SQL Server 2000 and Virtual Servers

  • I hope I can explain myself correctly here. I have a W2k Server running MSCS and SQL Server 2000 sp3 with one Virtual Server - one instance. I ran SQL Server setup to install/create a NEW Virtual Server instance - before the new cluster resource was created. We created the new Cluster resource and then reran SQL Server setup to install/create the new Virtual Server instance. Now the Install will not let me use the same name for the VIrtual Server that I used in the first failed attempt at the SQL Server install - says it is in use. I cannot uninstall it or find reference to it in the registry. Can anyone help? I guess my real problem is that I cannot find any good documentation - MS Site, BOL etc.. that walks me through the new virtual Server setup on 2000.

    Any help would be appreciated.

  • Did you install the instance on existing virtual server or you created a new virtual server, then try to install sql server on the virtual server?

     

  • We have had this Instance up and running for quite awhile. I did not do the original install way back when. We ran SQL Server 2000 Setup program yesterday to attempt to ADD a new virtual server instance to this installation. Because they jumped the gun and ran the SQL Server setup before the new cluster resource was created, had to cancel out of the setup program. Now the new Virtual Server name is 'in use' and I cannot rerun setup. Aren't there any step-by-step instructions from MS for creating a new Virtual Server in a Cluster?

  • Take the cluster resource offline, drop it and try it again.

    BOL search on 'How to create a new failover cluster (Setup)'. It has step-by-step instructions.

     

  • Thank you for the BOL search!!!

    That makes sense, To drop and recreate the MSCS cluster resource. But on the actual SQL Server Install, it is saying that the Virtual Server Name already exists - SQL Server Setup does not even get to the point where I define which cluster resource to use for the data files.. Does that mean that there is some kind of dependancy (hung) between the Virtual Server name that we definied in the first (failed) install and the Cluster resource?? Thank you for your help and patience.

  • You can use the existing virtual server name or if you do not like it, create another one. Both are valid options.

  • The new Virtual Server Name we are using is defined in WINS and DNS and is our preferred naming convention - so I will use the same Virtual Server name . I'll have a hard time convincing them to drop and recreate this new Cluster Resource as it will have to be done after-hours. I don't see how doing that will enable me to rerun SQL Server Setup without having the same problem? (Problem being - SQL Server setup is telling me that the Virtual Server name already exists). I think maybe I'm just missing something here.

  • That comes to the same issue of none-clustered installation. You may have to clean up registry(on every node) before try it again.

  • The registy on both Servers (both nodes in the cluster) have no reference to the new Virtual Server name - which is why I can't understand why SQL Server setup won't run saying the name already exists!!!

    I'll have to keep poking around. Thank you!

  • i have a related question see http://www.sqlservercentral.com/forums/shwmessage.aspx?forumid=54&messageid=99334 plz

    Have a look at http://www.microsoft.com/technet/treeview/default.asp?url=/technet/prodtechnol/sql/deploy/confeat/failclus.asp

    A W2K Advanced Server Cluster with 2 Nodes and SQL 2000 Enterprise Edition and 2 running SQL (normal state: one SQL Server on each node, failover: both SQL server on one node) has 5 different IP adresses and related names:

    1 physical node 1

    1 physical node 2

    1 virtual server 1

    1 virtual server 2

    1 cluster

    makes 5

    You do have already one SQL Server with one instance running on your system. Looks like a second SQL Server must be a named instance, not a default instance (my question in this forum and described in the MS link) - something you have to choose during SQL2K setup. SQL2000 setup recognizes that it has been started in a clustered environment.

    My problem is that in my cluster both SQL 7 servers are default instances and an upgrade is necessary (killer reason: Full Text Search is supported in Cluster environment only with SQL2000, not 7, and it is not working with 7). changing one default instance to a named instance means that the connect string changes   Failover is currently working properly for both SQL 7 instances. several times checked during installation of MS W2K Patches which required reboot.

  • liebla,

    In your SQL7 environment, how did you failover two defaults if one node die?

    SQL7 will run one version on a machine.

Viewing 11 posts - 1 through 11 (of 11 total)

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