Archives of the TeradataForum

Message Posted: Thu, 25 May 2006 @ 13:17:56 GMT

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

Apologize that my previous mail doesn't convey my question properly.

In the absence of statistics, the table size estimate made by the optimizer is based on Dynamic AMP Sampling. The script output below shows that the estimate made by dynamic AMP sampling was incorrect in one of the Explain outputs.

Observe the following in the script output

a. The estimate made on the size of the table the first time in the absence of statistics was very bad even though the distribution across the AMPs was even. Also observe that there is No Target Level Emulation data that may have influenced the erroneous estimate.

Note: Statistics were never collected on this table.

b. The estimate is better when statistics are collected. The estimate continues to be better after the statistics are dropped. Dynamic AMP Sampling sampling seems to work after statistics were collected and dropped; but not before. Why? Is this a known bug that's fixed in a newer release?

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