SQL Server Performance

A/P P/A cluster with SQL 2005

Discussion in 'SQL Server 2005 Clustering' started by mike carpendale, Feb 21, 2006.

  1. mike carpendale New Member

    Hi,
    I have a couple of questions around Clustering SQL serves, specifically Active/Passive and Passive/Active

    Details are

    Windows 2003 R2 64 bit STD
    Running on 2 HP DL385 with an MSA 1000
    SQL 2005 64 bit STD

    The Clients wants to make use of both servers in the cluster, ie each server node will have its own virtual server entity of SQL,
    From what I have read and gathered I#%92m thinking you do A/P install on node 1 pointing to the first virtual server entity, which will be (eg) SQL_V1 and will reside in c:program filesSQLHNSQL_V1
    To get he P/A install complete I ‘m thing some thing along the lines of this; Install SQL on node 2 pointing it to the second virtual server entity, which will be SQL_V2 and will reside in c:program filesSQLHNSQL_V2
    Each server virtual server entity will have its own data and log disks and will be running separate databases. I don#%92t want to load balance, juts have fault tolerance with making use of all of the resources in the cluster.

    Does this sound feasible?[?]
    Any help, comments would be much appreciated

    Thanks
    Mike
  2. mulhall New Member

    Aah, no. You're talking about Active/Active clustering.

    You don't install SQL Server on a node, you install it on the clustered virtual server.

    First of all you create a virtual server (see Microsoft Cluster Server (MSCS)), and this sets aside some of your shared disk space that you will use for a particular instance, a name and an IP address.

    Then you install an instance of SQL Server 2005 - it will realise the nodes are clustered and ask for a Virtual SQL Server name.

    Repeat for your second instance.

    Use Cluster Administrator to assign which node is active for which virtual server (and hence which sql instance)

Share This Page