Hello, I'm currently rolling out SMSP 8 to a newly architected SharePoint environment. I'm building the HA portion of the SMSP Manager and have a couple of questions.
1. I have configured a Windows NLB cluster and assigned the 2 servers appropriately. SMSP manager has been installed on both, with Control and Media services. Within the SMSP control panel, under Manager Monitor, I notice the status of Control and Media on HostA is down and the mode is Active, HostB both are up and active. The services within Windows are started on both A and B. Is this expected? If I disconnect the NIC on HostB, SMSP becomes unavailable.
2. Does NLB provide fault tolerance/HA or just load balancing? I don't see any documentation in the guide regarding how to architect using a Windows cluster, so I assume this is how to provide FT/HA.
To fix the issue with the services I ended up reinstalling the manager on the problem node. This caused things to work properly within SMSP, however after messing around with NLB for far longer than I should have, I have abandoned this solution.
It says in the document this should provide a level of HA to the SMSP manager. I think it only provides load balancing. When either of the nodes were offline, the Docave manager would hang on login, eventually timing out. When that system came back up, logins worked fine again.
This may have been an issue with NLB convergence, however weighing all variables I simply abandoned HA. Our Sharepoint environment is HA, the SMSP manager doesn't need to be HA given we only have about 4 SMSP admins.