Oracle gurus: Pls help me decipher this output.

[jlakhani]

what should i be looking out for and what component would i have to change?
Session Event Section
****************************************************
if average-wait > 0 then contention exists


EVENT TOTAL_WAITS TOTAL_TIMEOUTS AVERAGE_WAIT
------------------------------ ------------ -------------- ------------
latch free 4 2 .25
pmon timer 4,584 4,583 305.772469
rdbms ipc message 4,557 4,551 307.566381
rdbms ipc message 5,830 4,534 240.242196
rdbms ipc message 10 9 129669.9
rdbms ipc message 4,568 4,550 306.404772
enqueue 1 0 1
control file sequential read 30 0 .033333333
control file sequential read 57 0 .087719298
control file parallel write 13 0 .846153846
control file parallel write 4,550 0 .42043956

EVENT TOTAL_WAITS TOTAL_TIMEOUTS AVERAGE_WAIT
------------------------------ ------------ -------------- ------------
buffer busy waits 1 0 6
log file sequential read 8 0 .875
log file single write 10 0 .8
log file parallel write 1,309 0 .649350649
log file switch completion 1 0 8
log file sync 600 0 .713333333
log file sync 322 0 .757763975
db file sequential read 22 0 1.45454545
db file sequential read 1 0 1
db file sequential read 93 0 .903225806
db file sequential read 2 0 1.5

EVENT TOTAL_WAITS TOTAL_TIMEOUTS AVERAGE_WAIT
------------------------------ ------------ -------------- ------------
db file sequential read 234 0 .794871795
db file scattered read 38 0 3.39473684
db file scattered read 4 0 1.25
smon timer 49 45 28224.102
library cache pin 7 0 .857142857
library cache pin 3 0 11.3333333
library cache load lock 1 0 2
file identify 9 0 .333333333
file identify 6 0 1.5
SQL*Net message to client 56,766 0 .000510869
SQL*Net message to client 46,746 0 .000641766

EVENT TOTAL_WAITS TOTAL_TIMEOUTS AVERAGE_WAIT
------------------------------ ------------ -------------- ------------
SQL*Net message from client 56,765 0 24.6368185
SQL*Net message from client 11 0 .090909091
SQL*Net message from client 65 0 22.5692308
SQL*Net message from client 11 0 .090909091
SQL*Net message from client 149 0 926.38255
SQL*Net message from client 11 0 .090909091
SQL*Net message from client 11 0 .090909091
SQL*Net message from client 46,745 0 29.9330838
SQL*Net message from client 29 0 621.241379
SQL*Net message from client 11 0 .090909091
SQL*Net more data from client 127 0 4

EVENT TOTAL_WAITS TOTAL_TIMEOUTS AVERAGE_WAIT
------------------------------ ------------ -------------- ------------
SQL*Net more data from client 75 0 4.34666667

45 rows selected.


thanks
Jigar

[pando]

v$session_wait is used to diagnostic problems

so what is your problem? hanging sessions or what?

[JChiappa]

By no means I´m some type of guru, but I will say this :

a) normally is safe ignore common communication events like "sql*net xxx"

b) numbers per se mean NOTHING. See, it is like to say : "hey, I spend 1 gallon of gas going to home. Is it good or bad ?". The answer is : it DEPENDS ! How many miles you drove ?
There is something similar here, you MUST run the query, anotate the numbers, ** KEEP the db up and running **, and some time later (say, one day, one week) run again and compare : hey, in one day my xx event grows to N, and THEN we can say anything.

Regards,

Chiappa

[jlakhani]

PAndo
I have a script that gathers stats and I was trying to make sense out of it. Pls guide me in interpreting the output.

thanks
Jigar

[pando]

all those waits are described in documentation, check References section

[jlakhani]

REferences Pando? Are you refering to Oracle refernces" Performance tuning?

[pando]

......... I mean Oracle documentation, not Oracle books
tahiti.oracle.com