Archives of the TeradataForum

Message Posted: Fri, 04 Mar 2005 @ 20:50:47 GMT

	<Prev	Next>		<<First	<Prev	Next>	Last>>

Well...

We certainly do abort the later one, but we don't wait 4 minutes before we abort. We abort as soon as the global deadlock is detected. The 4 minutes is the interval at which we kick off global deadlock detection. On the average, you ought to be able to count of global deadlocks being broken in around 2 minutes.

I know, I know... I'm picking at nits. 8)

	<Prev	Next>		<<First	<Prev	Next>	Last>>