Another DR option is SQL Server Clustering. This happens to be my favorite approach, as it provides recovery of the entire
server, and full recovery can be achieved with
very little downtime.
In this post I am going to outline an A/P failover cluster. Basically, we've got two SQL Server instances, referred to as 'nodes', that use shared storage. One node is 'active', or the primary instance, and the other is 'passive', or the secondary instance. In the event of failure, ownership is transferred from the primary to to the secondary. At that point, the roles are switched, and the passive node becomes active.
We've got an alias on the network -- SQLCLUSTERSRV01 -- which targets whichever node is active, and satisfies any data requests to the shared storage. Everyone on your network talks to that alias. They do not know what node is active, or even how many nodes exist. I've tried to depict the general structure of the A/P cluster in the graphic below. I don't have Photoshop, so cut me some slack....
Put your v2012 installation media into the machine, and choose 'Installation'.
Choose 'New SQL Server failover cluster installation'.
Hit 'Ok' on the Setup Support rules dialog.
Accept the license terms.
Setup Support Rules dialog -- Review any errors or warnings that it produces.
Setup Role -- 'SQL Server Feature Installation' is checked. Hit 'Next'.
Feature Selection -- choose the features you need, hit 'Next'.
Feature Rules -- Please review any warnings that are returned.
Instance Configuration -- input the desired instance name and root directory.
Disk Space Requirements -- Be sure you have adequate space available.
Cluster Resource Group -- Input a name for the cluster resource group.
Cluster Disk Selection -- This is where you select the disks that will be shared.
Cluster Network Configuration - Input your IP and subnet(s).
Server Configuration -- Input your service accounts.
Database Engine Configuration -- On Server Configuration tab click 'Add Current User' to get yourself in there, and then 'Add', to input any additional DBAs.
Database Engine Configuration -- On Data Directories tab you will provide directory paths for your data files.
Error Reporting -- I never send the logging to MSFT, but it's your call.
Cluster Installation Rules -- You want to be sure there are no failures/warnings.
Ready to Install -- Go all the way down through the Summary, and check the Configuration file path, to be sure it is right.
Get some coffee....... Upon completion, just hit 'Close'. You're all good now.
Now if you go into the Windows Failover Cluster Manager (FCM), you will see your newly configured cluster server. You can get to the FCM through Administrative Tools, or Server Manager\Features, like I've pasted below. If I were you, I would become very comfy with the FCM. It provides a great deal of information regarding the status of your cluster; the nodes, the different cluster services, the storage... everything.
That's pretty much it for the installation of a SQL Server Failover Cluster. I will try to get back over time, with varied posts regarding the actual management and administration of the clusters.
Until then, this is a good read on Failover Clustering and the AlwaysOn Availability Groups: http://msdn.microsoft.com/en-us/library/ff929171(v=sql.110).aspx
I would also take a look at this video by Kendra Little in Brent Ozar's group. This is VERY good. http://www.brentozar.com/archive/2013/08/introduction-sql-server-failover-clusters-video/
No comments:
Post a Comment