Latency alerts on Render
Render has no latency alerts. It charts response times at P50, P75, P90 and P99 for web services on a Pro workspace or higher, but its notifications cover failed builds, deploys and cron jobs, unhealthy services and disk usage, not latency. To be alerted when a Render service gets slow, you need a tool that alerts on them. For what each percentile means, see P50 vs P95 vs P99 latency explained.
Where Render shows latency
- Open the web service in the Render Dashboard.
- Go to Metrics.
- The Response Times chart shows P50, P75, P90 and P99; the Percentile dropdown shows one at a time. A second chart counts requests by status code.
Response latency needs a Pro workspace or higher. Render keeps metrics for 7 days on Hobby, 14 days on Pro and 30 days on Scale and Enterprise.
What Render notifies on
| Event | Render notification |
|---|---|
| Build, deploy or cron job fails | Yes, by email or Slack |
| Service becomes unhealthy | Yes; the recovery notice goes to Slack only |
| Persistent disk passes 80% | Yes |
| Response time above a threshold | No |
| Error rate above a threshold | No |
Which percentile to alert on
Alert on P95 held for several minutes: the dashboard chart has no P95 line, but Render’s metrics stream, which Pro workspaces and up send to an observability provider from the Observability page, exports P50, P95 and P99. P90 can miss a growing tail, and P99 flaps on a quiet service. Pair it with the Render 5xx error rate.