What is uptime monitoring?

Checking on a schedule whether a service is reachable and healthy, and alerting the moment it is not. PingDog monitors uptime across HTTP, TCP, DNS, SSL, blockchain RPC, A2A agents and MCP servers.

What is delivered check?

One result handed to you at your chosen cadence. PingDog bills per delivered check — never per internal network probe — at a fixed rate, so your price never depends on how many other customers watch the same target.

What is poll-once-fan-out?

The load-bearing mechanism: when many customers watch the same target, PingDog runs one probe stream and fans each result out to everyone due for it. The target sees a single probe stream regardless of customer count — the fan-out is the anti-abuse.

What is proof of work (visible)?

PingDog makes vigilance visible: every check is counted, timestamped and charted, and the delivered-check count is a verifiable, signed record you can check against our public key — the receipt that a silent-but-working monitor is doing its job.

What is pay-as-you-go?

You pay once, up front, for exactly the checks you configure — no subscription, no auto-renewal, nothing to cancel. A schedule that runs "for 20 weeks" is a shape for the checks, paid in advance, not a recurring charge.

Up to 3 monitors at a 5 minutes interval, run with no payment and no wallet — the same builder in its zero-cost range, not a locked demo. Cross either floor and it becomes a normal paid order.

What is magic-link recovery?

Because there is no password and the key is the account, an optional recovery email lets you request a one-time link back into your dashboard. The email is a delivery address, never an identity.

What is interval floor?

The fastest interval the builder allows — 5 seconds, enforced server-side as an anti-abuse measure. The UI never offers a position the backend would refuse.

What is rpc staleness?

An RPC endpoint that answers requests but whose block height has stopped advancing. PingDog treats this as a failure, because reads succeed against a frozen view of the chain — worse than an endpoint that is plainly down.

A total number of checks shaped by a daily UTC active window, a day-of-week mask, and a span (repeat N weeks or run until a date), all paid once up front. Nothing recurs.

Reference

Glossary

The handful of terms it helps to know. Every one is a plain-language definition of exactly how PingDog behaves — no marketing gloss.

Uptime monitoring: Checking on a schedule whether a service is reachable and healthy, and alerting the moment it is not. PingDog monitors uptime across HTTP, TCP, DNS, SSL, blockchain RPC, A2A agents and MCP servers.
Delivered check: One result handed to you at your chosen cadence. PingDog bills per delivered check — never per internal network probe — at a fixed rate, so your price never depends on how many other customers watch the same target.
Poll-once-fan-out: The load-bearing mechanism: when many customers watch the same target, PingDog runs one probe stream and fans each result out to everyone due for it. The target sees a single probe stream regardless of customer count — the fan-out is the anti-abuse.
Proof of work (visible): PingDog makes vigilance visible: every check is counted, timestamped and charted, and the delivered-check count is a verifiable, signed record you can check against our public key — the receipt that a silent-but-working monitor is doing its job.
Pay-as-you-go: You pay once, up front, for exactly the checks you configure — no subscription, no auto-renewal, nothing to cancel. A schedule that runs "for 20 weeks" is a shape for the checks, paid in advance, not a recurring charge.
Free tier: Up to 3 monitors at a 5 minutes interval, run with no payment and no wallet — the same builder in its zero-cost range, not a locked demo. Cross either floor and it becomes a normal paid order.
Magic-link recovery: Because there is no password and the key is the account, an optional recovery email lets you request a one-time link back into your dashboard. The email is a delivery address, never an identity.
Interval floor: The fastest interval the builder allows — 5 seconds, enforced server-side as an anti-abuse measure. The UI never offers a position the backend would refuse.
RPC staleness: An RPC endpoint that answers requests but whose block height has stopped advancing. PingDog treats this as a failure, because reads succeed against a frozen view of the chain — worse than an endpoint that is plainly down.
Schedule: A total number of checks shaped by a daily UTC active window, a day-of-week mask, and a span (repeat N weeks or run until a date), all paid once up front. Nothing recurs.