Chris trying new fix

Over the past week, there have been issues where some miners have multiple simultaneous connections open to the server, which results in increased CPU load and wasted bandwidth. Chris thinks that the cause of the problem is the same as what caused the old server to go offline when he plugged in the new server - network teaming.

Sometime today, he's going to experiment with disabling teaming on the switch, to confirm the hypothesis that the switch poorly programmed. Hopefully, he'll have this issue resolved by the morning.

Chris identified the issue. It turns out that there are seven specific A4 workers that are establishing up to 134 connections every time there is a work restart.

Fortunately, the problem does not appear currently to be present anywhere else. We contacted the miner who is responsible to work on a solution.

Hopefully this is not a sign of A4 failures that happens over time. Whatever the case, we should have this issue resolved shortly.