![]() | |
![]() |
| | Thread Tools | Display Modes |
#41
| |||
| |||
|
|
You can create multiple instances without clustering. I think you are trying to fake a cluster to get around the "one default instance per computer" limitation. However, you still have only one default instance per cluster, so faking a cluster with a single node will not help. You can always create client aliases, but that gets time consuming and hard to manage. I still suggest DNS aliases for short name to IP address/port number combination. Check with your network guru. -- Geoff N. Hiten Senior SQL Infrastructure Consultant Microsoft SQL Server MVP "SQL Question" <SQLQuestion (AT) discussions (DOT) microsoft.com> wrote in message news:A627AC71-C0B6-42D4-B086-B0D104B260AB (AT) microsoft (DOT) com... Much thanks Geoff. Meanwhile I have tried and asked around and found more about the server setup that affects the SQL installation and my attempt of simplifying the way SQL instances are referenced. though the server has been set up as a single node cluster, it does not have a shared storage. It has a single SCSI drive that has a few partitions, with C used as system drive and quarom. This makes it impossible to assign different IP addresses to different instances, which is a requirement for the DNS aliasing. I don't know whether you have any other suggestions as to how I can achieve what I am trying -- to use a single one part name for the instances to be referenced, instead of using the machine-name\instance-name two part name. Thanks for your looking into this and advanced thanks for your possible reply. I will check back. "Geoff N. Hiten" wrote: The short answer is "no". Even though each SQL instance lives in its own virtual space, there can be only one default SQL instance per cluster. You can use a DNS alias to map a short name to the clustered instance name so when a Client connects to SQL1, they are redirected to the appropriate IP address and port number. -- Geoff N. Hiten Senior SQL Infrastructure Consultant Microsoft SQL Server MVP "SQL Question" <SQLQuestion (AT) discussions (DOT) microsoft.com> wrote in message news:E4116655-6B27-4E10-A307-C185E8CB1C6C (AT) microsoft (DOT) com... Hi, I am trying to install multiple SQL 2005 instances onto one machine, which has been configured as a single node cluster. Multiple virtual servers have been created, and I am trying to install one instance onto each virtual server set up for SQL installation, and the instance should all be default so that when referencing the sql instance only the virtual server's name is needed. Is this possible at all? The scenario: Server box name: ServerA Virtual server names: 1. SQL1 2. SQL2 3. SQL3 I logged into virtual server SQL1, installed a default sql instance (during the intallation the Virtual SQL Server Name page never appeared, an indication that the MSCS was not set up properly?). Then I (successfully, indicating that the cluster/virtual machine set up is done (probably) correctly?) logged in to SQL2, but when trying to install a default sql instance on this virtual server, the intallation recognizes an existing default intance and would not do the next default installation. There are a few reasons that I am not successful so far: 1. Single node cluster would not support this at all 2. The cluster service was not set up right 3. I am doing something wrong I just can not see what I am doing is wrong. Anyone can help pointing out what I should look at? Much thanks. |
![]() |
| Thread Tools | |
| Display Modes | |
| |