I'm not sure when I first saw this bug. I considered perhaps port forward had somehow not kicked in properly but that's something you'd expect from UPnP, etc. instead.
Occasionally, generally happening after I'm restarting Phex from a long session, Phex connects in Ultrapeer mode but is kicking off all leafs.
Peer connections at top.
Attempted Leaf connections below:
Error 503: No Leaf slots.
For some reason Phex is pushing off leafs despite declaring itself as an ultrapeer.
The two attachments below are a combination of snapshots taken over a 30 to 60 second period:
. .
The 'already connected' error messages might relate to WireShare also running. I verified at least a couple of same hosts as giving those error messages.
Possible Bug #2 - Connecting as Leaf with 5 Ultrapeer slots set:
I've found sometimes when I connected as a leaf I connect to more ultrapeers than I have set. It takes about an hour to reduce to 6 and another 30 mins to reduce to 5 ultrapeers: