Question 1

Is NodeBeat custodial?

Accepted Answer

No. Non-custodial and read-only. The agent scrapes metrics endpoints and remote-writes them. It holds no keys and takes no action on your host.

Question 2

Which chains and clients?

Accepted Answer

Ethereum (geth, reth, lighthouse, teku…) and Cosmos/CometBFT chains, detected automatically. Tell us your pair if it's exotic — onboarding takes 15 minutes.

Question 3

What happens if NodeBeat itself goes down?

Accepted Answer

Metrics buffering covers short outages (agent WAL retries). The data plane is independent of the portal: a portal outage never stops an in-flight page.

Question 4

How fast are pages, really?

Accepted Answer

A dead or hung Ethereum execution or consensus client reaches the pager in 5.6–8.6 seconds (9 measured runs, lab node → hosted backend). A client that still answers but stops advancing pages in about a minute; missed duties page as soon as the epoch's duties are checked.

Question 5

Can I self-host instead?

Accepted Answer

The agent, alert rules, and dashboards are open-source (Apache-2.0), so you can audit them or run them against your own Prometheus stack. The hosted backend is not open-source; SaaS buys you monitoring that doesn't go down with your box, maintained rules per chain and client, and not operating it yourself.

Question 6

How do I cancel?

Accepted Answer

Billing → Manage subscription → cancel. Your plan runs to the end of the period you paid for, then moves to Free; nothing is deleted. Changed your mind after paying? Any charge is refundable within 14 days.