[
https://issues.apache.org/jira/browse/HADOOP-1504?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Tom White updated HADOOP-1504:
------------------------------
Attachment: HADOOP-1504.patch
This patch fixes this serious bug by doing two things:
1. Restricting the instances that are to be terminated to the version of the
Hadoop AMI specified in the config file.
2. Listing the instances that will be terminated and prompting the user for
agreement to proceed.
I have also updated the wiki to draw attention to this bug.
> terminate-hadoop-cluster may be overzealous
> -------------------------------------------
>
> Key: HADOOP-1504
> URL: https://issues.apache.org/jira/browse/HADOOP-1504
> Project: Hadoop
> Issue Type: Bug
> Components: contrib/s3
> Affects Versions: 0.13.0
> Reporter: Doug Cutting
> Assignee: Tom White
> Priority: Blocker
> Fix For: 0.14.0
>
> Attachments: HADOOP-1504.patch
>
>
> If folks are using EC2 for things besides Hadoop, then the
> terminate-hadoop-cluster script (in src/contrib/s3/bin) will kill all
> instances, not just Hadoop instances. This could cause loss of work.
> http://developer.amazonwebservices.com/connect/thread.jspa?threadID=15699
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
