Retain cause of bind failure in Server.bind

• Key: HADOOP-3434
• Type: Improvement
• Status: Closed
• Resolution: Fixed
• Priority: Major
• Assignee: Steve Loughran
• Reporter: Steve Loughran
• Affects Version/s: 0.18.0
• Fix Version/s: 0.18.0

Description

When a datanode can't start listening on a port, there's no explanation why; the message/stack is lost:

Caused by: java.net.BindException: Problem binding to localhost/127.0.0.1:8022
at org.apache.hadoop.ipc.Server.bind(Server.java:168)
at org.apache.hadoop.dfs.DataNode.startDataNode(DataNode.java:279)
at org.apache.hadoop.dfs.DataNode.(DataNode.java:185)
at org.apache.hadoop.dfs.ExtDataNode.(ExtDataNode.java:55)
at org.smartfrog.services.hadoop.components.datanode.DatanodeImpl.sfStart(DatanodeImpl.java:60)

Comments

[Steve Loughran]

patch the file to retain the exception text and trace

[Steve Loughran]

The patch changes the output to :

Caused by: java.net.BindException: Problem binding to localhost/127.0.0.1:8022 : Address already in use
at org.apache.hadoop.ipc.Server.bind(Server.java:168)
at org.apache.hadoop.dfs.DataNode.startDataNode(DataNode.java:279)
at org.apache.hadoop.dfs.DataNode.(DataNode.java:185)
at org.apache.hadoop.dfs.ExtDataNode.(ExtDataNode.java:55)
at org.smartfrog.services.hadoop.components.datanode.DatanodeImpl.sfStart(DatanodeImpl.java:60)
Caused by: java.net.BindException: Address already in use
at sun.nio.ch.Net.bind(Native Method)
at sun.nio.ch.ServerSocketChannelImpl.bind(ServerSocketChannelImpl.java:119)
at sun.nio.ch.ServerSocketAdaptor.bind(ServerSocketAdaptor.java:59)
at org.apache.hadoop.ipc.Server.bind(Server.java:166)

I havent supplied a test; you'd need to bring up two data nodes listening on the same port to test it. I could add a test for this in my project.

[Steve Loughran]

retains the underlying cause for a bind failure during name node startup

[Lohit Vijayarenu]

+1 Looks good.

[Chris Douglas]

I just committed this. Thanks, Steve

[Hudson]

Integrated in Hadoop-trunk #509 (See http://hudson.zones.apache.org/hudson/job/Hadoop-trunk/509/)