[
https://issues.apache.org/jira/browse/BIGTOP-1528?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14209211#comment-14209211
]
jay vyas commented on BIGTOP-1528:
----------------------------------
adding blockers. for the record might be worth it to just merge BIGTOP-1526
(failure property file) with the vagrant recipe, since that fix will be best
tested alongside a working vagrant recipe which actually spins up a 2 node
cluster , and runs the failure executor on it
> Add cluster failure testing to Vagrant / docker recipes
> -------------------------------------------------------
>
> Key: BIGTOP-1528
> URL: https://issues.apache.org/jira/browse/BIGTOP-1528
> Project: Bigtop
> Issue Type: Test
> Components: tests
> Affects Versions: 0.8.0
> Reporter: jay vyas
> Fix For: 0.9.0
>
>
> In BIGTOP-1525, we clarified that Failure injection is something we are 100%
> behind: So
> - lets make it a first class citizen by adding a vagrant demo of it in suite
> of vagrant recipes. Those vagrant recipes can of course easily be adopted
> for the CI as well, i think, once it comes online, so we are paving the way
> for roman's work by doing this imo.
> - Maintaining a failure injection environment is serious business. We want
> to know that it works - and only way to do that is to run mapreduce jobs on a
> multinode cluster - *unit tests for the feature are not enough*.
> - Lets solidify an example of it in our vagrant based tests, so that people
> can see the failure injection framework in action, and we can build a real
> community around it.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
