Recommendations for Bugzilla hardware

[Max Kanat-Alexander]

On Wed, 2005-10-19 at 09:59 +0800, byron wrote:
> has anyone benchmarked mysql vs postgresql?

	I did some very brief, somewhat crude testing just now on landfill.

	Basically, I just ran a simple query over and over:

time ./buglist.cgi "resolution=---&order=bugs.bug_id" > /dev/null

	A query that returns 1750 bugs, when run about five times in a row,
averages out to this many "user" seconds for each DB:

	PostgreSQL: 3.1
	MySQL: 3.4

	That's entirely consistent with the times that it takes if I restart
the database servers and run the query "for the first time."

	I'm not sure how that would change under load, but since PostgreSQL has
a less-strict locking model, it might do better under a lot of readers
for Bugzilla than MySQL does.

	I'd be interested in further, more accurate benchmarking data, though.

	-Max
-- 
http://www.everythingsolved.com/
Competent, Friendly Bugzilla Services. And Everything Else, too.