Archives of the TeradataForum

Message Posted: Sat, 28 Jun 2003 @ 19:06:41 GMT

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

Pat,

Your reply helps. I will pass this condition on to our consultant.

I agree. Something is fishy. Too many 3rd parties involved in the story, which includes me.

It is understood that the data is striped across the AMPs to achieve parallelism. Since it is a shared nothing environment, I focused the question around what happens in an AMP. My apologies for not being clear here.

A thought came to mind regarding fragmentation when reading the replies received thus far. It seems quite probable that the more tables there are, the more likely a table could become fragmented as it grows. By this, I mean the data blocks would not be contiguous. Therefore, the data could be spread across more disks, causing more head movement, and thus more seek time.

Comments?

Thanks to everybody for your insights on this.

Charlie

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