Replication Deadlocks

Hello,
I am running multiple replication publication on a SQL 2k database, with
a separate server as distribution agent, then finally push subscriptions
to multiple databases across 5 servers.
All publications use row and column filtering in them.
My question is, can I use locking hints on the selection criteria for
the row filtering?
ie. Can I specify WITH (NOLOCK) on the row filtering so that it
eliminates the majority of the blocking / deadlocks?
Thanks
Dave
*** Sent via Developersdex http://www.codecomments.com ***
You could if it was a subquery. Do you have indexes on the join
condition/filter key in your filter?
Hilary Cotter
Looking for a SQL Server replication book?
http://www.nwsu.com/0974973602.html
Looking for a FAQ on Indexing Services/SQL FTS
http://www.indexserverfaq.com
"Dave" <Anonymous@.devdex.com> wrote in message
news:eqm8zv1THHA.5060@.TK2MSFTNGP02.phx.gbl...
>
> Hello,
> I am running multiple replication publication on a SQL 2k database, with
> a separate server as distribution agent, then finally push subscriptions
> to multiple databases across 5 servers.
> All publications use row and column filtering in them.
> My question is, can I use locking hints on the selection criteria for
> the row filtering?
> ie. Can I specify WITH (NOLOCK) on the row filtering so that it
> eliminates the majority of the blocking / deadlocks?
> Thanks
> Dave
> *** Sent via Developersdex http://www.codecomments.com ***
|||There is an index on the originating table/column, but the deadlocking
appears to be happening on the distribution database.
David
*** Sent via Developersdex http://www.codecomments.com ***

Replication Architecture question - Pull subscriptions on a SQL cluster...

Greetings:

I am working on a replication setup using transactional replication using with pull subscriptions and a separate distributor. The pull subscriptions are located on a SQL cluster using the virtual SQL Server name as the subscriber; when the box fails over, we get an error of missing replication.dlls. Researching further, we found that replication only works on one node of the subscriber.

Any ideas on what we did wrong in the setup?

Thanks,

Lee Everest

www.texas2oo.com/sqlblog

can you check the c:\program files\Microsoft Sql Server\80\com folder to see if the dll's exist there? If not try to copy and register them on the failing node.|||

We could do that, but i'm afraid that after I try to install one, several (or dozens) of others might show up missing, meaning that we didn't install it properly in the first place. The specific error that we get is:

Cannot load the DLL replsetoriginator extended procedure, or one of the DLLs it references. Reason: 126(The specified module could not be found.). The step failed.

I have no idea which DLL this function is located but could probably find out.

|||This is a missing or corrupt dll. When I have encountered these in SQL FTS clusters I have had to copy the dll's to the failed not and reregister them.|||So you are saying that the setup is correct, and that if the box fails over replication should pick right back up?|||

A friend of mine responded, below. Apparently replication shouldn’t miss a beat if failover occurs:

Never had this issue before. We are running 4 node clusters (active/active/active/passive) here and we have never had that issue. I'm going to guess it goes back to the install of SQL. I would just copy any missing dll from one node to the other. It sounds like SQL is looking for them so I think you should be safe just to copy them over.

Any thoughts?

|||That depends on the programming mechanism. Sometimes they need to be registered and sometimes they don't. It these dll's contain com compoents they will be, if not a simple file copy will work.

Replication Architecture question - Pull subscriptions on a SQL cluster...

Greetings:

I am working on a replication setup using transactional replication using with pull subscriptions and a separate distributor. The pull subscriptions are located on a SQL cluster using the virtual SQL Server name as the subscriber; when the box fails over, we get an error of missing replication.dlls. Researching further, we found that replication only works on one node of the subscriber.

Any ideas on what we did wrong in the setup?

Thanks,

Lee Everest

www.texas2oo.com/sqlblog

can you check the c:\program files\Microsoft Sql Server\80\com folder to see if the dll's exist there? If not try to copy and register them on the failing node.|||

We could do that, but i'm afraid that after I try to install one, several (or dozens) of others might show up missing, meaning that we didn't install it properly in the first place. The specific error that we get is:

Cannot load the DLL replsetoriginator extended procedure, or one of the DLLs it references. Reason: 126(The specified module could not be found.). The step failed.

I have no idea which DLL this function is located but could probably find out.

|||This is a missing or corrupt dll. When I have encountered these in SQL FTS clusters I have had to copy the dll's to the failed not and reregister them.|||So you are saying that the setup is correct, and that if the box fails over replication should pick right back up?|||

A friend of mine responded, below. Apparently replication shouldn’t miss a beat if failover occurs:

Never had this issue before. We are running 4 node clusters (active/active/active/passive) here and we have never had that issue. I'm going to guess it goes back to the install of SQL. I would just copy any missing dll from one node to the other. It sounds like SQL is looking for them so I think you should be safe just to copy them over.

Any thoughts?

|||That depends on the programming mechanism. Sometimes they need to be registered and sometimes they don't. It these dll's contain com compoents they will be, if not a simple file copy will work.