From: Mark R. <ma...@la...> - 2013-11-26 19:40:17
|
On 26-11-2013 19:28, Mark Rotteveel wrote: > The problem seems to be with connection pooling not closing the command > when the logical connection is closed (and the physical connection > returned to the pool). I'd argue that when connection pooling is used, a > close of the logical connection should close any open dependent objects > and only leave the connection itself open. Never mind my comment re connection pooling, I missed that Kjell asked why an explicit close and open of the connection fixed the issue. Mark -- Mark Rotteveel |