Hi Team,

I was just taking a look over the roadmap for Whirr 0.8.0. I think the top
priority should addressing the lack of deterministic behaviour on cluster
deployment - service start. A cluster should either start as expected or
fail all the time.

Some of the relevant issues are:

* https://issues.apache.org/jira/browse/WHIRR-221
* https://issues.apache.org/jira/browse/WHIRR-266
* https://issues.apache.org/jira/browse/WHIRR-295
* https://issues.apache.org/jira/browse/WHIRR-294
* https://issues.apache.org/jira/browse/WHIRR-468

One possible workaround the we should be able to implement pretty soon is
to give Whirr the ability to restart a specific role.

I am also thinking that as soon as we close the listed issues we can cut a
release.

What do you think?

-- Andrei Savu / andreisavu.ro

Reply via email to