add Fly HTTP health checks to detect unresponsive machines (#1214)

audio streaming app plyr.fm

after the 2026-04-02 outage, a machine became unresponsive but Fly
had no way to detect it — no health checks were configured. this left
the API down for ~6 minutes until manual intervention.

adds `[[http_service.checks]]` hitting GET /health every 10s with a
5s timeout and 30s grace period on startup, for both prod and staging.

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>

authored by

nate nowack

Claude Opus 4.6 and committed by

GitHub 1 month ago 1a80617c a5427155

+14

2 changed files

expand all

backend

fly.staging.toml

fly.toml

backend/fly.staging.toml

··· 34 34 hard_limit = 250 35 35 soft_limit = 200 36 36 37 + [[http_service.checks]] 38 + interval = "10s" 39 + timeout = "5s" 40 + grace_period = "30s" 41 + method = "GET" 42 + path = "/health" 43 + 37 44 [[vm]] 38 45 memory = '1gb' 39 46 cpu_kind = 'shared'

backend/fly.toml

··· 21 21 hard_limit = 250 22 22 soft_limit = 200 23 23 24 + [[http_service.checks]] 25 + interval = "10s" 26 + timeout = "5s" 27 + grace_period = "30s" 28 + method = "GET" 29 + path = "/health" 30 + 24 31 [[vm]] 25 32 memory = '1gb' 26 33 cpu_kind = 'shared'

Configure Feed

Configure Feed