fmadec: is this needed to be fixed soon? And when?

Bump for NIRs. For a single cryostat, a status command requires 14 controller commands spanning 1.5s or so. With three arms independently chattering that way the current command failure rate will get ugly. I have not been able to come up with a better/simpler alternative than adding an ion pump actor.

Bump. With three cameras this is becoming a problem.

Specifically, from 2023-04-11 15:46Z to 16:13Z, at 15s sampling, we got no good readings.

I decided that embracing the unavoidable contention should be done first: blocking the single TCP connection for all of the ~10 status commands, and expecting the first connection to have to wait for ~2s or so. Running at JHU, and seems to work.

Tagged 2.2.11