Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Mailing Lists -> Oracle-L -> RE: PCTINCREASE and the SYSTEM tablespace
Well I'm awaiting the day when it's kosher to have a locally managed system
tablespace. Until then I philosophize as follows... I agree with the
ideas/techniques covered Steve Adam's site. Been doing the custom sql.bsq
thing for years now with no ill effect. I believe in a pctincrease of either
0 or 100 but nothing in between. Even though you can custom size many system
objects based on past/present implementations something new eventually comes
along and has different requirements. That's where pctincrease of 100 comes
in. I'd rather "waste" a few MB than have inconsistent extent sizes. I
believe in powers of 2!!
IMHO,
Steve Orr
-----Original Message-----
From: root_at_fatcity.com [mailto:root_at_fatcity.com]On Behalf Of Allan
Robertson
Sent: Tuesday, September 12, 2000 11:20 AM
To: Multiple recipients of list ORACLE-L
Subject: PCTINCREASE and the SYSTEM tablespace
I have just started in my new place of work and we were discussing the pctincrease/smon/coalesce issue for tablespaces. The new place prefers to have pctincrease=0 and coalesce via a cron cobtrooled script. Fair enough ...
We then came to discuss the pctincrease value for the system TS. I have always left it at 50% for new sys object creation. At my new place they build the data dict/apps etc then set it to 0%.
I am getting mixed input from books, web sites and metalink on setting the value of pctincrease on the system TS, (including Steve Adams site & his custonizing sql.bsq doc). This is probably the wrong question to ask of a group of DBAs but ...... does anyone have any view on the subject?
Thanks
Allan
Switch : 0131 338 6060 Fax : 0131 338 6061 Email : allan.robertson_at_morse.com
Orders are accepted subject to Morse's January 2000 terms and conditions. Copies are available at www.morse.com <http://www.morse.com> and on request.
![]() |
![]() |