mutley1
MIS
- Jul 24, 2003
- 909
Hi All,
We currently have 12 SQL Databases spread over 4 old Proliant 6500 servers and are looking to implement a failover cluster using 2 DL380 G3 and a SAN. The setup that has been requested is Active / Active, but not for load balancing. It is envisaged that we would have 6 databases on node 1 and 6 databases on node 2, and if node 2 were to fail, it would switch to node 1 (all 12 running through node 1) and vice versa. If node 1 were to fail, it would switch to node 2 with all 12 running through node 2. I have a few questions regarding the set up and hope someone will be able to help me!
The assumptions I have made are that each set of 6 databases will be on different disk segments as the 2 nodes cannot read the same disk at the same time, and if the failover kicked in, it is valid to have the same node reading 2 disks simultaneously. For my example, let’s assume node 1 has 6 databases on disk D and node 2 has 6 databases on disk E.
1. Does the basic scenario seem valid, i.e. having 6 running off each node and then failing over to all running through 1 node?
2. I assume I need to install 2 instances on each node (or 2 on the one node which will copy over to the second). The default for node 1 will obviously point to disk D and the other to disk E?
3. Anything obvious I am missing in the basic though of set up architecture / does this setup seem a bit of a non starter / pointless?
Thanks in advance for any assistance,
Mutley.
We currently have 12 SQL Databases spread over 4 old Proliant 6500 servers and are looking to implement a failover cluster using 2 DL380 G3 and a SAN. The setup that has been requested is Active / Active, but not for load balancing. It is envisaged that we would have 6 databases on node 1 and 6 databases on node 2, and if node 2 were to fail, it would switch to node 1 (all 12 running through node 1) and vice versa. If node 1 were to fail, it would switch to node 2 with all 12 running through node 2. I have a few questions regarding the set up and hope someone will be able to help me!
The assumptions I have made are that each set of 6 databases will be on different disk segments as the 2 nodes cannot read the same disk at the same time, and if the failover kicked in, it is valid to have the same node reading 2 disks simultaneously. For my example, let’s assume node 1 has 6 databases on disk D and node 2 has 6 databases on disk E.
1. Does the basic scenario seem valid, i.e. having 6 running off each node and then failing over to all running through 1 node?
2. I assume I need to install 2 instances on each node (or 2 on the one node which will copy over to the second). The default for node 1 will obviously point to disk D and the other to disk E?
3. Anything obvious I am missing in the basic though of set up architecture / does this setup seem a bit of a non starter / pointless?
Thanks in advance for any assistance,
Mutley.
