On 28 August 2012 06:34, Iain Duncan <[email protected]> wrote:
> I expect this project will be doing a *lot* of writes. IE, maybe even all
> hits do some minor mutation. On the other hand, I have a hunch that the
> architecture could wind up being really clean in zodb as it's going to be a
> lot of object modelling (some non-realtime simulation related stuff) And I
> think transactions will be important, it will be all screwed up if there's
> any kind integrity jiggery pokery in the object model.

The ZODB is not optimized for this use-case. There's a couple of
reasons for this:

1. You'll get B-tree conflicts that can't be resolved on the server.
2. You'll need to load non-current object data too frequently, which
is expensive.
3. There's a global commit lock (primarily to establish order) which
incurs some latency.

You'll probable fare better with PostgreSQL because it's got knobs you
can tune and native support for concurrent B-trees.

\malthe

-- 
You received this message because you are subscribed to the Google Groups 
"pylons-discuss" group.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/pylons-discuss?hl=en.

Reply via email to