clawbot
9187cd47b6
check / check (push) Canceled after 0s
The prefix handler ran one PeerDown per peer, one after another, and stopped taking messages at 20,000 held-back messages. Sessions going down together queued up their walks, so a returning full-table peer could reach that limit. The table's PeersDown now takes several peers out of every prefix in one walk. One walk runs at a time. The next one takes every peer whose session went down meanwhile, so a message is held back for about two walks at most. The limit is now 100,000 messages, about 100 MiB. The status page shows how many messages each handler holds back. Model: opus-5-5