Re: stored outlines used but wrong plan executed

• Post for FREE via your newsreader at post.usenet.com ****

"Jonathan Lewis" <jonathan_at_jlcomp.demon.co.uk> wrote in message news:cijr2d$o32$1_at_hercules.btinternet.com... [...]
> I would be very surprised if there were no differences
> between the test and production systems that made
> the hints from imported outline illegal on arrival.

Of course there must be a difference! But in this case there was no evident difference (at least for me):

- no difference in documented and un-documented parameters
- same objects (same names as well)
- and, of course, same database version

The only difference I could recognize was that the two databases had a different "live" and slightly different statistics (unfortunately I was not able to export/import the statistics to find out if this was the problem). For example they were patched differently (I don't remember exactly, but e.g. prod: 9.2.0.1.0 -> 9.2.0.4.0, test: 9.2.0.1.0 -> 9.2.0.3.0 -> 9.2.0.4.0) and that on test the objects where dropped/created in a different order and/or more times in test (AFAIK the CBO doesn't use the timestamp of the objects like the RBO does to make a choice in case of equivalency, I'm wrong?). Therefore I concluded that some non-evident difference in the data dictionary lead to a non-used outline.

Chris

-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

• Usenet.com - The #1 Usenet Newsgroup Service on The Planet! *** http://www.usenet.com Unlimited Download - 19 Seperate Servers - 90,000 groups - Uncensored -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=