The problem is a lot of nodes are routed to it as not many are online in general much less since 0.19.5. I had 3 out of 4 nodes go down and freeze since 0.19.5 that I am trying to figure out why. That means all their workload is also going to the one that is remaining up.

Two should be back online in a few hours and will help lower the load. It doesn't help that it takes 1-3 days to bring a node back online when it goes down and it costs $650-2,000 a month to run a single full node.

The node is currently processing about 40TB/month.