repairing replicated tables

Hi all,
I have a sql2005 merge replication running nightly between SQl 2005 standard
and SQLexpress. (The subscription is on the SQLexpress) last night it began
failing. I ran a DBCC checkDB and found one table has 3 inconsistancies in
it and the lowest level of repair is repair with loss. This box is on its
way out and i need to band-aide it until the new one arrives and is setup.
My question is, do I need to drop the replication before putting the DB in
single user mode? Is there a way to repair the table without running DBCC
checktable with the repair option?
TIA,
Joe
Is the problem related to some indexes and do you get a RID ID error?
If so this is related to some indexes (on system tables IIRC) and you
can drop them and recreate them.
Run CHECKDB again note the object ID which is experiencing the error
and evaluate whether it is an index or not. Script out the index, drop
it and recreate it. There will be no data loss associated with this.
If it is table related you will have to bcp the data out noting where
failure occurs and then work around that/those rows using the firstrow
and lastrow options. You can use DBCC Page to look at the problem
pages if you need to be so granular in your data retrieval.
On Jan 15, 2:16 pm, jaylou <jay...@.discussions.microsoft.com> wrote:
> Hi all,
> I have a sql2005 merge replication running nightly between SQl 2005 standard
> and SQLexpress. (The subscription is on the SQLexpress) last night it began
> failing. I ran a DBCC checkDB and found one table has 3 inconsistancies in
> it and the lowest level of repair is repair with loss. This box is on its
> way out and i need to band-aide it until the new one arrives and is setup.
> My question is, do I need to drop the replication before putting the DB in
> single user mode? Is there a way to repair the table without running DBCC
> checktable with the repair option?
> TIA,
> Joe
|||Thank you, it was an index issue and droping and recreating the Indexes did
the trick.
Thanks again.
"Hilary Cotter" wrote:

> Is the problem related to some indexes and do you get a RID ID error?
> If so this is related to some indexes (on system tables IIRC) and you
> can drop them and recreate them.
> Run CHECKDB again note the object ID which is experiencing the error
> and evaluate whether it is an index or not. Script out the index, drop
> it and recreate it. There will be no data loss associated with this.
> If it is table related you will have to bcp the data out noting where
> failure occurs and then work around that/those rows using the firstrow
> and lastrow options. You can use DBCC Page to look at the problem
> pages if you need to be so granular in your data retrieval.
> On Jan 15, 2:16 pm, jaylou <jay...@.discussions.microsoft.com> wrote:
>