Remove the hardcoded ipc.client.connection.maxidletime setting from the TaskTracker.Child.main()

• Key: HADOOP-4040
• Type: Bug
• Status: Closed
• Resolution: Fixed
• Priority: Blocker
• Assignee: Devaraj Das
• Reporter: Devaraj Das
• Affects Version/s: 0.17.0
• Fix Version/s: 0.18.1

Description

This setting affects the dfs client idle timeout unless the jobconf of the user's job overrides it. I propose we remove this setting altogether.

Comments

[Raghu Angadi]

+1. This should go into 0.18.1 as well. The main problem is that it results in RPC calls timing out at client for wrong reasons. In DFS' case, is fatal for the tasks.

[Devaraj Das]

Attaching patches for 0.18 (this applies to trunk as well) and 0.17 (if the need is there for applying the patch on the 17 branch). The patches just removes the line that hardcodes the timeout from TaskTracker.java.

[Hadoop QA]

-1 overall. Here are the results of testing the latest attachment
http://issues.apache.org/jira/secure/attachment/12389162/4040-for0.17.patch
against trunk revision 690418.

+1 @author. The patch does not contain any @author tags.

-1 tests included. The patch doesn't appear to include any new or modified tests.
Please justify why no tests are needed for this patch.

-1 patch. The patch command could not apply the patch.

Console output: http://hudson.zones.apache.org/hudson/job/Hadoop-Patch/3147/console

This message is automatically generated.

[Owen O'Malley]

What is going wrong with this setting? It was originally put in to keep the task tracker connection to the job tracker from timing out. Is that still a concern?

[Owen O'Malley]

I just committed this. Thanks, Devaraj!

[Hudson]

Integrated in Hadoop-trunk #589 (See http://hudson.zones.apache.org/hudson/job/Hadoop-trunk/589/)