hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Koji Noguchi (JIRA)" <j...@apache.org>
Subject [jira] Commented: (HADOOP-4346) Hadoop triggers a "soft" fd leak.
Date Tue, 07 Oct 2008 16:36:45 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-4346?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12637539#action_12637539
] 

Koji Noguchi commented on HADOOP-4346:
--------------------------------------

bq. I guess our datanodes aren't that busy then

It also depends on the heapsize which would change the frequency of fullGC.
We use default heapsize of 1000M for the datanode.
Cluster that hit this issue was using 2048M (with 1024 file descriptor limit)


> Hadoop triggers a "soft" fd leak. 
> ----------------------------------
>
>                 Key: HADOOP-4346
>                 URL: https://issues.apache.org/jira/browse/HADOOP-4346
>             Project: Hadoop Core
>          Issue Type: Bug
>          Components: io
>    Affects Versions: 0.17.0
>            Reporter: Raghu Angadi
>            Assignee: Raghu Angadi
>         Attachments: HADOOP-4346-branch-18.patch, HADOOP-4346.patch
>
>
> Starting with Hadoop-0.17, most of the network I/O uses non-blocking NIO channels. Normal
blocking reads and writes are handled by Hadoop and use our own cache of selectors. This cache
suites well for Hadoop where I/O often occurs on many short lived threads. Number of fds consumed
is proportional to number of threads currently blocked. 
> If blocking I/O is done using java.*, Sun's implementation uses internal per-thread selectors.
These selectors are closed using {{sun.misc.Cleaner}}. Looks like this cleaning is kind of
like finalizers and tied to GC. This is pretty ill suited if we have many threads that are
short lived. Until GC happens, number of these selectors keeps growing. Each selector consumes
3 fds.
> Though blocking read and write are handled by Hadoop, {{connect()}} is still the default
implementation that uses per-thread selector. 
> Koji helped a lot in tracking this. Some sections from 'jmap' output and other info 
Koji collected led to this suspicion and will include that in the next comment.
> One solution might be to handle connect() also in Hadoop using our selectors.

-- 
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