[imapproxy] (no subject)

[Johny]

Good suggestion;

I just tested on a folder with close to 700 messages, moving from page to page
(I've got 20 messages per page in IMP). It took 1.95 (+- 5ms) seconds without
imapproxy, 2.95 (+-5ms) seconds with imapproxy.

The folder itself is 11.221.552 bytes, so reasonably sized.

Can't see how the proxy can speed anything up from where I sit ;) If the login
time was significant I'd likely have a look at what the proxy does with a view
to improving, as is the performance is significantly better in all my scenarious
without the thing there.

:)Johny

Quoting Anil Madhavapeddy <anil at recoil.org>:

> On Sat, May 24, 2003 at 10:10:29PM +1000, Johny wrote:
> > Which indicates a major increase in time caused by the proxy relaying the
> > message headers, this outweighs by a long margin the overhead of starting a
> new
> > imapd for every request and to my mind negates the value of the imapproxy
> in
> > it's current form. I didn't do any significant profiling on imapproxy,
> but
> > keeping an eye on 'top' didn't show a significant amount of CPU time used
> by the
> > proxy.
> > 
> 
> Out of interest, have you tried measuring on a mailbox with a 
> large number of messages (>1000) ?  I notice that these are the
> biggest culprits for slow responses when using IMAP; the common
> case of small folders is always nice and fast.
> 
> -- 
> Anil Madhavapeddy, http://anil.recoil.org
> University of Cambridge, http://www.cl.cam.ac.uk
>