[
https://issues.apache.org/jira/browse/WHIRR-301?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13054559#comment-13054559
]
Andrei Savu commented on WHIRR-301:
-----------------------------------
There are now 28 integration testing scenarios we should consider: 2 (cloud
providers) x 2 (operating systems) x 7 (services). I suggest that on EC2 we
should test against a Canonical Ubuntu AMI (10.04 LTS or 10.10) and the Amazon
Linux AMI (is there a standard Centos 5.x AMI available?). For cloudservers
it's clear: ubuntu 10.04 and centos 5.5.
Hopefully soon we are going to be able to rely more on Bigtop for testing
packaged artifacts on multiple operating systems.
> Improve code testability
> -------------------------
>
> Key: WHIRR-301
> URL: https://issues.apache.org/jira/browse/WHIRR-301
> Project: Whirr
> Issue Type: Improvement
> Reporter: Andrei Savu
> Assignee: Andrei Savu
>
> I'm thinking about the following changes:
> * use in-memory cluster state store in all service integration tests
> * it should be possible to run integration tests in parallel (easy with no
> state and port conflicts)
> * each service integration test should be executed on multiple OSes: Ubuntu /
> Debian and Centos. We should be able to set a flag to get this behavior.
> * refactor and write missing tests for ClusterStateStore (as discussed in
> WHIRR-288)
> I'm going to create a subtask for each fix.
> What do you think? What else should we do to improve the testing process?
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
