On Mon, 29 Nov 1999, Tim Bunce wrote: > Ignoring 'thread' (unsafe for production use) and 'debug' modes, the > normal 'fork' mode means that each client gets a seperate ProxyServer > process. And because of that, clients have no way to share connections > with each other. Is that necessarily the case? Perhaps the parent server could handle the connection-pooling and hand off live connection to the forked child. That way you could get the best of both worlds - de-coupled clients and a shared pool of database connections. It seems to me that moving to a single-process architecture has some BIG downsides when you consider the wide variation in database query runtimes. -sam
- pool of DB connections ? Oleg Bartunov
- Re: pool of DB connections ? Leslie Mikesell
- Re: pool of DB connections ? Oleg Bartunov
- RE: pool of DB connections ? Ed Park
- Re: pool of DB connections ? Randal L. Schwartz
- RE: pool of DB connections ? Chris Nokleberg
- Limitations of DBI::ProxyServer (... Tim Bunce
- Re: Limitations of DBI::Proxy... Randal L. Schwartz
- Re: Limitations of DBI::Proxy... Sam Tregar
- Re: Limitations of DBI::P... Tim Bunce
- Re: Limitations of DBI::P... Sam Tregar
- Re: Limitations of DBI::P... Jonathan Leffler
- Re: Limitations of DBI::Proxy... G.W. Haywood
- Re: Limitations of DBI::P... Jochen Wiedmann
- Re: pool of DB connections ? Leslie Mikesell
- Re: pool of DB connections ? Matt Sergeant
- RE: pool of DB connections ? Sheth, Niraj
- RE: pool of DB connections ? Oleg Bartunov