Add sql statements in SGA

Question

select a.address address,
s.hash_value hash_value,
s.piece piece,
s.sql_text sql_text,
u.username parsing_user_id,
c.username parsing_schema_id
from v$sqlarea a,
v$sqltext_with_newlines s,
dba_users u,
dba_users c
where a.address=s.address
and a.hash_value=s.hash_value
and a.parsing_user_id=u.user_id
and a.parsing_schema_id=c.user_id
and exists (select 'x'
from v$sqltext_with_newlines x
where x.address=a.address
and x.hash_value=a.hash_value
and upper(x.sql_text) like '%UNION%')
order by 1,2,3
On executing the above statement the server display list of sqltext present in sga. We know that the server first search for the record in sga and if it is not
present in SGA it searches in data files.
My question is ,adding all the sql statements in sga manually is advisable? will that improve the performance?

Answer 1

Vinodh2 wrote:
In production the users uses some sql statement frequently. So those statements should be in the sga for better performance. The dbms package said by you takes object as input. What is object here?
Did you check the last line I pasted from document?
The value for this identifier is the concatenation of the address and hash_value columns from the v$sqlarea view.Each SQL in SQL AREA has a hash_value to identify it. You can use this hash value to decide which SQL want to keep in shared pool.
Having Dynamic sql statements(records fetched based on varying filter condtion) called from interfaces will not help.
But having static sql statements helps.
In that case how to include those sql.Like said, you are introducing some more problems instead of any real gain by trying to micro manage how shared pool work. You should let Oracle take care it with it's own algorithm. Unless some rare cases, that you have limited shared pool and super busy system you want to pin some SQL in the shared pool with provide package.