[
https://issues.apache.org/jira/browse/HADOOP-15847?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16875810#comment-16875810
]
Steve Loughran commented on HADOOP-15847:
-----------------------------------------
Looking at the patch —I've made some comments
We should be setting the capacities to 0 in the core-site.xml and code; I'd
missed that in the HADOOP-15663 patch, where I did change core-default.xml and
defaults in Constants.
I've changed the title: if you could update the PR we'd be able to keep test
costs down to ~$0, even if a test run failed partway through
> S3Guard testConcurrentTableCreations to set r & w capacity == 0
> ---------------------------------------------------------------
>
> Key: HADOOP-15847
> URL: https://issues.apache.org/jira/browse/HADOOP-15847
> Project: Hadoop Common
> Issue Type: Sub-task
> Components: fs/s3, test
> Affects Versions: 3.2.0
> Reporter: Steve Loughran
> Assignee: lqjacklee
> Priority: Major
> Attachments: HADOOP-15847-001.patch, HADOOP-15847-002.patch
>
>
> I just found a {{testConcurrentTableCreations}} DDB table lurking in a
> region, presumably from an interrupted test. Luckily
> test/resources/core-site.xml forces the r/w capacity to be 10, but it could
> still run up bills.
> Recommend
> * explicitly set capacity = 1 for the test
> * and add comments in the testing docs about keeping cost down.
> I think we may also want to make this a scale-only test, so it's run less
> often
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
