Hi Folks..
I traced a session for a timely transaction and found about 30 of these :
SELECT ROWID
FROM
NominalBudgetNames
call count cpu elapsed disk query current rows
------- ------ -------- ---------- ---------- ---------- ---------- ----------
Parse 2 0.04 0.15 2 0 2 0
Execute 2 0.00 0.00 0 0 0 0
Fetch 2 0.01 0.02 2 2 8 2
------- ------ -------- ---------- ---------- ---------- ---------- ----------
total 6 0.05 0.17 4 2 10 2
Misses in library cache during parse: 2
Optimizer goal: CHOOSE
Parsing user id: 25 (MAGIC8)
Rows Row Source Operation
------- ---------------------------------------------------
1 TABLE ACCESS FULL NOMINALBUDGETNAMES
Rows Execution Plan
------- ---------------------------------------------------
0 SELECT STATEMENT GOAL: CHOOSE
1 TABLE ACCESS (FULL) OF 'NOMINALBUDGETNAMES'
********************************************************************************
You will notice that although the table is accessed via a full table scan, the cost of the operation is mininmal. Is it safe to say that this type of full table scan can be ignored as a peformance issue.
The queries which were costly were those which had many buffer/disk blocks queried but few rows fetched.
Thanks
Suresh
Once you have eliminated all of the impossible,
whatever remains however improbable,
must be true.