Please start any new threads on our new
site at https://forums.sqlteam.com. We've got lots of great SQL Server
experts to answer whatever question you can come up with.
| Author |
Topic |
|
doug-uvmmed
Starting Member
1 Post |
 Posted - 2012-03-12 : 15:49:24
|
| - Windows 2008R2 SP1, 64-bit, 3 node, node majority, cluster, currently patched- Dell M610 blades, dual quad processors, 48GB RAM each- 6 nics, 5 in use: 1 Public (1G), 2 SAN (MPIO, 10G each), 1 Private (Cluster, 1G)- Shared disks on Dell EqualLogic iSCSI SAN, current firmware, current Dell HIT (MPIO) and current MS MPIO DSM- All nics set up properly for iSCSI (jumbo frames, flow control, receive-side scaling, etc.)- SQL 2005 SP4, 64-bit, Enterprise, currently patched- 4 instances, 2 on each of 2 nodes, third node completely passive as failover for either of the other two- Cluster Validation shows cleanly, except for SAN nics which are on the same subnet- No cluster trafic enabled on the SAN nics, switches are isolated to SAN VLAN only- Cluster is set up for any node to accept any service or resource- Problem: •All instances do not appear as SQL Server service on all nodes, therefore failover to any node cannot/does not work•Not consistent: some instances appera on some nodes and not on others ... they are not always the same instance•For example:Node 1 •Instance A •Instance B Node 2 •Instance A •Instance B •Instance D Node 3 •Instance A •Instance B •Instance C •Instance D Note: Preferred nodes are set to all 3 on all instancesAny ideas from any of you what's causing this? And, ideas about how to get all instances on all nodes? The only thing I can think of is to: 1.Evict Node 1 (a quorum disk will have to be put in place and Disk Majority set)2.Then re-add Node 13.Evict Node 24.Then re-add Node 25.Remove quorum and reset to Node Majority This feels a bit draconian, and this is a production cluster that I have very little leeway in bring down for any length of time. Ideas? Thx... |
|
|
|
|
|
|
|
Windows 2008R2 Cluster SQL 2005 Instances Missing