The infrastructure also includes a fork-observer connected to each node as well as an addrman-observer for each node. Additionally, detailed Bitcoin Core debug logs are avaliable. The main part are the Grafana dashboards.
There’s a public version at public.peer.observer, which is redacted to not leak honeynode IP addresses. Happy to provide access to an internal version.
Grafana dashboard shown, for example:
in and outbound P2P messages per node
observed drop in inbound version coming in when linking lion was down
current inbound connections
block connection duration
transaction rejected from mempool
More automated alerting and anomaly detection is a todo.
Transcripts
Community-maintained archive to unlocking knowledge from technical bitcoin transcripts