Re: anyone seen this weirdness with sequences in 11gr2?
- From: Andre van Winssen <dreveewee@xxxxxxxxx>
- To: TESTAJ3@xxxxxxxxxxxxxx
- Date: Tue, 13 Apr 2010 14:31:22 +0200
relax pal, what is stated in the ML note *1050193.1 *is how it has always
worked for oracle sequences. It is not "cover your ass", it is more like not
reading the documentation properly. I remember people relying on group by to
do the order by in a select statement, well you can't. same here. you can't
rely on all consecutive values when it comes to sequences.
2010/4/13 <TESTAJ3@xxxxxxxxxxxxxx>
>
> oh give me a break, thats oracle's way of doing the Cover your ass thing,
> aka CYA as one of the definitions of a sequence is:
>
> a continuous or connected series
>
> This happens because of a bug during the segment creation, oracle attempts
> to do the insert, it fails, has already sucked off a sequence, they trap the
> error, ORA-14403 and redo the operation thus eating the next sequence for
> the insert and the first value is now lost.
>
>
> 14403, 00000, "cursor invalidation detected after getting DML partition
> lock"
> // *Cause: cursor invalidation was detected after acquiring a partition
> lock
> // during an INSERT, UPDATE, DELETE statement. This error is
> never
> // returned to user, because is caught in opiexe() and the DML
> // statement is retried.
> // *Action: nothing to be done, error should never be returned to user
>
>
> joe
>
> _______________________________________
> Joe Testa, Oracle Certified Professional
> Senior Engineering & Administration Lead
> (Work) 614-677-1668
> (Cell) 614-312-6715
>
>
>
>
>
> From: Andre van Winssen <dreveewee@xxxxxxxxx> To: jkstill@xxxxxxxxx,
> oracle-l@xxxxxxxxxxxxx Cc: robertgfreeman@xxxxxxxxx, ssibert@xxxxxxxxx
> Date: 04/13/2010 08:15 AM Subject: Re: anyone seen this weirdness with
> sequences in 11gr2? Sent by: oracle-l-bounce@xxxxxxxxxxxxx
> ------------------------------
>
>
>
>
> "..you should not rely on continuity of the values from a
> sequence.Sequences are not guaranteed to generate all consecutive values
> starting with the 'START WITH' value.."
>
> 2010/4/13 Jared Still <*jkstill@xxxxxxxxx* <jkstill@xxxxxxxxx>>
>
> On Thu, Apr 8, 2010 at 3:22 PM, Robert Freeman
> <*robertgfreeman@xxxxxxxxx*<robertgfreeman@xxxxxxxxx>>
> wrote:
> You only see this when it's associated with the creation of a table.
> Metalink document 1050193.1 explains this behavior in 11.2.
>
>
> Interesting new normal behavior.
>
> Thanks Robert.
>
> Jared Still
> Certifiable Oracle DBA and Part Time Perl Evangelist
> Oracle Blog: *http://jkstill.blogspot.com* <http://jkstill.blogspot.com/>
> Home Page: *http://jaredstill.com* <http://jaredstill.com/>
>
>
>
>
Other related posts:
