Not translated yet — showing English.
Public Status Page
The public status endpoint tells anyone whether the PlayServ platform is up, with no credentials. Point an uptime monitor at it, or use it to show a status indicator in your game or on your support page. It reports the platform's own health only: it never exposes project data.
Where the status comes from
The platform reports on itself. Every minute, each platform system writes a short health report, and the endpoint serves the latest verdict per system and point of presence.
| System | What it covers |
|---|---|
admin | The Backoffice and the platform API. |
data | Reading and writing platform data. |
transport | Realtime messaging. |
functions | The function gateway and runtime. It also turns yellow when deployed functions that the platform samples stop answering their GET /_health check. |
New systems may appear in the list over time, so treat it as open rather than fixed.
| Status | Meaning |
|---|---|
green | The system is healthy. |
yellow | The system is degraded: its own check was slow or failed, or it is restarting. |
red | The system stopped reporting. |
Red means silence, not necessarily an outage
A system never reports itself red. Red appears when no report has arrived for about three minutes, because a service that hangs cannot say so. The system may be down, or only its reporting may have stopped; either way, treat it as an incident until it turns green again.
This is a different measurement from Platform Health in the Backoffice, which probes the platform from outside. The two can disagree for a few minutes.
Current status
GET /status returns the verdict right now. Responses are cached for a few seconds, so polling more often than every 30 seconds gains nothing.
| Field | Meaning |
|---|---|
generated_at | When the response was produced. |
overall | The worst status across all items. Absent when there are no items. |
items[] | One entry per system and point of presence. |
items[].system · items[].pop | Which system, at which point of presence. |
items[].current_status | The verdict: green, yellow or red. |
items[].last_status | The status in the latest report, before silence is taken into account. |
items[].last_probe_at · items[].seconds_since_probe | When the latest report arrived, and how long ago. |
No items means no verdict, not healthy
If items is empty, there is no evidence either way, and overall is left out instead of defaulting to green. Show "status unknown", not "all good".
Daily history
GET /status/history returns one row per day for each system, oldest first, for one point of presence.
| Parameter | Values | Default |
|---|---|---|
days | 1–365 | 90 |
pop | A point of presence | The platform's own |
A value of days outside the range is rejected with 422. The per-minute reports behind each day are kept indefinitely, so older history doesn't thin out.
| Field per day | Meaning |
|---|---|
day | The UTC date. |
availability_pct | The share of the day's minutes that were not red, from 0 to 100. |
green_minutes · yellow_minutes · red_minutes | How the day's minutes split by status. They always add up to the minutes elapsed in that day. |
Using it in a game or a monitor
| Use | What to read |
|---|---|
| Uptime monitor | Poll GET /status every 30–60 seconds and alert when overall is red, or when it is missing for several polls in a row. |
| Status indicator in a game | Show green, degraded or down from overall, and "unknown" when it is absent. Don't block players on it: it reports the platform, not your game. |
| Support or status page | Draw the last 30 or 90 days from GET /status/history as daily bars, coloured by availability_pct. |
To be notified instead of polling, turn on health alerts in Notifications.
Next steps
- Platform Health — the detailed, signed-in health view in the Backoffice
- Notifications — Slack alerts when platform health changes
- Functions — the health of your own functions
Last updated on 14 Oct 2026