Conference decwet::winnt-clusters

    Hi,
    
    Just to addon to the above notes, when I shutdown the First Server, the
    database Salary did failover to the second Server. Thus on the Client
    PC, when I execute the ISQL_w and connect to the NT_Cluster1, I could
    see both the database Salary and Support. When I startup the first
    Server, the database Salary failback to the First Server. 
    
    When I do a refresh, I could only see the Salary database but not the
    Support database. I do not know what was the problems, why the Client
    PC could only see one database when both Server are up and running ?
    
    Do I need to apply any SQL service pack on the Client side too ?
    
    Another question:
    
    If we are creating an IP Failover object, do we add it to the failover
    group that we have created for the shared disk and SQL Failover object
    previously or do we create a new Failover group and add the IP failover
    object to the new group ? Which is correct or what differences does it
    make ?
    
    How can we tell or test whether the IP failover is running OK ?
    Can we have SQL Client to login to the SQL Server through IP, since
    there is a TCP/IP socket option provided under SQL Client Configuration
    utility instead of Named Pipes ?
    
    What are the steps or procedure how we can configure the IP Failover
    correctly and a simple test to verify that the IP failover is running
    fine using the Client PC ?
    
    Please advise as I need to get the Cluster running for a demo
    presentation. Thank you very much.
    
    
    
    
    

There are new procedures for SQL named pipe connections with V1.1. You will have to
configure a named pipe alias through the SQL client configuration Utility. Detailed
procedures can be found in the Cluster Adminstrators guide for V1.1 or look in the
Course materials I wrote.


As for the IP question, you would create an IP object and include it in the group.
Reference the Dual SQl with named pipe and socket connections Lab in the course materials
or  The Adminstrators guide.

Donna

    Hi Donna,
    
    Can't read your reply at every end of your lines.
    
    What about my Client PC when connected to the NT_Cluster1, I
    could only see one database called Salary ? I have actually created two
    database each on each disk in the BA350 and I should able to see both
    the database when connected to the Cluster through the ISQL_w command.
    
    As I am not very familiar with SQL Server, in the Admin guide for V1.1,
    it say for client conections using named pipes:
    
    1.	Type the SQL Server database alias
    2.	Type DLL name for the named pipes Net-Library
    3.	Connection String box,use this syntax:
    	\\ClusterName\pipe\ObjectName\query
    
    My current configuration is:
    
    I have two Server , one called A2100 and the other A1000
    NT Cluster Alias: NT_Cluster1
    On A2100, I created an SQL Failover Object called SQLfailoverobj1
    Using SQL Enterprise Manager, I created Database Device called HR, and
    Database called Salary.
    
    On A1000, I created an SQL Failover Object called SQLfailoverobj2
    I created Database Device called Tech and Database called Support.
    
    With the above info, I was not too sure what to type for the SQL Server
    database alias but guess that it could be the database Salary.
    Therefore I inputed the info as follow but could not get it connected.
    
    1.	In Server box, I type : Salary
    2.	DLL named pipes Net-Library, I type: DBNMP3.DLL
    3.	Connection string: \\NT_Cluster1\pipe\SQLfailoverobj1\query
    
    I have tried to change the Server box with the following names A2100
    and NT_Cluster1 but it will give the same errors :
    
    Unable to connect: SQL Server is unavailable or does not exist.
    Access denied
    
    But if I do not use the advanced option, I just simply use the default
    which I need only to key in the NT_Cluster1 and the login name sa
    followed by the passswd , I could get connected to the cluster. Is
    there any input which I have not done properly when trying to use the
    Advance option to login to the Cluster ?
    
    As for the client connections using IP, what is the DLL name for the
    TCP/IP Net-Library for Windows NT and Windows 95 ? There is no list to
    choose from the drop-down list TCP/IP Sockets.
    
    Please advise. Urgent.
    Thanks
                               
    Please advise.
    Thanks
    
    

    Hi,
    
    Anyone out there able to help me ? Urgently need help for demo
    presentation to our customers.
    
    Help
    Thanks

    Re: .-1
    
    >Anyone out there able to help me ? Urgently need help for demo
    >presentation to our customers.
    
    I too have been struggling with this scenario.
    I eventually got it to work on an NT WorkStation V4.0 client, but not
    on a Windows 95 client, so far....
    
    As the documentation states (Answers I entered):
    Using the client configuration utility | advanced
    Add Server Name: AlphaCluster
    DLL Name: dbnmp3
    Connection String: \\AlphaCluster\pipe\AlphaCluster\query
    OK
    Done
    
    When I connected from ISQL/w on the NT Client the database I created on 
    Drive S appeared in the DB: drop down box.
    If I then added the alternate pipe via the config utility I could then
    see the database created on Drive T, but not the db from Drive S. 
    So if you wish to be able to access both dbs from the client, connect
    to one database, then update the client config for the other database
    and then connect from the ISQL/w drop down menu via another new window to 
    the other database. 
    
    You will of course see both dbs when both are running on one server
    (failed over), but will lose if on failback...
    
    Clear as mud....???
    
    Hth,
    
    
    /Shaun

   Donna's reply reformatted. I'll get a pointer to the reference
   course material ...



There are  new  procedures  for  SQL named pipe connections with V1.1.
You  will  have to configure a named pipe alias through the SQL client
configuration  Utility.   Detailed  procedures  can  be  found  in the
Cluster Adminstrators guide for V1.1 or look in the Course materials I
wrote.

As for  the  IP question, you would create an IP object and include it
in  the  group.   Reference  the  Dual  SQl with named pipe and socket
connections Lab in the course materials or The Adminstrators guide.

Donna

   See the previous topic, reply 576.2, for access to the referenced
   course material.