Hash Based Column Sampling In 11g As Compared To RowBased...

Hi,
In 11g (11106), here is what is mentioned about the stats gathering,
http://download.oracle.com/docs/cd/B28359_01/server.111/b28274/stats.h...PFGRF30
In the 11g NF guide from OU, it is mentioned that prior to 11g , the
column sampling was done with the help of "Row Based Sampling". In
11g, if one uses, ESTIMATE_PERCENT=>AUTO_SAMPLE_SIZE than the sampling
technique is Hash Based. This is by default there from 11g onwards.
Here are my questions,
1)What is the older Row Based Sampling and how it works?
2) What is the new Hash Based Sampling and how it works?
3) What's the benefit of the new technique?
Thanks in advance.
Aman....

On Dec 19, 2:49 pm, Aman Sharma wrote:
> Hi,
> In 11g (11106), here is what is mentioned about the stats gathering,http://download.oracle.com/docs/cd/B28359_01/server.111/b28274/stats....
> In the 11g NF guide from OU, it is mentioned that prior to 11g , the
> column sampling was done with the help of "Row Based Sampling". In
> 11g, if one uses, ESTIMATE_PERCENT=>AUTO_SAMPLE_SIZE than the sampling
> technique is Hash Based. This is by default there from 11g onwards.
> Here are my questions,
> 1)What is the older Row Based Sampling and how it works?
> 2) What is the new Hash Based Sampling and how it works?
> 3) What's the benefit of the new technique?
> Thanks in advance.
> Aman....

This Oracle white paper might give a clue:
http://www.oracle.com/technology/products/bi/db/11g/pdf/twp_bidw_dss_p..._11gr1.

Regards,
Vladimir M. Zakharychev
N-Networks, makers of Dynamic PSP(tm)
http://www.dynamicpsp.com