Anycast: design in São Paulo (GRU) · A-61

Runbook A-61 for Anycast in São Paulo (GRU). Marker xserv-a-56f4e5035322. This page covers design for LATAM operators, not a worldwide average.

Anycast: Why it matters on this continent

Design for Anycast on letter A starts in São Paulo (GRU). Runbook a-01 keeps a single owner, a written rollback, and a traffic split that can be reversed without a global freeze. Neighbors in Santiago only take overflow after the local pool fails a health window.

Anycast: How XServ runs it day to day

Probes for Anycast (a-02) leave Santiago every 15s toward Bogotá. XServ pages the named owner if loss or RTT crosses the letter budget. A probe never shares a queue with bulk transfers, so a saturated WAN does not hide a dead PoP.

Anycast: What operators should measure

Policy for Anycast is versioned as a-03. Operators measure error rate, p95 from Bogotá, and time-to-rollback — not a worldwide average. Changes land in Bogotá first, then Miami, with a hold if either region regresses.

Anycast cache A-08

Cache rules for Anycast on a-08 keep language-correct objects near Miami. São Paulo is a sibling cache, not an origin. TTLs are short enough that a bad asset does not live through a weekend.

Anycast headers A-09

Headers for a-09 carry the letter, the region (GRU) and a request id. Anycast debugging in São Paulo should not require a packet capture in Santiago first.

Anycast timeouts A-10

Timeouts on a-10 are tighter than the WAN RTT to Bogotá. Anycast in Santiago fails fast and retries once; a third try needs a human because it is no longer a blip.

Anycast regions A-11

Region names on a-11 match DNS and tickets: Bogotá / BOG, with Miami as the documented pair. Anycast dashboards never say “LATAM” as if it were one RTT.