We should come up with at least a high level strategy for monitoring seesaw nodes. Some possibilities: * OpenTelemetry * Prometheus * ??? In any case, we should make it easy for alternative implementations to exist, should users need to plug in their own.