sneak
6b02cee68b
check / check (push) Canceled after 0s
The table's PeerDown visits every prefix under the write lock, which takes over a second on a full table. The prefix handler called it inline, so its queue filled and the streamer dropped updates. PeerDown now lets the lock go after every 10,000 prefixes. The handler runs it on a goroutine of its own, one peer at a time, and holds back the down peer's later messages until it returns, then applies them in order. Stop waits for this before its last flush, and logs the held-back messages it could not apply. Model: opus-5-5