I would like to ask what the “remediation” for the recent performance issues was. What was wrong? How did you fix that? (I didn’t see any post on the forum about that)
We use TCP when talking to Unbound internally, and exhausted its incoming-num-tcp. We increased that config value.