hadoop-mapreduce-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Amareshwari Sriramadasu (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (MAPREDUCE-2765) DistCp Rewrite
Date Mon, 08 Aug 2011 12:06:27 GMT

    [ https://issues.apache.org/jira/browse/MAPREDUCE-2765?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13080917#comment-13080917
] 

Amareshwari Sriramadasu commented on MAPREDUCE-2765:
----------------------------------------------------

First of all, the code needs go into a contrib project. So, you need to regenerate the patch
putting the code in contrib.
Also, build environment needs changes. Will this be blocked on mavenization of MapReduce?

Overall, design looks fine. Here are some comments on the code:
* CopyMapper:
  ** 
{noformat}
    if (targetFS.exists(targetFinalPath) && targetFS.isFile(targetFinalPath)) {
      overWrite = true; // When target is an existing file, overwrite it.
    }
{noformat}
Target file is overwritten irrespective of overwrite configuration? why?

* Dynamic\*
  ** DynamicInputChunk is not public?
  ** DynamicInputFormat creates FileSplits with zero length. Instead should it be created
with the size of chunk as the size of the split.
  ** DynamicRecordReader has commented code. Should remove it.

* CopyCommitter:
  ** Atomic commit should not delete the final directory. Should throw out an error if it
exists even before starting the job.
  ** deleteMissing() counts the files which do not exists at both source and target paths
as deleted entries.
  ** Preserving status for the root folder does not happen at all? Can you check?
  ** If I’m not wrong, preserveFileAttributes() does preserve only for directories. Can
we rename the method accordingly?
  ** The methods deleteMissing(), preserveFileAttributes() etc need more doc.
  ** Deleting attempt temp files happens in each attempt. Why are we doing delete again in
Committer? Committer should just delete the work path.

General comment:
All public classes and public methods need javadoc

Haven't looked at testcases.

> DistCp Rewrite
> --------------
>
>                 Key: MAPREDUCE-2765
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-2765
>             Project: Hadoop Map/Reduce
>          Issue Type: New Feature
>          Components: distcp
>    Affects Versions: 0.20.203.0
>            Reporter: Mithun Radhakrishnan
>            Assignee: Mithun Radhakrishnan
>         Attachments: distcpv2.20.203.patch
>
>
> This is a slightly modified version of the DistCp rewrite that Yahoo uses in production
today. The rewrite was ground-up, with specific focus on:
> 1. improved startup time (postponing as much work as possible to the MR job)
> 2. support for multiple copy-strategies
> 3. new features (e.g. -atomic, -async, -bandwidth.)
> 4. improved programmatic use
> Some effort has gone into refactoring what used to be achieved by a single large (1.7
KLOC) source file, into a design that (hopefully) reads better too.
> The proposed DistCpV2 preserves command-line-compatibility with the old version, and
should be a drop-in replacement.
> New to v2:
> 1. Copy-strategies and the DynamicInputFormat:
> 	A copy-strategy determines the policy by which source-file-paths are distributed between
map-tasks. (These boil down to the choice of the input-format.) 
> 	If no strategy is explicitly specified on the command-line, the policy chosen is "uniform
size", where v2 behaves identically to old-DistCp. (The number of bytes transferred by each
map-task is roughly equal, at a per-file granularity.) 
> 	Alternatively, v2 ships with a "dynamic" copy-strategy (in the DynamicInputFormat).
This policy acknowledges that 
> 		(a)  dividing files based only on file-size might not be an even distribution (E.g.
if some datanodes are slower than others, or if some files are skipped.)
> 		(b) a "static" association of a source-path to a map increases the likelihood of long-tails
during copy.
> 	The "dynamic" strategy divides the list-of-source-paths into a number (> nMaps) of
smaller parts. When each map completes its current list of paths, it picks up a new list to
process, if available. So if a map-task is stuck on a slow (and not necessarily large) file,
other maps can pick up the slack. The thinner the file-list is sliced, the greater the parallelism
(and the lower the chances of long-tails). Within reason, of course: the number of these short-lived
list-files is capped at an overridable maximum.
> 	Internal benchmarks against source/target clusters with some slow(ish) datanodes have
indicated significant performance gains when using the dynamic-strategy. Gains are most pronounced
when nFiles greatly exceeds nMaps.
> 	Please note that the DynamicInputFormat might prove useful outside of DistCp. It is
hence available as a mapred/lib, unfettered to DistCpV2. Also note that the copy-strategies
have no bearing on the CopyMapper.map() implementation.
> 	
> 2. Improved startup-time and programmatic use:
> 	When the old-DistCp runs with -update, and creates the list-of-source-paths, it attempts
to filter out files that might be skipped (by comparing file-sizes, checksums, etc.) This
significantly increases the startup time (or the time spent in serial processing till the
MR job is launched), blocking the calling-thread. This becomes pronounced as nFiles increases.
(Internal benchmarks have seen situations where more time is spent setting up the job than
on the actual transfer.)
> 	DistCpV2 postpones as much work as possible to the MR job. The file-listing isn't filtered
until the map-task runs (at which time, identical files are skipped). DistCpV2 can now be
run "asynchronously". The program quits at job-launch, logging the job-id for tracking. Programmatically,
the DistCp.execute() returns a Job instance for progress-tracking.
> 	
> 3. New features:
> 	(a)   -async: As described in #2.
> 	(b)   -atomic: Data is copied to a (user-specifiable) tmp-location, and then moved atomically
to destination.
> 	(c)   -bandwidth: Enforces a limit on the bandwidth consumed per map.
> 	(d)   -strategy: As above.    
> 	
> A more comprehensive description the newer features, how the dynamic-strategy works,
etc. is available in src/site/xdoc/, and in the pdf that's generated therefrom, during the
build.
> High on the list of things to do is support to parallelize copies on a per-block level.
(i.e. Incorporation of HDFS-222.)
> I look forward to comments, suggestions and discussion that will hopefully ensue. I have
this running against Hadoop 0.20.203.0. I also have a port to 0.23.0 (complete with unit-tests).
> P.S.
> A tip of the hat to Srikanth (Sundarrajan) and Venkatesh (Seetharamaiah), for ideas,
code, reviews and guidance. Although much of the code is mine, the idea to use the DFS to
implement "dynamic" input-splits wasn't.
> 	

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira

       

Mime
View raw message