[
https://issues.apache.org/jira/browse/HADOOP-12325?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14708852#comment-14708852
]
Akira AJISAKA commented on HADOOP-12325:
----------------------------------------
Thanks [~anu] for creating the patch. Two comments:
1. Would you add a whitespace before "took " in the log message?
{code}
+ LOG.warn(
+ "Slow RPC : " + methodName + "took " + processingTime +
+ " milliseconds to process from client " + client);
{code}
2. After running the regression test locally, I can't see any logs about sleep
RPC. I'm thinking the test should output the log as follows:
{noformat}
2015-08-24 15:16:31,771 WARN ipc.Server (Server.java:logSlowRpcCalls(438)) -
Slow RPC : sleep 1001 milliseconds to process from client 127.0.0.1:52257
{noformat}
Would you fix this?
> RPC Metrics : Add the ability track and log slow RPCs
> -----------------------------------------------------
>
> Key: HADOOP-12325
> URL: https://issues.apache.org/jira/browse/HADOOP-12325
> Project: Hadoop Common
> Issue Type: Improvement
> Components: ipc, metrics
> Affects Versions: 2.7.1
> Reporter: Anu Engineer
> Assignee: Anu Engineer
> Attachments: Callers of WritableRpcEngine.call.png,
> HADOOP-12325.001.patch, HADOOP-12325.002.patch, HADOOP-12325.003.patch,
> HADOOP-12325.004.patch, HADOOP-12325.005.patch
>
>
> This JIRA proposes to add a counter called RpcSlowCalls and also a
> configuration setting that allows users to log really slow RPCs. Slow RPCs
> are RPCs that fall at 99th percentile. This is useful to troubleshoot why
> certain services like name node freezes under heavy load.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
