Bluesky: requestCrawl heartbeat

[snarfed]

Bluesky had some network changes and outage(s) recently, which resulted in a could his-long periods where their relay stopped reconnecting to our subscribeRepos, most recently 7/11 1:17-10:45am PT. It happily reconnected once I sent a requestCrawl.

Jaz recommends we send requestCrawl regularly, either if we noticed that a relay isn't connected or all the time: https://discord.com/channels/1097580399187738645/1097580399187738648/1261023013902749779

this was due to a network outage, the relay gives up after attempting to reconnect like 5 times or smth
We've been dealing with getting our dual-homed IPs working this week and it's been a bit messy
FWIW our PDS distribution does a heartbeat requestCrawl whenvever it has activity (with a 30 second cooldown)
I would probably recommend doing the same, the relay won't infinitely retry connecting to your PDS
It does try on a restart tho