On Wed, Apr 7, 2010 at 1:39 PM, Andreas Jung <li...@zopyx.com> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
> According to the ZODB 3.10 release notes we will have dedicated
> processes on the ZEO server side for each client connection.
Where do they say that?
They say that "ZEO clients now work in forked processes." Is that what
you're referring to? That should probably have been worded
differently: "ZEO clients now work properly in forked processes."
In 3.10, ZEO servers allocate a thread for each client, but the change
notes don't mention that.
> What will
> be the major benefits of this change?
Making the server multi-threaded provides better performance.
> Our typical scenario is like this: we run 5-6 ZEO servers where each
> server manages one fat storage (50-100GB). Roughly 20-30 ZEO clients are
> connected to each ZEO server (10 ZEO clients with 2-3 works threads
> each) doing reads and writes plus a certain number (perhaps 5-10) ZEO
> clients doing mass export/read operations. Should we see any benefits
> from using a ZODB 3.10 server in such a scenario?
For more information about ZODB, see the ZODB Wiki:
ZODB-Dev mailing list - ZODB-Dev@zope.org