[
https://issues.apache.org/jira/browse/HADOOP-4189?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12632506#action_12632506
]
Amar Kamat commented on HADOOP-4189:
------------------------------------
bq. Null config values are not the right way to go. At best it should be "-1"
I am not sure if -1 internally selects the system defaults. If nothing is
passed, by default the history code selects filesystem's default values for
block size and buffer.
bq. In these specific cases I'm not sure we even need to make them configurable.
With the system defaults, the % recovery is very less as history info is not
frequently flushed to the disk. So controlling these parameters, one can
control the extent of recovery. Again there is a tradeoff between how much to
recover versus the dfs performance w.r.t low block-size/buffer-size.
bq. Do we need to bump up TaskUmbilicalProtocol's version?
I have added the comment in HADOOP-3245 but forgot to bump up the version
number. So fixing it in this patch.
> HADOOP-3245 is incomplete
> -------------------------
>
> Key: HADOOP-4189
> URL: https://issues.apache.org/jira/browse/HADOOP-4189
> Project: Hadoop Core
> Issue Type: Bug
> Components: mapred
> Reporter: Amar Kamat
> Assignee: Amar Kamat
> Priority: Blocker
> Fix For: 0.19.0
>
> Attachments: HADOOP-4189-v1.patch, HADOOP-4189-v2.patch,
> HADOOP-4189-v3.1.patch, HADOOP-4189-v3.patch, HADOOP-4189.patch
>
>
> There are 2 issues with HADOOP-3245
> - The default block size for the history files in hadoop-default.conf is set
> to 0. This file will be empty throughout. It should be made null.
> - Same goes for the buffer size.
> - InterTrackerProtocol version needs to be bumped.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
