Re: Zero rowcount ops in tkprof (only some queries!)

From: cam <kadmon_at_gmail.com>
Date: Sun, 29 Jun 2008 22:10:04 +0100
Message-ID: <2609b6ef0806291410w99025a5ne09d40b4b6517c21@mail.gmail.com>


Ooops - 10.2.0.4 on HP-UX 11.23, trace started with:

EXECUTE IMMEDIATE 'alter session set events ''10046 trace name context forever,level 12''';

in a logon trigger. tkprof generated with defaults and explain credentials.

c

On Sun, Jun 29, 2008 at 10:07 PM, cam <kadmon_at_gmail.com> wrote:

> Hello all,
>
> Can't quite understand this - I have just run a 26-hour level-12 10046
> trace on a slow process producing a 750M trc file. Imagine my disappointment
> to find that the key query I am interested in has no rowsource data - or
> only reports zeroes. Other queries have row totals - e.g.:
>
> Rows Row Source Operation
> ------- ---------------------------------------------------
> 3 SORT ORDER BY (cr=3 pr=0 pw=0 time=57 us)
> 3 TABLE ACCESS CLUSTER HISTGRM$ (cr=3 pr=0 pw=0 time=37 us)
> 1 INDEX UNIQUE SCAN I_OBJ#_INTCOL# (cr=2 pr=0 pw=0 time=23
> us)(object id 252)
>
> but for my huge query I just get an execution plan with zeroes in the Rows
> column:
>
> Rows Execution Plan
> ------- ---------------------------------------------------
> 0 INSERT STATEMENT MODE: ALL_ROWS
> 0 VIEW
> 0 SORT (ORDER BY)
> 0 HASH JOIN (ANTI)
> 0 NESTED LOOPS
> 0 NESTED LOOPS
> 0 HASH JOIN (OUTER)
> 0 HASH JOIN (RIGHT ANTI)
> .. etc.. etc
>
> I'd be interested in any ideas as to why this is. Also, am I correct in my
> assumption that the rowsource is arguably more valuable than the explain
> plan in that it shows what *actually* happened rather than an explain plan
> generated by tkprof 24 hours later - albeit likely a valid explain plan?
>
> Cheers for any insight,
> cam
>

--
http://www.freelists.org/webpage/oracle-l
Received on Sun Jun 29 2008 - 16:10:04 CDT

Original text of this message