Uptime overview
Use uptime monitoring to track endpoint health, create monitors, inspect failures, and understand availability trends.
Uptime monitoring checks important endpoints and services from the outside. Use it to see whether customers can reach your service, how response times are moving, and when failures opened or closed incidents.
What uptime shows
The Uptime page summarizes monitor status, availability percentage, response time, open incidents, and trends over the selected time range.
- Operational means the latest check passed.
- Degraded means the monitor is reachable but unhealthy or approaching a warning threshold.
- Down means the check failed.
- No data means the monitor has not produced checks in the selected window.
Review monitor health
- Open a project and select Uptime.
- Use the range and environment filters to focus the view.
- Scan the status, uptime percentage, response time, and 90-day timeline for each monitor.
- Open monitors that are down, degraded, or trending toward slower responses.
Create a monitor
Select New monitor to define what Errorgap should check.
- Endpoint: choose HTTPS, HTTP, TCP, or ping, then enter the display name and URL or host.
- Schedule: choose the interval, regions, and timeout.
- Assertions: define expected status codes and optional response text.
- Alerting: choose how many consecutive failures should open an incident and who to notify.
Monitor detail
A monitor detail page shows focused availability, response-time trends, SSL expiry context, recent checks, incidents, and alert rules. Use it when the monitor list shows a problem or when you need to understand a single endpoint's behavior over time.
Incidents and checks
Uptime incidents represent contiguous windows of down or degraded checks. Recent checks show individual check results, which is useful when a monitor flaps or only fails in one region.
Operate monitors
- Pause monitors during planned maintenance when alerts would be expected noise.
- Resume paused monitors as soon as the endpoint should be healthy again.
- Edit assertions when the service's expected response changes.
- Delete monitors only when their history is no longer needed.