RE: identify CTAS from v$ tables

In this case, dictionary managed tablespaces are being used, but the extents   are properly sized. CTAS and drop table are the only sql that need space management. I have suggested, to move towards delete/insert and use of global temporary table where possible.    

  Most of the scripts use CTAS and run once in 30 minutes or 1 hour.    

  Any suggestions are welcome.    

  Regards    

  Joseph

ramick <ramick_at_dotster.com> wrote:

        v\:* {behavior:url(#default#VML);}  o\:* {behavior:url(#default#VML);}  w\:* {behavior:url(#default#VML);}  .shape {behavior:url(#default#VML);}                Apologies if I missed something, but are you sure the ST enqueue locks are the result of CTAS?
   

  They could be anything that requires the allocation or deallocation of extents for any space management that is dictionary-managed.? Could be ?tablespaces of type temporary or permanent tablespaces.           

  From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Joseph Amalraj Sent: Tuesday, April 25, 2006 2:15 PM
To: Stephane Faroult
Cc: ganstadba_at_hotmail.com; oracle-l_at_freelists.org Subject: Re: identify CTAS from v$ tables    

    In this database, there are 'ST' enqueue locks and i have to identify the SQL statements issued by blocking and blocked sessions. I cannot use DDL triggers.      

    Thanks      

    Joseph

Stephane Faroult <sfaroult_at_roughsea.com> wrote:

    Joseph,

Would a DDL trigger be an acceptable solution to you? I haven't tried it recently, but I *think* that you can catch the statement text in the trigger. I am not fully certain about that, because since DDL statements are nothing else (ultimately) than DML statements applied to dictionary tables some are broken into their constituent DML components so fast that they let no track in the SGA. Otherwise perhaps that V$object_dependency could help too ...

HTH St鰨ane Faroult

Joseph Amalraj wrote:

> After going thru' the v$fixed_view_definition for v$sql, v$open_cursor
> and
> v$sqltext, I could only get the name of the table being created by the
> following query.
>
> sys_at_clfypurg > select kglnaobj from x$kgllk where
> kgllkmod = 3;
> KGLNAOBJ
> ------------------------------------------------------------
> MY_LARGE_TABLE
>
> if "alter session set sql_trace = true" is run before executing CTAS
> the trace
> does generate the CTAS sql statement.
>
> Can the sql be found from any x$ table ??
>
> Thanks
>
> Joseph
>
> */Michael McMullen /* wrote:
>
> Here's what TOAD uses but as I recall you can't get the underlying
> DDL statement for CTAS. Maybe changed in 10g.
> SELECT sql_text
> FROM v$sqltext_with_newlines
> WHERE hash_value = TO_NUMBER (:HASH)
> ORDER BY piece
>
>
  

--
http://www.freelists.org/webpage/oracle-l