nova team mailing list archive

[Ewan Mellor <ewan.mellor@xxxxxxxxxxxxx>]

On Fri, Sep 10, 2010 at 05:56:45PM +0100, Jay Pipes wrote:

> [Snip orm-refactor discussion]
> 
> So...what to do?  There are at least four options I can see:
> 
> 1) Go forward with this patch and add NoSQL stores back at some later
> time by ourselves

I just want to point out that there are two options here:

1a) Add NoSQL stores support under the db/api.py abstraction
1b) Add NoSQL stores support to SQLalchemy

> 2) Go forward with this patch and wait until SQLAlchemy properly
> supports key value stores
> 3) Delay this patch until after the Austin release and have a larger
> discussion about it here and at the next summit
> 4) Go back to the drawing board and try again with a less ambitious
> set of patches that incrementally changes the way the data stores
> work.


I'm offering no opinion on NoSQL vs SQL.  However, I would like to beg that
you don't fork at this early stage, even if only temporarily.  I would rather
delay the Austin release than release it prematurely and not be able to claim
that our primary contributors are right behind it.

Also, regardless of NoSQL vs SQL, one thing that I think we should protect is
our mantra of "shared nothing".  Regardless of which solution we choose, we
should be able to argue how we're moving in the direction of shared-nothing.
It's an important design choice that differentiates OpenStack from other
platforms.

Cheers,

Ewan.