Once-per-second "Failed to find node for hop #2 of our path." on dir auth startup, until new consensus is made
Starting somewhere in 0.4.7, we got a new bug where dir authorities log a string of
Sep 07 20:26:00.192 [notice] Failed to find node for hop #2 of our path. Discarding this circuit.
Sep 07 20:26:01.199 [notice] Failed to find node for hop #2 of our path. Discarding this circuit.
Sep 07 20:26:02.200 [notice] Failed to find node for hop #2 of our path. Discarding this circuit.
Sep 07 20:26:03.205 [notice] Failed to find node for hop #2 of our path. Discarding this circuit.
entries every time they restart. The log entries go away once there is a new consensus.
It is not a big deal, but it is still ugly and distracting for dir auth operators.
I believe what is happening is that we're picking our L2 vanguards (proposal 333, #40363 (closed)) before we've loaded our consensus, and then every time we try to make a preemptive circuit, we can't because we don't have any L2 vanguards to choose from.