3 commits

Author	SHA1	Message	Date
Pierre-Olivier Mercier	6b96ee8c2f	server: expose runtime metrics on /health for scheduler routing ... Adds HealthResponse carrying inflight count, total requests, 1/5/15-min EWMA load averages, uptime, and NumCPU so a scheduler can pick the least busy worker. A background sampler updates the load averages every 5s, stopped by a new idempotent Close method. Work endpoints (/collect, /evaluate, /report) are wrapped with a trackWork middleware; /health and /definition are excluded so polling traffic does not pollute the signal.	2026-04-16 16:52:00 +07:00
Pierre-Olivier Mercier	2fa44f69a4	server: return 500 status on collect errors instead of 200 ... Errors from provider.Collect() and json.Marshal were returned with HTTP 200, making failures invisible to monitoring, proxies, and clients that check status codes. Return 500 Internal Server Error so HTTP-level tooling can detect failures without parsing the response body.	2026-04-10 16:43:56 +07:00
Pierre-Olivier Mercier	ef7fffd4b7	tests: add coverage for options, types, and HTTP server ... - options_test.go: GetOption, GetFloatOption, GetIntOption, GetBoolOption with native types, JSON round-trips, missing keys, and wrong types - types_test.go: Status JSON marshal/unmarshal (strings, legacy ints, round-trip, unknown values), CheckTarget.Scope/String, BuildRulesInfo, empty-ID rejection - server_test.go: /health, /collect (success, error, bad body), /definition, /evaluate (all rules, disabled rule), /report (HTML, metrics, bad body), missing endpoints without CheckerDefinitionProvider	2026-04-10 16:35:50 +07:00