Re: Asynchonous CDC in 10GR2
- From: "Arul Ramachandran" <contactarul@xxxxxxxxx>
- To: "David Lord" <davidclord@xxxxxxxxxxxxxx>
- Date: Fri, 9 Jun 2006 11:02:02 -0700
David,
Yes, I noticed that yesterday. I wish it was based on range partitioning on
TIMESTAMP$, so I can have monthly partitions (that way I can do exchange
partition). No luck:( ..... I probably can achieve this by defining
multiple CHANGE SETs for a month using the begin_date, end_date - except
that it will still be partitioned on CSCN$. Any comments on this?
I tried dropping the change table using 'drop table ..', so that I can
re-create it as range partitioned on TIMESTAMP$. But Oracle would not allow
that, as the table can be dropped only via
DBMS_CDC_PUBLISH.DROP_CHANGE_TABLE.
1. After letting dbms_cdc_publish.create_change_table create the change
table, is it ok to (manually) add columns to it? I need to >>add a couple of
snapshot columns that are not present in the base table.
2. Would it be possible to implement a database insert trigger on the
change table to populate the snapshot columns of the change >>table?
I could add columns to the table. I also added a before insert trigger , the
trigger compiled fine, but would not execute. I knew this is the behavior in
Streams, same goes here as well. I believe the normal database triggers
won't work in this situation ...if you have any ideas, that would be great.
Thanks,
Arul
On 6/8/06, David Lord <davidclord@xxxxxxxxxxxxxx> wrote:
Arul,
The change tables are partitioned on a high value of the CSCN$. As to
adding columns, everything I can see suggests that the change tables are
perfectly normal, there doesn't appear to be any 'magic' going on. In
fact, when I logged a TAR for a performance problem with a query on a
change table, Oracle support suggested that it was okay to add indexes
to them (I ran into problems when I did because the purge job marked the
indexes as unusable).
Regards
David
On Wed, 7 Jun 2006 10:42:52 -0700, "Arul Ramachandran"
<contactarul@xxxxxxxxx> said:
> David,
>
> Thanks for your reply.
>
> I am aware of the ALTER_CHANGE_TABLE, but my intention is to add columns
> to
> the change table that are not present in the source table. Hence, I was
> wondering if I could manually alter the change table.
>
> Regarding, partitioning of change table, I see from your comment CDC
does
> partition it (probably a max val partition?).
>
> I don't have my environment yet, otherwise I can check out.
>
> Regards,
> Arul
--
David Lord
davidclord@xxxxxxxxxxxxxx
--
http://www.fastmail.fm - Or how I learned to stop worrying and
love email again
--
Arul
- Follow-Ups:
- Re: Deadlocks followed by ORA-07445 then database crashes
- From: Paula Stankus
- References:
- Asynchonous CDC in 10GR2
- From: Arul Ramachandran
- Re: Asynchonous CDC in 10GR2
- From: David Lord
Other related posts:
- » Asynchonous CDC in 10GR2
- » Re: Asynchonous CDC in 10GR2
- » Re: Asynchonous CDC in 10GR2
1. After letting dbms_cdc_publish.create_change_table create the change
2. Would it be possible to implement a database insert trigger on the
Arul,
The change tables are partitioned on a high value of the CSCN$. As to adding columns, everything I can see suggests that the change tables are perfectly normal, there doesn't appear to be any 'magic' going on. In fact, when I logged a TAR for a performance problem with a query on a change table, Oracle support suggested that it was okay to add indexes to them (I ran into problems when I did because the purge job marked the indexes as unusable).
Regards David
On Wed, 7 Jun 2006 10:42:52 -0700, "Arul Ramachandran" <contactarul@xxxxxxxxx> said: > David, > > Thanks for your reply. > > I am aware of the ALTER_CHANGE_TABLE, but my intention is to add columns > to > the change table that are not present in the source table. Hence, I was > wondering if I could manually alter the change table. > > Regarding, partitioning of change table, I see from your comment CDC does > partition it (probably a max val partition?). > > I don't have my environment yet, otherwise I can check out. > > Regards, > Arul -- David Lord davidclord@xxxxxxxxxxxxxx
-- http://www.fastmail.fm - Or how I learned to stop worrying and love email again
- Re: Deadlocks followed by ORA-07445 then database crashes
- From: Paula Stankus
- Asynchonous CDC in 10GR2
- From: Arul Ramachandran
- Re: Asynchonous CDC in 10GR2
- From: David Lord