[ 
https://issues.apache.org/jira/browse/MAPREDUCE-6622?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15159483#comment-15159483
 ] 

Ray Chiang commented on MAPREDUCE-6622:
---------------------------------------

In order to minimize impact, I didn't really do any real rewrite of the 
internals of CachedHistoryStorage.  I just swapped in the cache access.

Here's my understanding of getFullJob():
- It's using the HistoryFileManager "fileInfo" as a proxy to determine whether 
the job exists.
- If the job exists, it then checks the cache via loadedJobCache#getIfPresent()
- If the job isn't in the cache, it loads the file into memory
- If job history file has been deleted, getFullJob returns null


> Add capability to set JHS job cache to a task-based limit
> ---------------------------------------------------------
>
>                 Key: MAPREDUCE-6622
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-6622
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>          Components: jobhistoryserver
>    Affects Versions: 2.7.2
>            Reporter: Ray Chiang
>            Assignee: Ray Chiang
>              Labels: supportability
>         Attachments: MAPREDUCE-6622.001.patch, MAPREDUCE-6622.002.patch, 
> MAPREDUCE-6622.003.patch, MAPREDUCE-6622.004.patch, MAPREDUCE-6622.005.patch
>
>
> When setting the property mapreduce.jobhistory.loadedjobs.cache.size the jobs 
> can be of varying size.  This is generally not a problem when the jobs sizes 
> are uniform or small, but when the job sizes can be very large (say greater 
> than 250k tasks), then the JHS heap size can grow tremendously.
> In cases, where multiple jobs are very large, then the JHS can lock up and 
> spend all its time in GC.  However, since the cache is holding on to all the 
> jobs, not much heap space can be freed up.
> By setting a property that sets a cap on the number of tasks allowed in the 
> cache and since the total number of tasks loaded is directly proportional to 
> the amount of heap used, this should help prevent the JHS from locking up.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to