can i add a column to an article in transactionalreplication

hi
how can i add a column to the table at some specified location after publishing the article in transactional replication.is there any procedure or method.please explain
thanks®ards
reddy
Reddy,
sp_repladdcolumn is the standard way to add a column to replication. However
this will not permit you to place the column at a specific position in the
table. This is not normally an issue, as TSQL is not usually written in a
way that makes it order dependant - columns are referred to by name and not
by position. If this is a requirement for you, you'll need to drop the
publication, add the column then recreate the publication.
HTH,
Paul Ibison
|||paul,
thank you.
in transactional replication on subscriber side we have to set autoid property to no.now i am setting this manually.
we have script to set the autoid 'not for replication',like that any script you have to set the autoid property NO for all tables in a replication.please help.
thanks®ards
reddy
|||Reddy,
I haven't used this myself, but Hilary has posted up a
reference to a script he uses:
http://communities2.microsoft.com/co.../newsgroups/en
-us/default.aspx?
dg=microsoft.public.sqlserver.replication&mid=2239 a0b8-
d572-4d86-ac92-5109662bef6b
This updates to Yes - Not for Replication. However, you
could modify this script for your use. You'll still need
to alter the script for each table though.
Also I'd like to know a little more about why the
subscriber columns were originally set to have an identity
attribute and now it's being removed - (it's normally the
opposite :-)).
Regards,
Paul Ibison
|||paul,
first i will explain the way how i am doing transactional replication.first i will change the autoid column of publisher to not for replication.then i will restore the same in subscriber and set the auto id property to no.why means in transactional publis
her it self handles the autoid property so no need in subscriber.
is there any mistake in the above process please reply
thanks & regards
reddy
|||Reddy,
this is correct for a nosync initialization for standard
transactional replication (RO subscribers).
Regards,
Paul Ibison
|||paul
thanks
regards
reddy