| 1 | ### Understand the alert |
| 2 | |
| 3 | The Netdata Agent monitors your HTTP endpoints. You can specify endpoints the Agent will monitor in the Agent's Go module under `go.d/httpcheck.conf`. |
| 4 | The Agent calculates the average response time for every HTTP request made to the endpoint being monitored per hour. The Agent also calculates the average response time in a 3-min window. |
| 5 | |
| 6 | The Netdata Agent compares these two (average) values. If there is a significant increase in 3-min average, then it will trigger a warning alert when the response time 3-min average is at least twice as much as 1-hour average. The alert will escalate to critical when the response time 3-min average reaches three times the average amount per hour. |
| 7 | |
| 8 | ### Troubleshoot the alert |
| 9 | |
| 10 | To troubleshoot this issue, check for: |
| 11 | |
| 12 | - Network congestion in your system's network and/or in the remote endpoint's network. |
| 13 | - If the endpoint is managed by you, then check the system load. |
| 14 | |
| 15 | ### Useful resources |
| 16 | |
| 17 | 1. [HTTP endpoint monitoring with Netdata](/src/go/plugin/go.d/collector/httpcheck/integrations/http_endpoints.md) |
| 18 |