[
https://issues.apache.org/jira/browse/HADOOP-1085?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Arun C Murthy updated HADOOP-1085:
----------------------------------
Attachment: HADOOP-1085_20070311_1.patch
Here is a patch which removes 'port rolling' from the test-cases... a bit of a
work in progress since it seems to weirdly cause some streaming tests to fail.
Highlights:
1. Removes port-rolling by passing port '0' and letting the OS pick a ephemeral
port.
2. Changed org.apache.hadoop.ipc.Server to bind in its constructor and not in
the 'run' method to be sure of the ephemeral port.
3. Since the 'port' changes from the one passed (which previously is stored in
some data-members/configuration), this resets them in all necessary places.
4. Removed support for 'dataNodeFirst/taskTrackerFirst' flag to the
Mini{DFS|MR}Cluster since they cannot be supported now (we can't know the
NameNode/JobTracker ephemeral ports until we start them).
5. Fixed all the test-cases and deprecated the unecessary constructors in
Mini{DFS|MR}Cluster which cannot be supported.
> Remove 'port rolling' from Mini{DFS|MR}Cluster
> ----------------------------------------------
>
> Key: HADOOP-1085
> URL: https://issues.apache.org/jira/browse/HADOOP-1085
> Project: Hadoop
> Issue Type: Bug
> Components: test
> Affects Versions: 0.12.0
> Reporter: Arun C Murthy
> Assigned To: Arun C Murthy
> Fix For: 0.13.0
>
> Attachments: HADOOP-1085_20070311_1.patch
>
>
> The rolling of ports in these 2 clusters lead to a lot of timing issues and
> failed test cases; as witnessed in our patch process.
> The way around is to let the OS pick the port for the NameNode/JobTracker and
> let the the DataNode/TaskTracker query them for the port to connect to and
> then use that port.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
