Correct, but that behavior, (not being able to reset credentials) is a "database implementation detail". ;-) Imagine if SQL Server could reset credentials, in a manner analogous to how sp_reset_connection works today. Then it would be possible to write a connection pooling implementation that works the way ChongQing wants it to. It doesn't, of course, work that way today.
OTOH, reset of credentials (and flushing user specific database objects, etc) might require enough database "bandwidth" that it would invalidate the cost-saving (which considers session setup overhead in addition to other items) of pooling. Cheers, Bob Beauchemin [EMAIL PROTECTED] -----Original Message----- From: Marsh, Drew [mailto:[EMAIL PROTECTED]] Sent: Friday, May 10, 2002 10:39 AM To: [EMAIL PROTECTED] Subject: Re: [DOTNET] question about connection pool ChongQing Xiao [mailto:[EMAIL PROTECTED]] wrote: > In ADO.NET, the connection pool will be used if the > connection string is the same. This works fine > if I can map many different user to the same userID in the > Database. But how it will work if I want to distinguish > different user in the database (say for audit purpose)? If I > map one end user to one database user, can the connection > still be reused? They can't be shared between database users anymore because the connection has been established with the set of creditials supplied. So, each database user will have their own pool (puddle?) of connections. HTH, Drew You can read messages from the DOTNET archive, unsubscribe from DOTNET, or subscribe to other DevelopMentor lists at http://discuss.develop.com. You can read messages from the DOTNET archive, unsubscribe from DOTNET, or subscribe to other DevelopMentor lists at http://discuss.develop.com.