Monitor

The Monitor is a continuous service that performs comprehensive health checks on validated relays using NIP-11 and NIP-66 protocols. It is the most complex service in BigBrotr, producing multiple metadata types per relay per cycle.

Purpose

Knowing that a relay exists is not enough. The Monitor answers: How healthy is this relay right now? It collects relay information documents (NIP-11), runs six different health tests (NIP-66), stores results as content-addressed metadata, and optionally publishes monitoring data as Nostr events.

How It Works

For each validated relay in the current batch:

NIP-11: Relay Information

Fetches the NIP-11 Relay Information Document via HTTP(S). This JSON document describes the relay’s capabilities, limitations, and software:

• Relay name and description
• Software name and version
• Supported NIPs list
• Relay limitations (max message length, max subscriptions, etc.)
• Payment/admission policies

NIP-66: Health Tests

Runs six independent health tests:

Test	What It Measures	Networks
RTT	WebSocket round-trip time — 3-phase: open, read, write (ms)	All
SSL	Certificate validity, expiration date, issuer, SANs, cipher	Clearnet only
DNS	A/AAAA/CNAME/NS/PTR records, TTL	Clearnet only
Geo	Country, city, coordinates, timezone, geohash (GeoLite2-City)	Clearnet only
Net	AS number, ISP name, network ranges (GeoLite2-ASN)	Clearnet only
HTTP	Server header, X-Powered-By	All

Result Storage

Each test produces a Metadata object with a specific MetadataType. Metadata is content-addressed: the SHA-256 hash of the canonical JSON data field becomes the id. Same data always produces the same id, providing automatic deduplication.

Results are stored via relay_metadata_insert_cascade, which atomically inserts across relay, metadata, and relay_metadata tables.

Event Publishing

When configured with Nostr keys, the Monitor publishes monitoring results as Nostr events:

• Kind 0 — profile metadata (Monitor identity)
• Kind 10166 — replaceable relay monitoring announcement (capabilities, networks)
• Kind 30166 — parameterized replaceable relay discovery event (per-relay health data)

These events are broadcast to configured relays, allowing other Nostr clients to consume health data.

Configuration

config/services/monitor.yaml

interval: 600  # seconds between cycles

processing:
  chunk_size: 200
  allow_insecure: false
  compute:
    nip11_info: true
    nip66_rtt: true
    nip66_ssl: true
    nip66_geo: true
    nip66_net: true
    nip66_dns: true
    nip66_http: true

networks:
  clearnet:
    timeout: 30
    max_tasks: 100
  tor:
    enabled: true
    timeout: 90
    max_tasks: 10
    proxy_url: socks5://tor:9050

geo:
  city_database_path: static/GeoLite2-City.mmdb
  asn_database_path: static/GeoLite2-ASN.mmdb
  geohash_precision: 9

Usage

# Run continuously
python -m bigbrotr monitor

# Run a single monitoring cycle
python -m bigbrotr monitor --once

# With debug logging for troubleshooting
python -m bigbrotr monitor --log-level DEBUG

Metrics

The Monitor exposes Prometheus metrics on port 8003:

• service_counter{name="relays_checked"} — total health checks performed
• service_counter{name="nip11_success"} — successful NIP-11 fetches
• service_counter{name="events_published"} — monitoring events broadcast
• cycle_duration_seconds — time per monitoring cycle