Extensive IO and CPU for simple queries

Question

Hi,
I have a machine using oracle 9.2.0 and solaris 10.
For every simple queries, it very big IO and a lot of CPU. This is only happend on one particular machine(We have same version db and soalris on other mahcine and it works fine).
One example queris is when I use Enterprise Manager to get the "configuration" information of the instance,it use 50% IO. I get the trace file and tkprof as following:
SELECT UNIQUE sp.name, sp.sid, DECODE(p.type, 1, 'Boolean', 2, 'String', 3,'Integer', 4, 'Filename', ' '), sp.value, p.issys_modifiable, p.description FROM v$spparameter sp, v$parameter p WHERE sp.name = p.name ORDER BY sp.name,sp.sid
call count cpu elapsed disk query current rows
Parse 4 0.02 0.01 0 0 0 0
Execute 4 0.00 0.00 0 0 0 0
Fetch 9 4.36 34.12 7980 0 0 783
total 17 4.38 34.13 7980 0 0 783
Misses in library cache during parse: 1
Optimizer mode: ALL_ROWS
Parsing user id: 5 (SYSTEM)
Rows Row Source Operation
261 SORT UNIQUE (cr=0 pr=0 pw=0 time=1214116 us)
261 HASH JOIN (cr=0 pr=0 pw=0 time=1221296 us)
361485 MERGE JOIN CARTESIAN (cr=0 pr=0 pw=0 time=370609 us)
261 FIXED TABLE FULL X$KSPSPFILE (cr=0 pr=0 pw=0 time=19777 us)
361485 BUFFER SORT (cr=0 pr=0 pw=0 time=6413 us)
1385 FIXED TABLE FULL X$KSPPCV (cr=0 pr=0 pw=0 time=4180 us)
1379 FIXED TABLE FULL X$KSPPI (cr=0 pr=0 pw=0 time=7001 us)
It seems Oracle FTS the X$KSPPCV and X$KSPPI.
Can anybody give me some suggestion to improve the performance?
thanks.

Answer 1

Is there a difference in the query plans on the two machines?
Did you analyze the SYS and SYSTEM schemas on one system and not the other?
Are there different initialization parameters on the two machines?
What do you mean by "it use 50% IO"? I'm not sure what that means and I'm not sure how you're measuring that.
Justin
Distributed Database Consulting, Inc.
http://www.ddbcinc.com/askDDBC