Alerts & Notifications
Alerts watch the metrics flowing through your event system and notify your team the moment one crosses a threshold you define.
Inbox
Active alerts are listed with their state, the rule that fired, the condition, the observed value versus threshold, and how long it's been active. If something critical is firing, a callout at the top lets you jump straight to investigating the affected events, acknowledge it, or mute it for 15 minutes, 1 hour, 4 hours, or 24 hours. Resolving an alert requires the admin/operator role.
Summary stats across the top show how many alerts are firing or pending, how many are critical, and your team's mean time to acknowledge (MTTA) and mean time to resolve (MTTR).
Rules
Each rule lists its name, condition, severity, and whether it's enabled, with quick edit/delete actions. To create one, click New alert rule and set:
| Field | What it controls |
|---|---|
| Metric | DLQ depth or inflow rate, error/duplicate/retry rate, publish-failure rate, p95 latency, or throughput |
| Operator & threshold | e.g. "greater than 5%" |
| Measured over | the window the metric is evaluated across (1 minute to 1 hour) |
| Hold before firing | how long the condition must stay true before the alert actually fires, to avoid noise from brief blips |
| Scope | limit the rule to a specific consumer, broker, or event type |
| Severity | critical, warning, or info |
| Notify channels | which channels get notified when it fires |
While you're setting it up, a live preview shows the metric's current value and whether the rule would fire right now against real traffic.
History
Resolved alerts are kept here with their severity, when they fired, when they resolved, and how long they were open. Useful for spotting recurring problems.
Channels
Add a channel by name, kind (Slack, webhook, or log), and target, and choose which severities route to it. Use Test to send a sample notification, and check Recent deliveries to confirm alerts are actually reaching the channel.