hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Sanjay Radia (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-1985) Abstract node to switch mapping into a topology service class used by namenode and jobtracker
Date Thu, 07 Feb 2008 20:23:07 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-1985?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12566778#action_12566778
] 

Sanjay Radia commented on HADOOP-1985:
--------------------------------------


>1) On namenode startup, the namenode keeps on accepting datanode registrations and lazily
resolving them. (as in the patch)
>2) Block report is not processed if the datanode in question has not been resolved

Throwing a block report away because the NN has not figured out the DN's rack location is
an expensive proposition.
The reason being that block reports generate a lot of memory allocations and GC. 
Repeated BRs cause the startup time of the NN to increase. 
Hence when the registration comes in and if the rack location of the DN has not been determined,
the NN should NOT request a
block report. The block report should be requested at a future heartbeat when the rack location
has been determined.
The problem is that a reply of  "DNA Register" is interpreted as 
    a) please send registration
    b) please send BR (my recent patch added a random delay between steps a and b)

Looks like we need to separate (a) and (b).
This would require a new NN-DN command.

Perhaps:
DNA_REGISTER
DNA_BR

[DNA_REGISTER_AND_BR] -- i don't see how a combined command would be used.
Q. Does rack location determination occur AFTER the registration?




> Abstract node to switch mapping into a topology service class used by namenode and jobtracker
> ---------------------------------------------------------------------------------------------
>
>                 Key: HADOOP-1985
>                 URL: https://issues.apache.org/jira/browse/HADOOP-1985
>             Project: Hadoop Core
>          Issue Type: New Feature
>          Components: dfs, mapred
>            Reporter: eric baldeschwieler
>            Assignee: Devaraj Das
>             Fix For: 0.17.0
>
>         Attachments: 1985.new.patch, 1985.v1.patch, 1985.v10.patch, 1985.v11.patch, 1985.v2.patch,
1985.v3.patch, 1985.v4.patch, 1985.v5.patch, 1985.v6.patch, 1985.v9.patch, jobinprogress.patch
>
>
> In order to implement switch locality in MapReduce, we need to have switch location in
both the namenode and job tracker.  Currently the namenode asks the data nodes for this info
and they run a local script to answer this question.  In our environment and others that I
know of there is no reason to push this to each node.  It is easier to maintain a centralized
script that maps node DNS names to switch strings.
> I propose that we build a new class that caches known DNS name to switch mappings and
invokes a loadable class or a configurable system call to resolve unknown DNS to switch mappings.
 We can then add this to the namenode to support the current block to switch mapping needs
and simplify the data nodes.  We can also add this same callout to the job tracker and then
implement rack locality logic there without needing to chane the filesystem API or the split
planning API.
> Not only is this the least intrusive path to building racklocal MR I can ID, it is also
future compatible to future infrastructures that may derive topology on the fly, etc, etc...

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message