Pull cleanup round 900
Check pods longer when down vs check dead ones randomly via cron, once dead dead then gone gone. Podmin could always undelete them.
Use a function for is cli? I put in boot.php since it is used all over, is that bad form?
Fix checking dead pods that do not send back well-known or nodeinfo as json but html that barfs things up
Fix DNS for a pod's A being 127.0.0.1 (I could not see a easy way or package to say external IP). I suspect an A of 10.10.10.10 would be about the same.
Offer podmins some data on why in the email they get.