relay-search lists flakey as down for an hour or two, when it's never been reachable
https://metrics.torproject.org/rs.html#details/5481936581E23D2D178105D44DB6915AB06BFB7F is marked as offline, and the "downtime" section varies over time but it always seems to be more than one hour and less than two hours.
That's because the bridge firewalled its ORPort, so it never gets the Running flag from Serge, so bridgedb won't give it out. Makes sense -- it is the snowflake bridge, so it receives its connections via the bridge-side snowflake proxy, not direct to the ORPort.
But metrics seems to have hit an edge case: it does something weird when it has no known "time it was last up" value.
(Not a big deal, but it's been sitting in a tab for some days now and I figured I should file the ticket.)
To upload designs, you'll need to enable LFS and have admin enable hashed storage. More information