[
https://issues.apache.org/jira/browse/HIVE-16295?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15965660#comment-15965660
]
Steve Loughran commented on HIVE-16295:
---------------------------------------
Thanks for starting this
1. We're making changes to FileOutputFormat so that it doesn't require an
instance of {{FileOutputCommitter}}, just any committer which also supplies a
working directory. This lets us add new committers alongside the existing one,
without playing games trying to subclass what is already a complex game.
1. All work is focused on getting the netflix "staging" Committer out the door
first; the other one, which I'd started before netflix offered theres, does
things inside S3a which could best be viewed as "dark magic". It will offer
even more performance, but I'm neglecting it for now. The netflix one is in use
in production, and has all its failure/abort algorithms thought out and
implemented.
1. I'm keeping the magic committer tests working, but not going to consider
that one ready to use until it passes lots of tests. Consider it a speedup for
the future.
The netflix committer itself has two subclasess, "directory" and "partitioned",
the directory one propagating a directory tree, the partitioned one expects
paths like "dateint=20161116/hour=14"; it has a different conflict policy than
the directory one.
Algorithm for the staging committer is
# tasks write to a local temp dir
# task abort: delete the files
# task commit: PUT the files as multipart uploads to their final destinations,
*do not commit the put*. Instead the data needed for the commit is saved to the
cluster FS, and committed using the normal algorithm
# job commit: load in the output of all committed tasks, commit them. Failure
to commit triggers revert: delete all files already committed, abort the rest
of the list.
# job abort: abort the output of all uncommitted tasks by reading in the files
and aborting those uploads.
# retry logic? Whatever we is implemented by the AWS SDK (mutliple attempts to
POST/PUT parts) and in S3A (retries of that final commit POST)
Nothing is visible until job commit; there's still a window of non-atomicity
there, but its the time for N posts where N=#of files; this can be parallelised
easily as it uses little bandwidth per post (unlike the uploads).
In tests, the dir committer works for the intermediate output of MR jobs saving
data to part-000x directories; the partitioned one good for spark output which
doesn't save the intermediate data, and wants to output partitioned style.
> Add support for using Hadoop's OutputCommitter
> ----------------------------------------------
>
> Key: HIVE-16295
> URL: https://issues.apache.org/jira/browse/HIVE-16295
> Project: Hive
> Issue Type: Sub-task
> Reporter: Sahil Takiar
> Assignee: Sahil Takiar
>
> Hive doesn't have integration with Hadoop's {{OutputCommitter}}, it uses a
> {{NullOutputCommitter}} and uses its own commit logic spread across
> {{FileSinkOperator}}, {{MoveTask}}, and {{Hive}}.
> The Hadoop community is building an {{OutputCommitter}} that integrates with
> S3Guard and does a safe, coordinate commit of data on S3 inside individual
> tasks (HADOOP-13786). If Hive can integrate with this new {{OutputCommitter}}
> there would be a lot of benefits to Hive-on-S3:
> * Data is only written once; directly committing data at a task level means
> no renames are necessary
> * The commit is done safely, in a coordinated manner; duplicate tasks (from
> task retries or speculative execution) should not step on each other
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
