TOOLS
Open-source operator tooling by BeeHive. Built for Monad node operators, cross-referenced against docs.monad.xyz/node-ops for every check and config defaults aligned with the latest Foundation announcements.
🔧
Public RPC Catalog
Live latency board for 19 public Monad RPCs
Pings 11 mainnet + 8 testnet endpoints every 60s, ranks by median latency, exposes 'Add to MetaMask' per network. Foundation, Alchemy, Ankr, dRPC, OnFinality, Tenderly, thirdweb, bloXroute, Tatum, MonadInfra, Natsai.
OPEN →
🩺
Operator Scripts
Pre-flight check + host setup — single-file bash
Two single-file bash scripts that take a fresh Ubuntu 24.04 box to a fully-configured Monad validator: monad-doctor (50+ readiness checks across hardware, OS, network, security, monad-specific config) and monad-validator-setup (13 idempotent install steps, --network=testnet|mainnet, --node-type=validator|full, --with-monitoring). Cross-referenced against docs.monad.xyz.
OPEN →
📊
Monitoring Stack
NEWSelf-hosted Grafana + Prometheus for your Monad node
Three Docker containers (Prometheus + Grafana + RPC exporter) plus node-exporter that scrape Monad's bundled OpenTelemetry collector on :8889. Renders a 47-panel dashboard: sync, vote delay, consensus events, system resources, disk, network, txpool, raptorcast, errors. Auto-installer handles UFW, chrony, hostmetrics overlay, healthcheck. Loopback-only by default; --public flag exposes Grafana :3000.
OPEN →
WHY WE PUBLISH THIS
We run Monad validator (testnet) and built these tools for ourselves first. They encode the lessons we learned the hard way: a kernel in the buggy 6.8.0-{56..59} range causing client hangs, NVMe stuck on 4096-byte LBA tanking triedb performance, vm.swappiness=60 inflating p99 vote_delay metrics. Open-sourcing them is the cheapest way to help other operators avoid the same traps — and it lets the community audit our work.
We run Monad validator (testnet) and built these tools for ourselves first. They encode the lessons we learned the hard way: a kernel in the buggy 6.8.0-{56..59} range causing client hangs, NVMe stuck on 4096-byte LBA tanking triedb performance, vm.swappiness=60 inflating p99 vote_delay metrics. Open-sourcing them is the cheapest way to help other operators avoid the same traps — and it lets the community audit our work.