[
https://issues.apache.org/jira/browse/MAPREDUCE-4283?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13414823#comment-13414823
]
Siddharth Seth commented on MAPREDUCE-4283:
-------------------------------------------
Overall, lgtm. Minor nits.
- AggregatedLogFormat.getContainerLogsReader should accept a ContainerId
instead of a String AggLogFormat should be responsible for converting the
ContainerId to whatever format it uses.
- If there's nothing to read (start == end) -
AggregatedLogsBlock.readContainerLogs closes an unopened pre block.
Also, I think AggregatedLogsFormat should be Private / Unstable for now. Will
create a separate jira.
> Display tail of aggregated logs by default
> ------------------------------------------
>
> Key: MAPREDUCE-4283
> URL: https://issues.apache.org/jira/browse/MAPREDUCE-4283
> Project: Hadoop Map/Reduce
> Issue Type: Improvement
> Components: jobhistoryserver, mrv2
> Affects Versions: 0.23.3, 2.0.1-alpha
> Reporter: Jason Lowe
> Assignee: Jason Lowe
> Attachments: MAPREDUCE-4283.patch, MAPREDUCE-4283.patch,
> MAPREDUCE-4283.patch
>
>
> Similar to the manner in which the nodemanager webUI displays container logs,
> it would be very useful if the historyserver showed the trailing 4K or so of
> the aggregated logs with a link to see the full log.
> When debugging issues the relevant errors are usually at the end of the log,
> so showing just the last few K can enable quick diagnosis without waiting for
> what can be many megabytes of log data to download.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
