// Open network Β· service status
The network. At a glance.
Live service health, measured availability and a transparent record of every interruption, straight from the monitor.
Connecting to the monitorβ¦
No service is marked healthy until a real check is received.
- Pending: No recent data
- Pending: No recent data
- Pending: No recent data
- Pending: No recent data
- Pending: No recent data
- Pending: No recent data
- Pending: No recent data
01Availability
Platform services
- Operational
- Slow responses
- Disruption
- No recent data
- Striped = partial coverage
// STATUS.FEED loading...
02Response time
Every millisecond counts.
βAverage Β· last 24 h
βSlowest bucket
0/ 288Buckets measured
View response-time data
| Time | Response time |
|---|
End-to-end check duration, not in-game ping. CDN checks include release metadata and 4 KiB partial downloads. Missing samples are never joined across gaps.
03Maintenance
Heads up.
Schedule unavailable
Waiting for a monitoring snapshot.
// Monitor
- Probe
- β
- Check interval
- β
- Unknown after
- β
- Collecting since
- β
- Retention
- 90 days
04Incident history
The record, not the guesswork.
Last 90 days Β· newest first
// INCIDENTS.LOG offline
History unavailable
Incidents open after two consecutive failed checks and close after two successful checks. Single failures still appear in availability. Detection is not a root-cause diagnosis.
How we measure availability
Measurements come from the configured operator probe. A check runs every minute. This is a single monitoring vantage point, not a global service-level guarantee. Community game servers and end-to-end sign-in are not covered.
Availability = successful checks Γ· completed checks. Slow but successful responses remain available. Coverage = completed checks Γ· scheduled minute slots in the selected UTC calendar days. Unobserved periods, including before monitoring began, are unknown β never counted as uptime. Maintenance does not erase failed checks.
Daily bars show the worst observed result; stripes indicate incomplete coverage. The response graph averages successful checks in five-minute buckets. Data older than three minutes becomes unknown, even if the last result was healthy.
Collection start is not yet known. History is retained for 90 days. A monitor or network outage may prevent observations; incident timestamps do not prove uninterrupted downtime between checks.