Adding an instance to a SQL2008 Cluster

  • How would one go about adding an instance to a SQL2008 cluster? I am running a Server 2008 Cluster with the default instance installed and working beautifully, but now I need to install a 2nd instance and can find no way to do it. I can install a stand alone SQL2008 instance, or create a new cluster, but I don't want to do either of those, I want a new instance on my same cluster. I have new disks ready to go, but no matter what path I use to get to the installation, I am either looking at no cluster information at all, or am being required to provide a new cluster name. I can't seem to find anyone who has actually done this and was hoping someone here might at least get me pointed in the right direction. I appreciate anything you guys can tell me.

  • Each instance on a cluster needs to be installed on it's own resources. So, in Cluster Administrator, add a new group, and add your (new) Drive resources to that group. You should then be able to install a second instance to those drives/group. Don't forget that you will need a virtual SQL computer name / ip for the instance you are installing.

    Wayne
    Microsoft Certified Master: SQL Server 2008
    Author - SQL Server T-SQL Recipes


    If you can't explain to another person how the code that you're copying from the internet works, then DON'T USE IT on a production system! After all, you will be the one supporting it!
    Links:
    For better assistance in answering your questions
    Performance Problems
    Common date/time routines
    Understanding and Using APPLY Part 1 & Part 2

  • This issue is occurring on a Windows Server 2008 R2 cluster. As far as I can thell there aren't any groups in the cluster setup anymore, just services and applications.

  • I think I have the root of the problem, but no solution yet. Installing SQL 2008 in a cluster on Server 2008 R2 requires SP1 to be slipstreamed in. After wrestling with this all day I remember I had to do that originally setting up the cluster. However, when I try to run my command line install I get this error:

    The timestamp marker on log files has not been set. the timestamp must be set prior to it being used for the first time.

    Error code 0x84B20001.

    Do I have to figure out the syntax to add some timestamp parameter to my command line call of the install? Any input would be greatly appreciated.

    Thanks,

    Scott

  • Has anyone here actually added a 2nd instance of SQL2008 on a Server 2008 R2 Cluster? I can't be the first one to run into this issue.

  • Hi, I do have 3 sql2008 instance on win2008 cluster server and its working fine.

    try this link. i have use this for my set up.

    http://www.mssqltips.com/tip.asp?tip=1687

    in your case, i think you'll need the "part 3" of that procedure under "Installing SQL Server 2008 on a Windows Server 2008 cluster"

    hope that helps..

  • My question is around installing the second instance, which nothing I can find covers. The only options I get when running the setup are to create a new cluster, adding a node, or modifying the existing instance. I don't really want another cluster, just another instance on the same cluster. Does anyone have the process for adding an instance to an existing cluster in Server 2008 R2?

  • paksyeet,

    Are you running 3 separate clusters or 3 instances on one cluster?

    Thanks,

    Scott

  • wierd.. but i do have..

    1 cluster(win 2008) with 2 nodes plus 3instance of ms sql server 2008.

    the above link help me with the set-up..

    please read my post carefully.. go to the part 3 of the installation process or here (http://www.mssqltips.com/tip.asp?tip=1709) and find the "Installing SQL Server 2008 on a Windows Server 2008 cluster"

    or from your option.. select the 'New SQL Server failover cluster installation' then add node the node for you ms sql cluster to work.

    to sum up..

    1. install a new failover cluster (named instance)

    2. add node for the named instance to work as sql failover cluster.

  • paksyeet,

    When I run the new SQL cluster option, it will not let me choose to run a new instance on the same existing cluster. It makes me provide a new cluster name, which I don't want to do. Are you sure you aren't running 3 separate clusters based on that option? This may be a difference between 2008 and 2008 R2, but I can't find anyone who has done this on R2 yet.

    Thanks,

    Scott

  • Hi Scott;

    sorry for the late reply, can you point me from the procedure i gave you in which you encounter creating that new server?

    As far as i can remember, those 'new cluster' are just part of your the sql instance name. or should i say this new cluster is only a resource group of sql since each named instance has its own resource.

  • Scott Wise (3/4/2010)


    How would one go about adding an instance to a SQL2008 cluster? I am running a Server 2008 Cluster with the default instance installed and working beautifully, but now I need to install a 2nd instance and can find no way to do it. I can install a stand alone SQL2008 instance, or create a new cluster, but I don't want to do either of those, I want a new instance on my same cluster. I have new disks ready to go, but no matter what path I use to get to the installation, I am either looking at no cluster information at all, or am being required to provide a new cluster name. I can't seem to find anyone who has actually done this and was hoping someone here might at least get me pointed in the right direction. I appreciate anything you guys can tell me.

