![](//teradataforum.com/imgs/ptdfw1.gif) |
![](//teradataforum.com/imgs/ptdfw2.gif) |
Archives of the TeradataForum
Message Posted: Tue, 09 Mar 2010 @ 08:34:59 GMT
Subj: | | Re: Data Skew vs Single Amp |
|
From: | | Barrineau, Claybourne |
Barry and David... Thank you for your input. Your insight has been extremely helpful.
In response to Barry, we are primarily focusing on a single JI which pre-joins our Fact and 2 Dimensions. Time is 'as of yesterday' and our
fact table is rebuilt nightly from scratch. The JI currently has a multi-level PPI which is working wonderfully, though highly skewed. We are
testing the results out in PROD now....just trying to rationalize the pros and cons of a skewed, ultra-efficient (all single-amp access) JI versus
an evenly distributed, relatively efficient (~70% single amp access) JI.
I'll let you know what we come up with.
Thanks,
Clay
| |