tuning through tkprof stats
555329Dec 16 2008 — edited Dec 18 2008
Hello,
Working on oracle 10G R2/Windows, I have the following Script, and the tkprof output gotten.
SQL Script*:
SELECT t1.ALL_CUSTOMER_ID
FROM DEV.ALL_CUSTOMER t1
ORDER BY :"SYS_B_0"
*_{color:#ff0000}Note:{color}_* the table all_customer contains more than 490000 rows, and seeing the stats below, the number of rows compared to the number of query seems pretty good.
all_cutomer_id is the primary key and all_customer_i1 is the index based on the primary key.
the index clustering factor is not so far away from the blocks# of the all_customer table, so there is not to worry about, I think.
the only hints is the event SQLNet message from client consumming the whole amount of times.* *{color:#ff0000}Question:{color} How Could I diminish get rid of this or diminish it impact on the whole SQL response time?*
call count cpu elapsed disk query current rows
------- ------ -------- ---------- ---------- ---------- ---------- ----------
Parse 2 0.00 0.00 0 0 0 0
Execute 2 0.00 0.00 0 0 0 0
*{color:#ff0000}Fetch 7692 1.23 1.32 0 9910 0 984153{color}*
------- ------ -------- ---------- ---------- ---------- ---------- ----------
total 7696 1.23 1.32 0 9910 0 984153
Misses in library cache during parse: 0
Optimizer mode: FIRST_ROWS
Parsing user id: 58 (SPA5000)
Rows Row Source Operation
------- ---------------------------------------------------
492077 INDEX FULL SCAN ALL_CUSTOMER_I1 (cr=4955 pr=0 pw=0 time=2092 us)(object id 102366)
Rows Execution Plan
------- ---------------------------------------------------
0 SELECT STATEMENT MODE: FIRST_ROWS
492077 INDEX MODE: ANALYZED (FULL SCAN) OF 'ALL_CUSTOMER_I1' (INDEX)
Elapsed times include waiting on following events:
Event waited on Times Max. Wait Total Waited
---------------------------------------- Waited ---------- ------------
SQL*Net message to client 7692 0.00 0.00
{color:#ff00ff}*{color:#ff0000} SQL*Net message from client 7692 1.07 9.56{color}*
*{color:#000000}Thanks for your help{color}*
{color}