Hello, ls!
You wrote on 19 Jan 2002 03:19:02 -0800:
lc> Hi
lc> If we face ORA-4031 will
lc> alter system flush shared_pool
lc> defragment (coalesce) thee free chunks?
some memory will be freed however... doubt about coalescing.
lc> I understand that in Oracle 7 this command was useless to coalesce
lc> or defragment the shared pool (at least this is what Metalink's Note
lc> suggests). However in Oracle 8 and 8i does it still apply or it's
lc> improved and we can fix ORA-4031 (temporary solution) by flush the
lc> shared pool manually OR does Oracle defragment shared pool
lc> internally?
afaik - no, oracle don't defragment shared pool internally by itself.
anyway - flush not releases _all_ memory in shared pool, some objects
(kept or are being executed now) could remain;
see dox about cursor_space_for_time parameter...
lc> Also is there any X$ view that shows the free chunks remaining with
lc> their respective sizes? (I will try with a small shared pool of
lc> course! If such x$ exists)
visit
http://www.ixora.com.au/home.htm
Your comments, pls...
- Best regards, Igor Ushkalo (igorus!) - igorus(at)mail.ru, ICQ #19972198
Received on Sat Jan 19 2002 - 06:46:19 CST
