Comments on: Advanced Oracle Troubleshooting Guide, Part 5: Sampling V$ stuff with WaitProf. Really fast. Using SQL!

By: Venu B

Venu B — Wed, 23 Nov 2011 01:46:21 +0000

Tanel,

I get the following error message when I run it on Solaris. Any idea? It works fine on Linux or AIX.

SQL> @snapper ash 5 1 ALL
Sampling SID ALL with interval 5 seconds, taking 1 snapshots…

– Session Snapper v3.13 by Tanel Poder @ E2SN ( http://tech.e2sn.com )

———————————————————————–
Active% | SQL_ID | EVENT | WAIT_CLASS
———————————————————————–
declare
*
ERROR at line 1:
ORA-00979: not a GROUP BY expression
ORA-06512: at line 831
ORA-06512: at line 1364

Thanks,
Venu

By: Mehul S.

Mehul S. — Mon, 02 Aug 2010 14:40:40 +0000

Coolest use of ‘connect by’
Simplicity is the ultimate elegance!!!

Cheers
Mehul S.

By: Tanel Poder

Tanel Poder — Fri, 21 May 2010 07:03:12 +0000

@Lasse Jenssen Same here! ;)

By: Lasse Jenssen

Lasse Jenssen — Fri, 21 May 2010 06:15:03 +0000

You are right! I’m on 9i. Should have testet an earlier version, but I was confident I did run the v3 against 9i earlier. I was obviously wrong. Fantastic script by the way. After starting using it, I keep wondering how I managed before ;-)

By: Tanel Poder

Tanel Poder — Thu, 20 May 2010 10:07:55 +0000

@Lasse Jenssen

Are you running it on 9i? Snapper v3 doesn’t run on 9i yet, once I have the time I’ll change it… so you need to use snapper_v2 on 9i…

By: Lasse Jenssen

Lasse Jenssen — Thu, 20 May 2010 07:42:59 +0000

Get an error with the latest version (3.11) of snapper.sql:

sys@FD> @snapper out 1 1 3
Sampling with interval 1 seconds, 1 times…
type snrec is record (stype varchar2(4), statistic# number, name varchar2(64));
*
ERROR at line 30:
ORA-06550: line 476, column 42:
PLS-00302: component ‘STATE’ must be declared
ORA-06550: line 475, column 11:
PL/SQL: Statement ignored

Have anybody else gotten this error?

By: Advanced Oracle Troubleshooting Guide, Part 7: Sampling latch holder statistics using LatchProf | Tanel Poder's blog: Core IT for Geeks and Pros

Thu, 01 Oct 2009 05:51:45 +0000

[...] is a script similar to WaitProf, only it samples latch holder statistics from V$LATCHHOLDER. As V$LATCHHOLDER contains a SID column [...]

By: Tanel Poder

Tanel Poder — Sat, 22 Aug 2009 07:38:32 +0000

The object name can be deducted from the file# and block_id (parameter1,parameter2) in dba_extents.

I use a script like this, it takes file# and block# as parameters:

select owner, segment_name, partition_name, tablespace_name, extent_id
from dba_extents
where file_id = &1
and &2 between block_id and block_id + blocks – 1;

By: Tanel Poder

Tanel Poder — Sat, 22 Aug 2009 07:37:07 +0000

High CPU usage is good in some sense – meaning that the session can do its work on CPU without having to wait for anything much.

Of course if the job/session is still slow then I would run snapper on the session and see whether its perhaps hard parsing a lot or doing lots of logical IOs. In first case the application code must be changed (or cursor_sharing=force parameter can be used as a dirty workaround), in second case the heavy logical IOs may be due an inappropriate nested loop joins looping through the inner rowsources way too many times (revisiting blocks a lot etc)

Of course there are many more reasons for high CPU usage but these two are the first common ones that come into my mind.

By: Maxx

Maxx — Wed, 19 Aug 2009 16:32:49 +0000

This is a great script. Needed something like this, as we have very tight change control and getting a trace approve take like 2 days and untill then we are expected to see what the waits are and stuff like that. Couple of suggestions (maybe)… Yes the script looks at p1, p2, p3 values and point to the datafile#, and block# which can then tell us what table its waiting on. It would be nice to someway have the script where it gives you p1,p2,p3 value and also tell you the object name. Just a thought.

Also you mentioned. “Even though the CPU usage is in the top of the list, it has only used 21.63% of total response time, so the rest 78% of the IO time is the “problem” here.” — what are some suggestions if its the other way around ?? Lets say i have 78% of CPU usage..what can I do to get that down?? any suggestions ?