From time to time I happen to perform procedures which require a lot of time (10+ minutes sometimes). The actual problem is that if the connection happens to drop during those 10 minutes since there is no actual transfer (it is a really, really long stored procedure which might have to go through 20-30k rows and seriously rearrange them and might have to "wait" quite often in the process) the tcp stack does not throw an exception. I could edit the tcp timeout of my workstations and reduce it, but it might not always be possible (or desirable on slower connections).
I have currently added a 'watchdog' thread which does some really nasty stuff every X minutes (kills the thread and creates an entirely new connection, such long-lasting commands are possible, but not expected to happen often). Is there a proper way to do that in the .net provider (or firebird at all) ? Any other suggestions are welcome, thanks in advance. -- Sanity is a sin! ------------------------------------------------------------------------------ _______________________________________________ Firebird-net-provider mailing list Firebird-net-provider@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/firebird-net-provider
