I have a challenge!!!!!!!!!!!
Requirements are:
A. To replace the 2 nodes of SQL 2005 Cluster on Windows 2003 SP1, with a
set of 2 new nodes (hardware).
B. Keep all the settings of SQL 2005, Databases, .etc. the same way they
are.
C. WITH THE SAME IP ADDRESSES (no change to any IP address).
I know the theory is to implement this KB:
http://msdn2.microsoft.com/en-us/library/ms191545.aspx
My question is:
Can I do the following?
1. Remove the second OLD node, using the same ms191545 above.
2. Add (Join) the 1st New node to the cluster.
3. Remove the last OLD node.
4. Add the NEW second node.
Are the above steps feasible?
What are the things I have to be aware about (so I will not break anything).
How can get thru having the same IP addresses with no change on the 2 new
nodes?
Thanks in advance
wit1 wrote:
> I have a challenge!!!!!!!!!!!
> Requirements are:
> A. To replace the 2 nodes of SQL 2005 Cluster on Windows 2003 SP1, with a
> set of 2 new nodes (hardware).
> B. Keep all the settings of SQL 2005, Databases, .etc. the same way they
> are.
> C. WITH THE SAME IP ADDRESSES (no change to any IP address).
Maybe I'm missing something in your 3rd requirement. When you specify
"same IP address," it implies that your requirement is referring to the
cluster _node_, not a cluster group being hosted by the cluster. Why is
that, and what difference does it make?
As the cluster group is moved from one node to another during failure or
intentional move, there should be an IP address for that group that is
separate and unique from the node(s) that own it.
|||I meant that if node 2 (for example) has IP addresses of
1.2.3.4 for the Virtual IP (Cluster group IP)
2.3.4.5 for public (for the NODE itself not the Virtual IP),
and
1.1.1.1 for private (heart beat)
, the new (coming node) that will replace that node 2 should have the same
IP addresses.
So the plan is to un-install SQL from node 2, then evict it, after that join
the NEW node 2 and install SQL with same IP addresses.
Hope that clears things out.
"horseradish" <millardjk@.gmail.com> wrote in message
news:%237hk1iojHHA.5084@.TK2MSFTNGP03.phx.gbl...[vbcol=seagreen]
> wit1 wrote:
with a[vbcol=seagreen]
they
> Maybe I'm missing something in your 3rd requirement. When you specify
> "same IP address," it implies that your requirement is referring to the
> cluster _node_, not a cluster group being hosted by the cluster. Why is
> that, and what difference does it make?
> As the cluster group is moved from one node to another during failure or
> intentional move, there should be an IP address for that group that is
> separate and unique from the node(s) that own it.
No comments:
Post a Comment