm.schedule.addrss is default port 8030 and yarn.resourcemanager.address
is default port 8032. The need for the different value here explains why we
get the ConnectException from the mr client trying to talk to 0.0.0.0:8032.
The follow-on questions this leads me to are:
* Why is do we need to override these in our minimrcluster tests?
* Why isn't this need documented and where should it be?
* Why isn't this a problem on other HBase-MR jobs?
* What can we do to properly propagate this so that future MR tests won't have
this problem?
I feel that these need to be answered and hopefully addressed before we commit
a fix.
bq. Can you explain folding a little bit ?
I'm using folding and merging as synonyms -- what the copyConfigurationValues
does.
> Make mapreduce tests pass on hadoop2
> ------------------------------------
>
> Key: HBASE-8258
> URL: https://issues.apache.org/jira/browse/HBASE-8258
> Project: HBase
> Issue Type: Bug
> Components: mapreduce
> Reporter: stack
> Priority: Blocker
> Fix For: 0.95.1
>
> Attachments: 8258-plain.txt, 8258-v1-hadoop-2.0.txt,
> 8258-v2-hadoop-2.0.txt
>
>
> HBASE-7904 was a first attempt at making this work but it got lost in the
> weeds.
> This is a new attempt at making hbase mapreduce jobs run on hadoop2 (w/o
> breaking mapreduce on hadoop1)
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
