Archives of the TeradataForum

Message Posted: Fri, 25 Oct 2002 @ 18:01:20 GMT

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

We use "canary queries," too. However, that is not effective nor, necessarily, very accurate. We'd like to see the DBS/OS report something like rolling averages of the queue lengths for various things (disk, cpu), so that we could see the relative demand for resources.

As soon as the users get warmed up in the morning, the box goes to 100% utilization on all AMPs but there's a vast difference in performance between one query driving all AMPs to 100% utilization (and having that 100% all to its lonesome) and 25 queries driving all AMPs to 100% (with each query getting 4%)

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