Re: Failover Clustering Virtual IP Address
- From: "Linchi Shea" <linchi_shea@xxxxxxxxxxx>
- Date: Sun, 26 Feb 2006 00:18:14 -0500
When you install a SQL2000 instance in a cluster, the disk resource
group--that you'll have to select during the install--should not have the IP
resource and the network name that your instance will use.
In some envrionments, it's common for the server admin folks to create an IP
recource and a network name resource when they give the server to the DBA to
install SQL Server, and that IP address and that network name are supposedly
for the SQL instance to use. This is not a bad practice because it
effectively reserves the IP address and the network name. But since SQL
Server Setup will actually create the IP resource and the network name
resource for the SQL instance, so if you provide the same IP address during
the setup, the install will complain. If this is the case, just delete the
network name resource and the IP resource after you have recorded the IP and
network name, and then re-do the install. When asked for IP address, type in
the address you have recorded.
You definitely should not just pick any unused IP address. Most places do
manage their IP addresses.
Linchi
"sqlclusterhelp" <misty.pratt@xxxxxxxxx> wrote in message
news:1140926061.794221.240270@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
Hi there,
I am installing SQL 2000 on a Microsoft Windows Server 2003 clustered
environment. This is a two node active/passive setup. The servers and
OS were setup/installed and are being housed by an external company.
The setup is as follows:
Groups - Cluster Group, GRP1, GRP2 (different names of course). GRP1
and GRP2 then have a number of resources inside of them ie. GRP1 IP
Address, GRP1 Network Name, and different disks.
I have been provided with the public/private IP addresses of the
machines, nodes, groups etc. When I install SQL and get to the step
"Failover Clustering" I am a little uncertain what IP Address I should
be using. Obviously, when I try using the Public IP address of say
GRP1 (which contains the disk for where I want my first SQL instance to
reside) it returns "This already exists as GRP1 resource"...and the
same for all other IP Addresses that I know of.
I realise I then need to define a new "Virtual IP Address" that is not
currently being used by any other groups/resources. The question I
have is - Should this new Virtual IP Address be provided by the company
housing the servers or can I just pick any unused IP Address to use for
this purpose? Ie. if GRP1's public IP is 192.150.20.100 can I just
define 192.150.20.101 as the Virtual IP Address to use?
When I take this approach and just select a new IP Address it continues
to the next step... although I am not going to follow through with the
install until I have clarified which IP Address I should use.
Thanks.
.
- References:
- Failover Clustering Virtual IP Address
- From: sqlclusterhelp
- Failover Clustering Virtual IP Address
- Prev by Date: Failover Clustering Virtual IP Address
- Next by Date: Re: adding a new sql server node to an existing sql cluster
- Previous by thread: Failover Clustering Virtual IP Address
- Next by thread: Re: adding a new sql server node to an existing sql cluster
- Index(es):
Relevant Pages
|
Loading
