id: meta-datacenters
name: Meta data center locations
domain: datacenters.atmeta.com
kind: cloud_provider
mode: html
priority: 1
license: "Publicly published operator pages (Meta Data Centers); facts only"
attribution: "Meta — datacenters.atmeta.com"
homepage: https://datacenters.atmeta.com/all-locations/
coverage: { global: true, operators: ["Meta"] }
implementation: cloud_regions
fetch:
  level: 1
  maxLevel: 2
  rpm: 6
  concurrency: 1
  respectRobots: true
  maxCreditsPerRun: 0
discovery:
  seeds:
    - { url: "https://datacenters.atmeta.com/all-locations/", group: facility_pages, pageType: facility_index }
schedule:
  facility_pages: weekly
extractors:
  sites:
    match: ["/all-locations/?$"]
    kind: facility
    parser: cloud_meta_datacenters_v1
defaults:
  operatorName: Meta
  operatorKind: hyperscaler
  facilityType: hyperscale
  isHyperscale: true
notes: >
  33 sites as server-rendered cards (state/country, town, "$X billion investment", "YYYY break ground", jobs).
  Operating status is not stated: break-ground year in the future → announced, within the last year →
  under_construction, older → unknown (left to reconciliation). Info sheets are PDFs (1–2 MB) and are not fetched.
  robots.txt allows everything; WordPress sitemaps exist for the blog (used by meta-news).