performance issue (imap spool on san)

[Phil Pennock]

On 2006-07-28 at 11:05 +0100, Andrew Findlay wrote:
> Headers are not usually very large, so I would be more inclined to
> the idea that the index should store every header (perhaps with a
> blocklist to avoid things like Received:)

Those fine folk at Cambridge who introduced replication have a number of
other patches; what you suggest is in fact done:
 <URL:http://www-uxsup.csx.cam.ac.uk/~dpc22/cyrus/performance.html>

and there's a lot more interesting stuff on David Carter's cyrus pages.
Top-level, predictably:
 <URL:http://www-uxsup.csx.cam.ac.uk/~dpc22/cyrus/index.html>
 http://www-uxsup.csx.cam.ac.uk/~dpc22/cyrus/index.html

-- 
"Everything has three factors: politics, money, and the right way to do it.
 In that order."  -- Gary Donahue