Re: very impressed by 11g

"Gerard H. Pille" <ghp_at_skynet.be> wrote in message news:4f9703d1$0$3116$ba620e4c_at_news.skynet.be...
| Months ago we upgraded our main database to 11g, and after tuning a
number of selects, we were
| running happily along. A couple of weeks ago, we noticed we'd forgotten
to change the
| compatible parameter to 11.2 or something, so, yesterday we set that
straight.
|
| And brought that system to it's knees.
|
| There was nothing to be done about it, we didn't find any specific
queries that were
| responsible, but the system kept hitting 100% cpu.
|
| Finally, we doubled the number of processors, and that solved it. Now
and then, the system
| still reaches 50%, but mostly it stays way below it.
|
| That parameter can not be set back, it appears, so we were very lucky we
had spare processing
| power, or we would still be struggling.

In principle the compatible parameter is about physical data structures, not about features. So you wouldn't expect (for example) changes in optimizer plans from changing this parameter. The sort of thing you might expect is to see some extra work being done because you have data stored in one physical format that has to be converted in real time to a different physical format. (I can't think of any real examples of this kind of need, though, that would explain your observations.)

-- 
Regards

Jonathan Lewis
http://jonathanlewis.wordpress.com
Oracle Core (Apress 2011)
http://www.apress.com/9781430239543