[
https://issues.apache.org/jira/browse/HADOOP-2119?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12567287#action_12567287
]
Amar Kamat commented on HADOOP-2119:
------------------------------------
I guess you meant 'spare matrix' conceptually. AFAIK java doesn't have any
inbuilt implementation for sparse matrix. This data structure will be optimized
for non-running tasks (virgin + failed) since
1) For virgin tasks there are no insertions
2) For failed tasks we just append in the front.
Wondering on the following points
1) I am not sure how easy it would be to insert into this structure.
2) Also how easily we can maintain the sortedness. This would be required for
failed/running tasks.
3) Also the implementation and space complexity will also be _HIGH_.
For HADOOP-2014, any of the designs mentioned in the table would work fine with
an extra count (in the TIP) of number of trackers having the split locally. I
think the strategy mentioned
[here|https://issues.apache.org/jira/browse/HADOOP-2014?focusedCommentId=12567015#action_12567015
] along with (8) [in the above table] for running tasks would be efficient in
terms of memory, space and implementation. Comments?
> JobTracker becomes non-responsive if the task trackers finish task too fast
> ---------------------------------------------------------------------------
>
> Key: HADOOP-2119
> URL: https://issues.apache.org/jira/browse/HADOOP-2119
> Project: Hadoop Core
> Issue Type: Bug
> Components: mapred
> Affects Versions: 0.16.0
> Reporter: Runping Qi
> Assignee: Amar Kamat
> Priority: Critical
> Fix For: 0.17.0
>
> Attachments: hadoop-2119.patch, hadoop-jobtracker-thread-dump.txt
>
>
> I ran a job with 0 reducer on a cluster with 390 nodes.
> The mappers ran very fast.
> The jobtracker lacks behind on committing completed mapper tasks.
> The number of running mappers displayed on web UI getting bigger and bigger.
> The jos tracker eventually stopped responding to web UI.
> No progress is reported afterwards.
> Job tracker is running on a separate node.
> The job tracker process consumed 100% cpu, with vm size 1.01g (reach the heap
> space limit).
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
