Hello Team,
I have configured fail-over clustering on two virtual nodes having Windows 2012 R2 Operating system. On windows cluster i have configured SQL server SQL Server Always-On Availability groups. SQL server has 3 instances which are distributed on both servers. The cluster owner is holding two instance & passive node is holding single instance. File witness share is configured on domain controller.
Recently we faced an issue with one of the the database instance which is hosted on passive node & it went to hung state. We checked cluster events & found that during issue FSW was unavailable that lead this issue but same time the we don't find issue with owner node who is holding two database instance. Same issue had been re-occurred when we rebooted Domain Controller for Patching activity.
We have distributed these database instances on both nodes because we want to utilize both nodes as having less memory on virtual machine i.e. 8 GB RAM each node.
If i keep all the database instances on the owner node & FSW is unavailable due to any reason all is working fine.
Herewith i am little bit confused to keep database instance on single node or distributed on both node due to RAM constraint.
What will be best practice with respect to Windows Fail-over clustering ? what settings to be done at cluster side to keep alive my database instance even my FSW is unavailable ?
Please suggest.
Regards,
Vinod