Oracle FAQ | Your Portal to the Oracle Knowledge Grid |
![]() |
![]() |
Home -> Community -> Mailing Lists -> Oracle-L -> RE: outline not working on 1 instance of RAC cluster
Ok, I did the oradebug, and low and behold, the values were 0s. Turns out the
instance was bounced yesterday.
So, this implies that enabled outlines don't survive an instance bounce. On the surface, this makes sense, since it is an "alter system" command that enables it, but apparently, you can't set this in the spfile.
SQL> ALTER SYSTEM SET USE_STORED_OUTLINES = STEVE_OUTL scope=spfile sid='*'; ALTER SYSTEM SET USE_STORED_OUTLINES = STEVE_OUTL scope=spfile sid='*'
*ERROR at line 1:
So, how do you make an outline persistent? Do you have to issue the alter system in the database startup trigger?
And while we're talking about it, there's got to be an easier way to see whether an outline is active?
Wayne
On Fri Apr 13 17:02 , "Pal, Raj" sent:
>RE: outline not working on 1 instance of RAC cluster
>
>
>To check for sure, try this:
>
>SQL> oradebug setmypid
>
>SQL> oradebug dumpvar sga sgauso 100
>
> -If 2nd & 3rd set of (8) numbers are != 0, then use_stored_outlines=true.
>
> (Eg. qolprm sgauso_ [700000010020F58, 700000010020F7C) = 00000001
00074445 4641554C 54000000 00000000 00000000 00000000 00000000 00000000)
>
> -If both 2nd & 3rd sets are all 0s… use_stored_outlines=false.
>
> (Eg. qolprm sgauso_ [700000010020F58, 700000010020F7C) = 00000000
00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000)
>
>But the best way is just to "alter system set use_stored_outlines = true;"
anyway. Altering the system again if you are expecting it to be on should not
cause an issue.
>
>Raj.
>
>
>
>-----Original Message-----
>
>From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On
Behalf Of Wayne Adams
>
>Sent: Friday, April 13, 2007 4:27 PM
>
>To: oracle-l_at_freelists.org; Tony Aponte
>
>Subject: RE: outline not working on 1 instance of RAC cluster
>
>Yes (as far as I know). I was thinking about that when I was writing up the
>
>problem, but I couldn't figure out how to confirm. I couldn't find a parameter
>
>or V$ view that would tell me for sure.
>
>Wayne
-- http://www.freelists.org/webpage/oracle-lReceived on Fri Apr 13 2007 - 18:33:10 CDT
![]() |
![]() |