[ 
https://issues.apache.org/jira/browse/HDFS-15402?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17129509#comment-17129509
 ] 

Wei-Chiu Chuang commented on HDFS-15402:
----------------------------------------

Initially I thought it was broken by HADOOP-15696 (KMS, httpfs and Hadoop web 
UI all use Jetty) but no. Hadoop web UI defaults to 10 second idle time out and 
the patch doesn't change the behavior of Hadoop web UI.

Instead, something was probably broken when we updated Jetty from 6 to 9 in 
Hadoop 3.0

> Requesting http jmx metrics leads to too much CLOSE-WAIT on datanode
> --------------------------------------------------------------------
>
>                 Key: HDFS-15402
>                 URL: https://issues.apache.org/jira/browse/HDFS-15402
>             Project: Hadoop HDFS
>          Issue Type: Bug
>          Components: metrics
>    Affects Versions: 3.1.3
>            Reporter: Sean Chow
>            Priority: Major
>
> We access  {{http://127.0.0.1:50075/jmx}}  to get datanode metrics 
> periodically. But there is too much CLOSE-WAIT socket state that lead the 
> normal webhdfs request failed.
>  
> {code:java}
> $ ss -ant|grep 127.0.0.1:50075 |grep CLOSE-WAIT |head -10
> CLOSE-WAIT 122 0 127.0.0.1:50075 127.0.0.1:37296 
> CLOSE-WAIT 122 0 127.0.0.1:50075 127.0.0.1:26499 
> CLOSE-WAIT 122 0 127.0.0.1:50075 127.0.0.1:47470 
> CLOSE-WAIT 122 0 127.0.0.1:50075 127.0.0.1:42852 
> CLOSE-WAIT 122 0 127.0.0.1:50075 127.0.0.1:40281
> $ ss -ant|grep 127.0.0.1:50075 |grep CLOSE-WAIT | wc -l 
> 6729
> lsof -i:37296
> COMMAND PID USER FD TYPE DEVICE SIZE/OFF NODE NAME
> java 101015 hdfs 3044u IPv4 271157177 0t0 TCP 
> localhost:50075->localhost:37296 (CLOSE_WAIT)
> {code}
>  
> The pid 101015 is the datanode's process id.
> I use {{cdh6.1.1}} and {{apache-hadoop-3.1.3}} in my production, and both of 
> them have the same issue. When the metric retriving script stop, the number 
> of CLOSE-WAIT does not increase anymore.
>  The version apache-hadoop-2.9.2 does not have this issue with the same 
> retriving metric script.
>  



--
This message was sent by Atlassian Jira
(v8.3.4#803005)

---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org

Reply via email to