Downloads
The data behind the site as CSV files, refreshed with every update (every 6 hours). Exchange and network data comes from PeeringDB, so its acceptable use policy applies; member lists come from the exchanges themselves; city populations are from GeoNames (CC BY 4.0).
| exchanges.csv | Every exchange: id (PeeringDB's, or 9000001 and up for exchanges added by hand), name, location, peers (the highest of the three counts), IX-F, website and PeeringDB member counts, the count it states, mismatches, ghost flag, 10G price, route server sessions from its public looking glass (blank when there's none to read), and whether it's a Virtual IX with where that comes from (export, declared or override). | 1,335 |
| fees.csv | Port fees as each exchange publishes them on its website, read automatically: check them on the exchange's page. | 618 |
| networks.csv | Every network seen at an exchange: ASN, name, and the PeeringDB ids of its exchanges. | 23,584 |
| facilities.csv | Every facility with an exchange: exchanges, total peers (added up), unique networks reachable through them, networks listed in the building, and how many of those don't peer on its exchanges. | 2,169 |
| ghosts.csv | Ghost exchanges: why, and since when (already_when_tracking_began = 1: it was a ghost before the first update, so 'since' is that update's date). | 251 |
| metro-gaps.csv | Metro areas of 500,000+ people with no exchange (or only ghosts), with the nearest real exchange. | 939 |
| address-checks.csv | Every network whose IP addresses or port speed in PeeringDB differ from its exchange's IX-F export. | 8,457 |
| ixf-exports.csv | Every exchange's IX-F export: whether it could be read on the last update, what went wrong, since when, what the export says about itself, and whether PeeringDB's IX-F import is on for the exchange. | 466 |
| changes.csv | Every change ever recorded, newest first, with its source: PDB (PeeringDB), IXF (the IX-F list), WEB (the exchange's website). | 5,509 |
Each exchange page also has its own fix list as CSV (networks to chase between what the exchange publishes and PeeringDB), and its address and port speed differences.
Read-only API
The same data as JSON, one file per exchange and per network: plain static files, rebuilt with every update (every 6 hours), with no key or sign-up. They may be cached for 5 minutes, and other websites can load them straight from a browser (CORS is open). Please fetch only what you need, and no more often than the data changes.
| /api/index.json | Every exchange: id, name, city, country, peers, ghost flag, and the addresses of its file and page. |
| /api/ix/<id>.json | One exchange, by its PeeringDB id (9000001 and up for exchanges added by hand):
its counts (peers, PeeringDB, IX-F export, website list, the count it states), the share of its own list that PeeringDB has,
ghost flag, peak traffic, 10G price, route server sessions, how each source fared on this update
(data_status), and every member with the sources that list it, its addresses, port speed and route server routes. |
| /api/net/<asn>.json | One network: its PeeringDB record's type, policy and organization, every exchange it's on with the same per-exchange details plus how PeeringDB and the exchange's list compare (comparison), and the buildings it lists in PeeringDB with the exchanges in each. |
| /ix/<id>/badge.svg | The exchange's badge (each exchange page has the code to embed it). |
Every file has "schema": 1 and the time it was built. New fields may be added at any time; the schema number goes up only
when a field is removed or changes meaning. A source that couldn't be read on the last update is null, with the reason
in data_status: these files never carry an old copy. A network or exchange that's gone returns 404. PeeringDB's
acceptable use policy applies to the data that comes from it.