May 20, 2014 at 7:05 am
Hi,
I am planning to move an existing standalone sql instance to a newly installed clustered infrastructure.
The challange here is; i must use the standalone server as a member of the cluster.
Existing sql infr. is as below;
Microsoft SQL Server 2008 R2 (SP1) - 10.50.2550.0 (X64) Enterprise Edition (64-bit) on Windows NT 6.1 <X64> (Build 7601: Service Pack 1).
Only one instance is running.
System databases are on local hard drive (D:).
Some user databases are on local disk (D:) some are on storage (F:) which is host bus (fiber) connected to sql server.
What I am planning to do is;
Connect host bus adapters to existing storage area on which databases are already hosted.
Check if OS can access to the drive (F:) and are db files seen.
Install Failover Clustering on new server
Install SQL server and Set SQL server as clustered service
Set existing storage area (F:) as resource to clustered SQL server service
Shutdown all applications connected to old sql server
Kill all connections to databases
Detach all databases from old server.
Copy locally (D:) hosted databases to storage (F:).
Shutdown SQL services on old server.
Attach all user databases to new sql server clustered service.
Redirect all aplications to new sql cluster name.
??? Here is it applicable that creating an alias in DNS which resolves old server name to cluster virtual ip address ?
Verification of users that all applications are running ok with the new server.
Reinstall old server OS and join it to cluster
Could this method work or do i need some modifications ?
May 20, 2014 at 7:26 am
It sounds to me like you're planning to switch the fibre volume with the standalone instances system databases and connect it to a new server as a clustered instance, is this assumption correct?
Is the exsiting stand alone instance a default or named instance?
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
May 20, 2014 at 7:39 am
Your assumption is partially right.
Standalone server is connected via fibre to F: drive (LUN).
System databases and some user databases are hosted locally on D: (local drive) other user databases hosted on F: (LUN).
I will only move user databases from D: to F: (local to LUN). I will not move system databases.
Yes the standalone one is default instance.
May 20, 2014 at 7:56 am
So, you have a new server for your planned cluster to add to the current machine with the stand alone instance.
Do you have a maintenance window or agreed downtime period in which to perform this move?
I would build the new cluster on the new single node primarily with new storage presented. Get the instance up and running and move across any system objects such as logins, jobs, etc.
Also get across as many user databases as possible then point the applications connection strings to the new clustered instance.
If you're wanting to use the existing servername you'll need to shut the server down before you can use the servername as the VNN on the new clustered instance.
-----------------------------------------------------------------------------------------------------------
"Ya can't make an omelette without breaking just a few eggs" 😉
Viewing 4 posts - 1 through 4 (of 4 total)
You must be logged in to reply to this topic. Login to reply