Hi Dave,
I just did an experiment. I opened up a RAM usage graph supplied by SiteGround before logging into WF Central. I focused on our one server that has the most sites (83). These sites get very little traffic. As an illustration, this is how that graph peaked when I logged in: https://take.ms/XhrJt
For reference, this is what that server’s load has looked like for the past few days: https://take.ms/jGxkf . In other words, this peak is pretty significant and was caused when logging into WF Central. In the first illustration, note how after the peak the server returns to nearly normal levels. I stayed logged into WF Central during most of the duration after the peak on the above graph. While in WF Central I sorted by the “critical” option and then did some minor updates across 10 websites. I also removed a couple of websites that were not connecting and added another site. In other words, it seems to me that things work well with all but the initial login which is adding a significant resource usage.
As an analogy, I’m thinking about the Windows operating system. A few generations ago, Windows wanted to ensure they could advertise a fast boot time. One way that they achieved this marketing goal was to delay startup apps over the first few minutes while the machine was booting up. This kept the initial load lower and greatly decreased the start up times.
Can the same kind of principal be applied to WF Central? Maybe batch these initial calls over a longer period of time?