RPC health

Blockchain RPC node monitoring

An RPC endpoint that answers but has stopped advancing is worse than one that is down — your app keeps reading stale state and never errors. PingDog calls the chain on your schedule and catches exactly that.

What PingDog checks

  • A valid JSON-RPC response to eth_blockNumber
  • The block height is a well-formed hex number
  • The head is not stale — the block advances between checks, not frozen

Node operators and dapp teams live and die by RPC health, and "up" is not enough. A node can respond to every request while quietly stuck behind the tip of the chain — a syncing pause, a peer partition, a bad upgrade. Reads succeed, but against a frozen view of the world.

PingDog treats staleness as a first-class failure: it compares the block height across checks and flags a head that has stopped moving, so a stuck node pages you instead of silently corrupting your app's view of state. This is one of the checks built specifically for the crypto-infrastructure audience PingDog serves.

Anonymous, pay-per-check

Like everything on PingDog, this check needs no account: your key is the account, you pay once at $0.0005 per delivered check in USDC on Base, and nothing renews. A real free tier runs up to 3 monitors at a 5 minutes interval with no payment at all. See how pricing works.

Frequently asked

What counts as a stale RPC node?

One whose reported block height does not advance between consecutive checks. PingDog remembers the previous height and flags a head that has stopped moving, even when the endpoint is still answering requests.

Which chains work?

Any Ethereum-compatible endpoint that answers eth_blockNumber over JSON-RPC — mainnets, L2s, testnets and private chains alike.

Build this monitor → — the free tier needs no payment, and the price is just your check count times $0.0005.