October 1, 2007 at 9:10 am
The title says most of the issue that I'm currently testing.
I'm currently using two stand alone SQL Server 2005 sp2 developer machines on windows 2k3 sp2 enterprise and I would like to add a cluster over those installations without rebuilding any SQL Server.
At the moment, the cluster is working properly, quorum has been added, DTC also, work fines. However I'm not sure how the two stand alone SQL Server can be reconfigured to behave in a cluster mode and of top of that is it feasible first?
Your helps will be most welcome.
Thks
October 1, 2007 at 9:25 am
I'm not sure there is any easy way to go about this short of hacking the registry which could do more harm than good. I would suggest an alternative approach. Install a new named MSSQL virtual server in the cluster and migrate your databases, etc. to this instance. Post migration, uninstall the default instance.
Tommy
Follow @sqlscribeOctober 1, 2007 at 9:29 am
That's what I fear, reinstalling SQL even a new instance.
The other issue I'm currently struggling with now is in the cluster, I don't have the SQL Server resource which might be added when adding a new clustered installation only.
Although I like editing (read hack) the registry I must stay with supported tools provided for doing such tasks.
I'll give it a try, installing a new instance.
October 1, 2007 at 9:37 am
The resources can always be added after the fact as well. But as you pointed out - hacking should be your last resort. You should be able to install a new instance (named) in the cluster and just dettach, reattach your databases, script your logins, jobs, etc. The just uninstall the default instance.
Good luck -
Tommy
Follow @sqlscribeOctober 1, 2007 at 10:01 am
Creating the SQL server resource in Cluster admin isn't going to be your problem. What IS going to be your problem is that you would have to modify at LEAST one of the 2 dev installs, unless you got "lucky" and someone installed one of them into a named instance.
Now - checklist time (probably incomplete, but should be a running start):
- do the machines have the resources available to run BOTH instances?
- do the two instances have different names (instance name).
- does each instance access the data files on volumes that are cluster resources
- do you have the 6 IP addresses set up correctly?
- do you have the 2 cluster groups set up, each with their respective IP address resources, volumes, generic services and machine names set up.
Although a fun exercise - unless this is a play environment you can afford to have down for a while, trying to do this by hand isn't usually recommended. There's a big difference between what CAN be done, and what SHOULD be done. Chances are that unless you planned ahead of time to do this, something isn't set up in a way conducive to make this happen, meaning - you will likely have to run a reinstall of SQL on one or both machines.
----------------------------------------------------------------------------------
Your lack of planning does not constitute an emergency on my part...unless you're my manager...or a director and above...or a really loud-spoken end-user..All right - what was my emergency again?
October 1, 2007 at 11:17 am
Hi Matt
Thks for your answer.
What I want to do is what SHOULD be done with proper tools to do it (using Best Practices as often as possible) although I always like to know up to which extend it CAN be done 😉 (In case we are absolutely forced to do it regardless of consequences)
Yes it is a test environment before going into the wild. I'm testing different cluster setup and their feasibility. From what I've found on the web as well as in forums, I believe the only way to have a clustered environment for SQL Server is to start from fresh. Leveraging an already setup installation is mostly unfeasible without proper hack.
Still I was testing my luck if some DBA's here did have that kind of setup or when through something similar.
Viewing 6 posts - 1 through 5 (of 5 total)
You must be logged in to reply to this topic. Login to reply