Open Stack

On 2015-05-08 18:54:25 +0800 (+0800), Li Tianqing wrote:
[...]
> I argue that why we do not do this in upstream. For that most
> production do this.

Where? Do you mean recording these recommendations in deployment
documentation? Or setting it up that way in our integration tests?

> And if you do this you will find that there are many work need do.
> The community applies the laziest implementation.
[...]

It sounds like the community (in this case the operators deploying
and running Trove in production on large scales at various OpenStack
Foundation member companies) have this working. Or by "the
community" do you mean something else entirely? Perhaps DevStack?
That's not _meant_ to be a production deployment, just a means of
getting the service started as simply as possible so as to be able
to test patches for it.
-- 
Jeremy Stanley