Hi Friends,
I have 150Gb database with 30 rollback segs with optimal settings!! for normal operations no problems!! But we are doing massive updates for some of big tables!! So we are using vb programs and connecting to the database thru odbc dsn. My first question is how to set transaction use
Ave !
Is there any reason not to
keep maxextents always unlimited :) Extentsize should
be reasonable of
course.
Br.
Jorma
-Original Message-From: ext venkat Rama
[mailto:[EMAIL PROTECTED]Sent: 16 July, 2003 07:34To:
Multiple recipients of list ORACLE-LSubject: Rollback
Segs
crash.
-Mensaje original-De: venkat Rama
[mailto:[EMAIL PROTECTED]Enviado el: miércoles, 16 de julio de
2003 6:34Para: Multiple recipients of list
ORACLE-LAsunto: Rollback Segs
Hi Friends,
I have 150Gb database with 30 rollback segs with
optimal settings!! for normal operatio
t
me if I'm wrong).
Hope
this helps.
Cheers
Lee
-Original Message-From: Fermin Bernaus
[mailto:[EMAIL PROTECTED]Sent: 16 July 2003 09:09To:
Multiple recipients of list ORACLE-LSubject: RE: Rollback
Segs
Which database version are you using? as long as I know,
do it is to
disable all except the one we want it to use.
-Mensaje original-De: Lee Cullip
[mailto:[EMAIL PROTECTED]Enviado el: miércoles, 16 de julio
de 2003 10:54Para: Multiple recipients of list
ORACLE-LAsunto: RE: Rollback Segs
set
transaction use rollback segment
llback seg.
Cheers
Lee
-Original Message-From: Fermin Bernaus
[mailto:[EMAIL PROTECTED]Sent: 16 July 2003 11:19To:
Multiple recipients of list ORACLE-LSubject: RE: Rollback
Segs
Yes you are right, sorry but I made a mistake, since we
are using Baan IVc4 which works agai
c rollback segment, so the only way for us to do it is
to disable all except the one we want it to use.
-Mensaje original-
De: Lee Cullip [mailto:[EMAIL PROTECTED]
Enviado el: miércoles, 16 de julio de 2003 10:54
Para: Multiple recipients of list ORACLE-L
Asunto: RE: Rollback Segs
set trans
When we enter data or delete date(modifying data
blocks) , that will generate some redo.
and my doubt is:
when we delete some data in the database, that will
further insert information into Rollback segment
transaction tables (means furtherly it is modifying
some block of oracle).
Will this g
2 AM
> To: Multiple recipients of list ORACLE-L
> Subject: Redo for Rollback segs
>
> When we enter data or delete date(modifying data
> blocks) , that will generate some redo.
>
> and my doubt is:
>
> when we delete some data in the database, that will
> further i
OTECTED]
>To: Multiple recipients of list ORACLE-L <[EMAIL PROTECTED]>
>Subject: RE: Redo for Rollback segs
>Date: Fri, 27 Jul 2001 07:36:50 -0800
>
>Yes, __ALL__ changes to any of the database blocks are recorded in the
>Redo.
>
>Regards,
>
>- Kirti Deshpande
>
Rachel Carmichael wrote:
>
> to expand on this:
>
> changes to the rollback segments will generate entries in the redo logs.
> They ARE database blocks.
>
> In fact, rolling back something will generate change vectors as well.
>
> Rachel
kinda dumb question, do changes to rollback segmants us
oo.com> cc:
Sent by: Subject: Redo for Rollback segs
[EM
of list ORACLE-L <[EMAIL PROTECTED]>
>Subject: Re: Redo for Rollback segs
>Date: Fri, 27 Jul 2001 09:51:30 -0800
>
>Rachel Carmichael wrote:
> >
> > to expand on this:
> >
> > changes to the rollback segments will generate entries in the redo logs.
> > The
;
>Subject: Re: Redo for Rollback segs
>Date: Fri, 27 Jul 2001 10:57:40 -0800
>
>
>Yes, it will.
>
>The concepts manual is your friend.
>
>Jared
>
>
>
>
>
> prasad maganti
> list ORACLE-L <[EMAIL PROTECTED]>
On Fri, 27 Jul 2001,Rachel Carmichael scribbled on the wall in glitter crayon:
->It is? Ah, but will it bring me tea and toast when I am sick in bed?
no, but it will put you to sleep.;-)
->
->Rachel. it's a silly sort of day today, too many meetings with too many
->sales reps
i'm not allow
If rb's are small, then there is a good chance that
just when they are about to aged out of the buffer
cache, hey presto! another session needs them again -
so you end up with a nice scenario where they spend
their life in the cache and cause virtually no IO.
As they get larger and larger, the ti
Thanks.. makes sense...
Can I use x$bh or something to determine whether it's nicely squeezed into the
buffer at some point in time?
On Mon, 12 Nov 2001 02:00:21 -0800, you wrote:
>If rb's are small, then there is a good chance that
>just when they are about to aged out of the buffer
>cache, he
Generally I try monitor v$filestat to see what
proportion of IO is due to rollback...The ideal is of
course nil, but to do so pretty much means giving
everyone a couple of dozen 1555's every day :-)
So you end up with the balancing act of trying to make
them as small as possible without introduci
Thanks... good info...
On Mon, 12 Nov 2001 06:45:21 -0800, you wrote:
>Generally I try monitor v$filestat to see what
>proportion of IO is due to rollback...The ideal is of
>course nil, but to do so pretty much means giving
>everyone a couple of dozen 1555's every day :-)
>
>So you end up with t
19 matches
Mail list logo
