Platform Status & Transparency

Trust in a B2B/B2G platform survives outages only when the operator is transparent during them. The status page solves this by going beyond the typical green-square theatre. A background job, CollectStatusSamplesJob, runs every 60 seconds and probes each component the platform depends on: API health endpoint, MongoDB ping, Redis ping, Auth service, the static web apps.

Each probe writes a ServiceHealthSample document with status, latency and timestamp into a rolling-window collection with a 90-day TTL — long enough for SLA reporting, short enough to stay cheap. The visible /status page reads aggregated samples and renders per-service uptime over the last 24 hours, 7 days and 90 days, plus current state. Operators can publish StatusIncident records from the sys console with a title, severity, affected services and a timeline of updates — these render at the top of the status page during active incidents and below the fold afterwards as historical post-mortems.

The client-side fetcher refreshes every 30 seconds without a page reload so visitors always see fresh state, and — critically — if the fetch times out the UI fails red ('We can't reach the status service either') rather than silently showing stale green. The page is fully translated into EN/FR/ES/SV and is linked from the footer of every marketing page and from inside the admin app, so any user, customer or prospect who suspects an outage gets a same-second answer without escalating to support.