As I mentioned in a previous blog, I learnt something new from Anjo Kolk’s SAN presentation at the Oak Table day at the UKOUG conference. Instead of looking at the maximum read time for a file since the instance started or the average in a statspack report, you can reset the maximums.
SQL> select * from v$version;
BANNER
----------------------------------------------------------------
Oracle Database 10g Enterprise Edition Release 10.2.0.1.0 - Prod
PL/SQL Release 10.2.0.1.0 - Production
CORE 10.2.0.1.0 Production
TNS for 32-bit Windows: Version 10.2.0.1.0 - Production
NLSRTL Version 10.2.0.1.0 - Production
SQL> select file#, maxiortm from v$filestat;
FILE# MAXIORTM
---------- ----------
1 82
2 71
3 62
4 61
5 9
(Do some work in another session)
SQL> /
FILE# MAXIORTM
---------- ----------
1 84
2 71
3 62
4 61
5 23
(Now I want to reset the max timing, which is the maximum since the instance started, so that I can check it again after running a specific workload)
SQL> exec dbms_system.kcfrms();
PL/SQL procedure successfully completed.
SQL> /
FILE# MAXIORTM
---------- ----------
1 0
2 0
3 0
4 0
5 0
That’s just one of a few things I learnt from the two presentations by Anjo that I could attend. Thanks!
This also resets the maximum times on v$session_event, v$system_event and (for 10g) v$service_event. (Not that you can see the maximum on v$system_event unless you redefine the view).
See also:
http://www.jlcomp.demon.co.uk/log_file_writes.html
and
http://www.jlcomp.demon.co.uk/system_i.html
Jonathan Lewis
Thanks Jonathan. Anjo only mentioned it in the context of i/o investigations, but it’s good to know it applies in other areas too.
I was vaguely aware of this but probably depend too much on statspack and tracing and don’t make enough use of the V$ views. I need to spend more time looking at dbms_system.
The hints presentation was terrific, by the way, so that’s a few things I’ve learnt from you this week. I’m still thinking about that RSS feed!
Cheers,
Doug