    Hi Scott

    after reading this topic through i think you are getting confused. To install a second (named) clustered instance of SQL Server 2008 you select the option "New SQL Server fail over cluster installation". You will be asked to provide a new virtual network name and IP address along with an instance name. Then add other nodes by running setup directly on each node (as you did for the default instance).

    Remember a new SQL Server fail over cluster installation is the very action of adding a second instance. Clustered instances do not work in the same way as non clustered.

    i.e. on a stand alone machine server name MYSERVER with a default and named instance would be

    MYSERVER (default)

    MYSERVER\INST1 (named)

    On a cluster, each instance you install needs a new virtual network name (which is essentially the same as a computername)

    MYSERVER-A\INST1 (named)

    MYSERVER-B (default)

    I hope this goes some way to explaining this for you

    Regards

    Perry

    -----------------------------------------------------------------------------------------------------------

    "Ya can't make an omelette without breaking just a few eggs" 😉

  • Perry Whittle (3/18/2010)


    Scott Wise (3/4/2010)


    How would one go about adding an instance to a SQL2008 cluster? I am running a Server 2008 Cluster with the default instance installed and working beautifully, but now I need to install a 2nd instance and can find no way to do it. I can install a stand alone SQL2008 instance, or create a new cluster, but I don't want to do either of those, I want a new instance on my same cluster. I have new disks ready to go, but no matter what path I use to get to the installation, I am either looking at no cluster information at all, or am being required to provide a new cluster name. I can't seem to find anyone who has actually done this and was hoping someone here might at least get me pointed in the right direction. I appreciate anything you guys can tell me.

    Hi Scott

    after reading this topic through i think you are getting confused. To install a second (named) clustered instance of SQL Server 2008 you select the option "New SQL Server fail over cluster installation". You will be asked to provide a new virtual network name and IP address along with an instance name. Then add other nodes by running setup directly on each node (as you did for the default instance).

    Remember a new SQL Server fail over cluster installation is the very action of adding a second instance. Clustered instances do not work in the same way as non clustered.

    i.e. on a stand alone machine server name MYSERVER with a default and named instance would be

    MYSERVER (default)

    MYSERVER\INST1 (named)

    On a cluster, each instance you install needs a new virtual network name (which is essentially the same as a computername)

    MYSERVER-A\INST1 (named)

    MYSERVER-B (default)

    I hope this goes some way to explaining this for you

    Regards

    Perry

    Perry, are sure that will work on cluster?

    http://social.msdn.microsoft.com/Forums/en/sqlsetupandupgrade/thread/906a8c25-ce3c-402b-9a06-648bd83bdfde%5B/url%5D

    Scott, i think if you need two instances on cluster then you should go for Active\Active clustering with two named instances.

    EnjoY!
  • GTR (3/18/2010)


    Perry, are sure that will work on cluster?

    http://social.msdn.microsoft.com/Forums/en/sqlsetupandupgrade/thread/906a8c25-ce3c-402b-9a06-648bd83bdfde%5B/url%5D

    Scott, i think if you need two instances on cluster then you should go for Active\Active clustering with two named instances.

    you dont need active\active.

    you can have a default and a named instance on clustered environments they dont need to be all named instances!

    My post suggests exactly the same as the link you posted. Install a new instance with new IP and virtual network name, then add nodes to this instance in exactly the same way as you do for the default instance. Clustered and non clustered machines work differently. On non clustered machines multiple instances are assigned to the same computername this does not happen in a clustered environment!

    The following details differences between clustered and non clustered environments,

    on a stand alone non clustered machine with server name MYSERVER with a default and named instance you would use

    MYSERVER (default)

    MYSERVER\INST1 (named)

    On a cluster, with node names NODE1 and NODE2 each instance you install needs a new virtual network name (which is essentially the same as a computername)

    MYSERVER-A\INST1 (clustered named)

    MYSERVER-B (clustered default)

    Whether clustered or non clustered the rules apply you may have one default instance only the rest are named. You may have 2 instances active on one cluster node this will provide an active\passive configuration.

    -----------------------------------------------------------------------------------------------------------

    "Ya can't make an omelette without breaking just a few eggs" 😉

  • OK, i made this real easy. The screenshots below shows the registry entries created when installing SQL Server and the instance configuration in the installer. Notice the default instance (MSSQLSERVER) is using network name NETWORKNAME1. Re run the installer selecting "Create new failover cluster" (which installs a new instance) and you have to supply a new network name, they must be unique each time. Dont confuse this with the installation to a non clustered machine, they are totally different.

    If you're still unsure PM me and i'll see if i can help more

    -----------------------------------------------------------------------------------------------------------

    "Ya can't make an omelette without breaking just a few eggs" 😉

Viewing 15 posts - 1 through 15 (of 16 total)

You must be logged in to reply to this topic. Login to reply