Hi, I just ran:
explain select max(changelog_id) from changelog 

changelog_id is an int and the primary key of changelog, which has
184551 rows.

And the gist of the explaination is:
TABLE SCAN                         828
RESULT IS COPIED , COSTVALUE IS   1657

It seems to me that it ought to be cheaper to walk down the index and
find the max, in fact I should think the cost would be in the 'hood of a
simple match.

Is this a missing feature, a bug, something that is caused by my data or
what? 

BTW, select count(*) and select count(changelog_id) on that table has
the same strategy / costvalue

Investigating a bit further I have found out that the SYSPRIMARYKEYINDEX
isn't used, how can that happen for a table that has 184551 rows?

-- 
 Regards Flemming Frandsen aka. Dion/Swamp http://dion.swamp.dk
_______________________________________________
sapdb.general mailing list
[EMAIL PROTECTED]
http://listserv.sap.com/mailman/listinfo/sapdb.general

Reply via email to