Docs / Monitoring & alerts
Create an alert
3 minute read
Servers, apps and databases are checked every minute. An alert rule watches one metric on one resource and tells Slack, Discord, email or a webhook when it crosses a line, and again when it recovers. Monitoring and alerts are free.
1. Pick a resource and a metric
Open Monitoring & alerts and choose Create alert. Pick the resource, then the metric to watch.
| Metric | Applies to | Example rule |
|---|---|---|
| CPU | Servers, apps, databases | CPU above 80% for 5 minutes |
| Memory | Servers, apps, databases | Memory above 90% for 10 minutes |
| Disk | Servers, databases | Disk above 80% |
| Outbound traffic | Servers | Outbound traffic above 5 MB/s for 30 minutes |
| Deploy failed | Apps | Any failed deploy |
| Backup failed | Servers, databases | Nightly backup did not finish |
2. Set the threshold and duration
The duration stops short spikes from waking anyone up: the value has to stay over the line for the whole window before the alert fires. Start loose and tighten once you know what normal looks like on the graphs.
- Short windows (1–5 minutes) for things users notice straight away, like CPU on a busy API.
- Longer windows (10–30 minutes) for slow trends, like memory or traffic.
- No window for events such as Deploy failed.
3. Choose where it goes
Pick one or more channels. Slack, Discord and webhooks are set up once under Where alerts go and then reused by every rule; email goes to any address.
simplicloud alerts create \--resource reports-runner --metric cpu --above 80 --for 5m \--notify slack:#ops --notify email:[email protected]
4. Send a test and watch the state
Choose Send test next to the rule to check the message arrives. In the Alert policies table each rule shows its state and current value.
| State | Meaning |
|---|---|
| OK | The value is inside the line. |
| Alerting | The rule fired; you were notified. You hear again when it recovers. |
| Paused | Not evaluated. Use it during maintenance; choose Resume after. |