[
https://issues.apache.org/jira/browse/HBASE-12718?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14254476#comment-14254476
]
stack commented on HBASE-12718:
-------------------------------
bq. To have it in both, I'd need to do some refactoring.
Yep. No harm, right?
bq. Unit tests ideally tell us where.
Yeah, but until we have that, leave TAG where it is? If TAG starts failing
after a particular patch has been committed, thats good enough signal the patch
has mangled our ACID-on-a-row.
TAG would fail on occasion. Can't finger it identifying a bad patch but now we
are in a new realm of passing unit tests most of the time, it could find a
dirty commit. Filed HBASE-12735 and assigned yourself [~jmhsieh]
> Convert TestAcidGuarantees from a unit test to an integration test
> ------------------------------------------------------------------
>
> Key: HBASE-12718
> URL: https://issues.apache.org/jira/browse/HBASE-12718
> Project: HBase
> Issue Type: Bug
> Components: hbase, integration tests, test
> Affects Versions: 1.0.0, 2.0.0
> Reporter: Jonathan Hsieh
> Assignee: Jonathan Hsieh
> Fix For: 2.0.0, hbase-11339, 1.1.0
>
> Attachments: hbase-12718.patch
>
>
> TestAcidGurantees has a main method so that it can be executed via the
> commandline. In the past this was run and we'd use external tools to inject
> faults while it executed for an extended period of time. We've had the IT
> framework we'd like to use the ChaosMonkey automation with it.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
