[
https://issues.apache.org/jira/browse/HIVE-4920?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13721442#comment-13721442
]
Brock Noland commented on HIVE-4920:
------------------------------------
:)
Yeah I was thinking about it and I think what we should do instead is replace
our set of slaves if they are all lost due to a price spike and then block
while we cannot get hosts due to price. I have this patch working and will
submit it soon.
> PTest2 spot instances should fall back on c1.xlarge and then on-demand
> instances
> --------------------------------------------------------------------------------
>
> Key: HIVE-4920
> URL: https://issues.apache.org/jira/browse/HIVE-4920
> Project: Hive
> Issue Type: Improvement
> Reporter: Brock Noland
> Assignee: Brock Noland
> Priority: Critical
> Attachments: Screen Shot 2013-07-23 at 3.35.00 PM.png
>
>
> Today the price for m1.xlarge instances has been varying dramatically. We
> should fall back on c1.xlarge (which is more powerful and is cheaper at
> present) and then on on-demand instances.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
