"warm starting" vs non-"warm starting"?

Peter Teoh

unread,

Feb 13, 2008, 9:50:28 PM2/13/08

to

I created an index on obj$ today:

SQL> create index ooee on obj$(name, obj#);
create index ooee on obj$(name, obj#)
*
ERROR at line 1:
ORA-00701: object necessary for warmstarting database cannot be
altered

Don't quite understand this, and did some other things, and after some
time, I discovered this - if u persistently apply the same SQL:

SQL> create index ooee on obj$(name, obj#);
create index ooee on obj$(name, obj#)
*
ERROR at line 1:
ORA-00701: object necessary for warmstarting database cannot be
altered

SQL> create index ooee on obj$(name, obj#);
create index ooee on obj$(name, obj#)
*
ERROR at line 1:
ORA-00604: error occurred at recursive SQL level 1
ORA-00060: deadlock detected while waiting for resource

First there must be a difference between warmstarting and non-
warmstarting. What is it? (what is the difference between all the
YYY$ tables vs any other tables - be it in SYS schema or not? why is
it forbidden to modify these tables vs those non-YYY$ tables - there
must be a reason? this is because the database is already opened
anyway, so, is it a precautionary measure, or is there a physical
reason not to do this? if it is a physical reason, there there must
be some other error, which is not likely. therefore, if this is a
precautionary measure, the list of warm starting tables that is
advised not to be modified must be stored somewhere, and if we can
modify these, the error will go away, right?)

Why the difference in error message - there must be something
executing ASYNCHRONOUSLY IN THE BACKGROUND?

Please enlighten me :-).

phil_h...@yahoo.com.au

unread,

Feb 14, 2008, 12:35:41 AM2/14/08

to

Um, is that when you type 'ior warm'?

Now I feel old...

-- Phil

Mladen Gogala

unread,

Feb 14, 2008, 1:59:34 AM2/14/08

to

On Wed, 13 Feb 2008 21:35:41 -0800, phil_h...@yahoo.com.au wrote:

> Um, is that when you type 'ior warm'?

If my memory serves me right, he should re-initialize the database
with "ior init" and watch for the "before image" file (VAX$BI).

--
Mladen Gogala
http://mgogala.freehostia.com

Mladen Gogala

unread,

Feb 14, 2008, 2:00:04 AM2/14/08

to

On Wed, 13 Feb 2008 21:35:41 -0800, phil_h...@yahoo.com.au wrote:

> Um, is that when you type 'ior warm'?

If my memory serves me right, he should re-initialize the database

sybrandb

unread,

Feb 14, 2008, 4:39:35 AM2/14/08

to

Peter,

Can you explain why you tamper with the datadictionary, consequently
invalidating your license?

YOU SHOULD NEVER EVER MODIFY SYS UNLESS ON INSTRUCTION OF ORACLE!!!

--
Sybrand Bakker
Senior Oracle DBA

gazzag

unread,

Feb 14, 2008, 6:18:58 AM2/14/08

to

On 14 Feb, 02:50, Peter Teoh <htmldevelo...@gmail.com> wrote:
> I created an index on obj$ today:
>
> SQL> create index ooee on obj$(name, obj#);
> create index ooee on obj$(name, obj#)

[snip]

> Why the difference in error message - there must be something
> executing ASYNCHRONOUSLY IN THE BACKGROUND?
>
> Please enlighten me :-).

One of the crossbeams is out of skew on the treadle.

You've shown us _what_ you've done, now can you enlighten us as to
_why_? :)

-g

Frank van Bortel

unread,

Feb 18, 2008, 3:10:10 PM2/18/08

to

ior i/w (did c exists, or was oir i "cold"?)
Check out sql.bsq for ior :)
--

Regards,
Frank van Bortel

Top-posting in UseNet newsgroups is one way to shut me up