| Endpoint | Chain | State | Round trip from you | Behind head | Uptime 24h |
|---|---|---|---|---|---|
| rpc.ordofi.networkthe endpoint to use: nearest site, automatic failover | Robinhood Chain | checking… | — | — | — |
| us.ordofi.networkUS site, Chicago, next to the sequencer | Robinhood Chain | checking… | — | — | — |
| eu.ordofi.networkEU site, Frankfurt | Robinhood Chain | checking… | — | — | — |
| arc.ordofi.networkCircle's L1, its own node | Arc | checking… | — | — | — |
| aeva.ordofi.networkAeva's L1 mainnet, our own archive node | AevaChain | checking… | — | — | — |
Round trip from you is a warm eth_blockNumber, the second of two calls, so it excludes the TLS handshake a wallet pays once.
Behind head compares each endpoint's block against the highest any of them reported in the same instant; on a chain that makes ten blocks a second, one or two is normal.
Uptime 24h is each site's /health as seen from the other continent, once a minute, so it is not us grading our own homework.
No incidents recorded. When one happens it is written here, with what broke and what we did, in plain words.
Every figure above comes from one of these. They are open, keyless, and safe to poll.
GET /health | Liveness, chain head, our node's lag, every fresh upstream. 200 when serving, 503 while draining. |
|---|---|
GET /status.json | The machine-readable version of this page for one site: latency percentiles, request and error counts, and the peer record. |
GET /metrics | Prometheus text. Scrape it. |
Any of those works on rpc.ordofi.network, arc.ordofi.network, and on the per-site names above. A terminal that wants to fail over on its own should watch ownNodeLag and freshUpstreams.atHead from /status.json.