I think a lot of haproxy users employ the use of intelligent health checks. A health check could be a php,python...whatever script that can assess current load and return a 204 or 503..(or some other appropriate down http response.)
On Mon, Aug 30, 2010 at 11:12 AM, Ben <[email protected]> wrote: > Hi -- Haven't heard a response to this message and wondering if it got > bounced or something. > > Ben > > On Mon, Aug 23, 2010 at 11:36 PM, Ben <[email protected]> wrote: >> Hello -- >> >> Are there any provisions for distributing connections according to >> external metrics, e.g. load as reported by uptime or other utilities? >> Obviously this would have some scalability issues but I'm struggling >> with a way to balance between database servers on heterogenous >> hardware. >> >> Best regards, Ben >> > >
