| T.R | Title | User | Personal Name
 | Date | Lines | 
|---|
| 575.1 | cross-server joins are a SQL feature; not a Clusters feature | LJSRV1::BOURQUARD | Deb | Thu Jan 23 1997 15:53 | 0 | 
| 575.2 | Will version 1.1 implement this feature or not? | NEOV00::KLOPEZ | Ay Caramba! | Fri Jan 31 1997 14:15 | 21 | 
|  |     Hello,
    
    I thought this was a Cluster software feature. How come other vendors
    are claiming they can have two different databases opened on their
    clusters. 
    
    How would this work? Suppose you have a database on server 1 and a
    different database on server 2, how would the failover take place?
    would the clients of server 1 suffer if database 2 would failover?
    Would we need to shutdown SQL server on the surviving node?
    
    Our customer was curious on how would this take place,
    
    We have sold two Windows NT clusters, and have losed many more
    customers, who don't see the benefit of having two systems and not been
    able to run SQL server on both.
    
    Thanks for your information
    
    Katya
                     
 | 
| 575.3 | joins across servers is different from sql server running on both cluster servers | LJSRV1::BOURQUARD | Deb | Mon Feb 03 1997 10:13 | 16 | 
|  | I believe Clusters 1.0 Service Pack 1 supported dual SQL servers -- 
the ability to run SQL Server on both servers of a cluster.  If not, then
it's definitely in Clusters V1.1 (or V1.0 Service Pack 2 if you're running
NT351).
This is a very different feature from performing joins across
tables residing on different servers which is a feature (or not) 
of SQL Server and is what the base note was questioning.
There are other notes in this conference regarding dual SQL
servers which may provide the answers to your other questions.
(I don't know the answers or I'd repeat them here...)
Hope this helps to eliminate the confusion.
- Deb 
 | 
| 575.4 | V1.1 allows running SQL on both nodes | DECWET::CAPPELLOF | My other brain is a polymer | Fri Feb 07 1997 13:18 | 17 | 
|  | >    How would this work? Suppose you have a database on server 1 and a
>    different database on server 2, how would the failover take place?
>    would the clients of server 1 suffer if database 2 would failover?
>    Would we need to shutdown SQL server on the surviving node?
    
    The following features will be in NT Clusters V1.1 (for NT 4.0) and NT
    Clusters V1.0+SP2 (for NT 3.51).
    
    With these versions of the cluster software, you can run the
    MSSQLServer process on both nodes at the same time.  Server 1 exports
    Database 1 to clients, and Server 2 exports Database 2 to clients.  If
    Server 1 fails, Server 2 will export both Database 1 and Database 2.
    
    This does not imply that NT Cluster software enables cross-node joins
    between databases.  NT Cluster software does NOT enable this
    functionality.  As Deb pointed out in .-1, this is an SQL feature, not
    a cluster feature.  Sorry.
 |