Re: AW: Query Optimizer selection incorrect access path

[CRPence <CRPbottle@xxxxxxxxx>]

On 05 Nov 2012 23:10, Birgitta Hauser wrote:
> It [<ed:> the query] simply uses the access path of the logical
> file, but not the logical file itself, i.e. it ignores the
> select/omit clause.

  That a query could use the AccPth while ignoring the Slt\Omt 
specifications can be true *only* for a DYNSLT LF.  Without a Dynamic 
Select specification in the DDS, a Keyed Access Path actually includes 
or omits the rows with the Access Path itself, so there would be no 
ability for the query to ignore those specifications when implementing 
with that AccPth.  Whether that has any relevance for the OP scenario is 
unknown by me; unstated by the OP.

> <<SNIP>>
>
> As an aside an SQL index has (per default) a page size of 64 K (while
> the default page size for a DDS described logical file is 8K), i.e.
> performance may be better.

  The PagSiz for a keyed DDS PF or LF is calculated based on the size 
of the key; i.e. there is no one specific default.  A parameter kwd was 
added to the CRTLF and CRTPF to enable specification of the Page Size; 
the default is *CALC.  IIRC a "short" key actually uses a 4K page size, 
and smaller page sizes can be better for non-query [RLA] activity and 
effects on journal; i.e. for effective memory footprint.