Re: Keep (Recycle) Buffer Question

The tool Odat is has a function specially designed for this purpose. The function is called "database report". By click it, it will automatically explain plan all the sql statements in your v$sqlarea, after that, it will give you a report, which tells your, which table has been through full table scans, how big is the table, etc, etc.

check it out at
http://members.bellatlantic.net/~vze2n2ja/

"Peter Bruhn" <lbruhn_at_miss.wu-wien.ac.at> wrote in message news:al1ofe$mku$1_at_bird.wu-wien.ac.at...
> Hi,
>
> I am not a DBA, but I have followed the long thread of the
 cach/keep-buffer
> discussion a bit. We have performance problems in our DB all the time. So
 I
> checked, if we use different buffers for different DB-objects. We do not.
 What
> I have understood from the above discussion that it is useful to put small
> lookup tables (+indexes) and some large tables (+indexes) that are heavily
> used into a seperate buffer that is large enough to hold all these
 objects.
> This buffer can be called keep-buffer. Then you put large tables (+indexes
> [?]) that are used very seldom into a third rather small buffer, call it
 the
> recycle buffer.
>
> Now we have a large database with many, many tables. How do I identify
 what
> tables to put in what buffer. Do I have to make these decisions "by hand".
 Or
> is there any tool (query) to help me make this decision? Can I find out,
 what
> tables often whipe out the current cache and are not reused, so I can put
 them
> to the recycle cache? Can I find out what tables are whiped out often and
 then
> reloaded soon, so I can put into the keep cache?
>
> Thanks for any hints, comments, flames,
> Peter
Received on Tue Sep 03 2002 - 16:15:28 CDT