Carl Bowden wrote:
> > It's the switching specifically that fails? There must
> > have been
> > 'opening' statements vefore this, were these OK then?'
>
> yes. the sites are work fine for serveral hours. it seems to
> be when there has been a quite patch and then apache has to
> respawn more treads ( or may be has not kill the old ones
> correctly? )
I can't rule it out, but it's not too likely. Switching only happens
after the DB connection has been established within the handler child
previously.
> its also seems to be fine when there is heaps
> of activity one DB (say SmartWEB_001) the errors start when
> some connects to a site from another DB (say SmartWEB_002
> which was going ok when everything started up)
I don't think it's related, but you never know: are you using the same
password/username for MidgardDatabase, or are yo uusing unique
combinations for each? If you're using unique combinations, make sure
MySQL allows for plenty of connections. I usually set things up with a
common user/pass for each DB to minimize connections.
Emile
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
