Alerts

Alerts notify you when things go wrong with your crons or queues — and when they recover. Recuro supports three notification channels and four alert types.

Alert types

Type	Description	When it fires
failure	A cron or queue has failed beyond the configured threshold	Cron: consecutive failures reach alert_threshold. Queue: job fails after all retries exhausted
warning	First sign of trouble	Cron: first failure in a sequence. Queue: first job failure before retries
recovery	Previously failing resource is healthy again	Cron succeeds after being in an alerted state
limit_warning	Monthly usage limit approaching	Team exceeds the soft limit for their plan

Alert lifecycle

Cron alerts

1. A cron execution fails — consecutive_failures increments by 1
2. If this is the first failure, a warning alert is created
3. If consecutive_failures reaches alert_threshold and the cron is not already alerted, the cron is marked as alerted and a failure alert is sent to your notification channel
4. When the cron succeeds again, consecutive_failures resets to 0, the alerted flag is cleared, and a recovery alert is sent

Queue alerts

When a queue has alerts_enabled: true:

1. A job fails for the first time — a warning alert is created
2. The job enters the retry flow (if retries are enabled)
3. If all retries are exhausted and the job is dead-lettered, a failure alert is sent

Queue alerts do not have a recovery mechanism since jobs are one-off.

Notification channels

Configure your notification channel in Settings > Notifications:

Channel	Description
email	Sends alert emails to all team members
slack	Posts to a Slack webhook URL
discord	Posts to a Discord webhook URL

For Slack and Discord, provide a webhook URL in the notification settings. You can test the webhook from the settings page.

Alert data

Each alert record contains:

Field	Description
type	failure, warning, recovery, or limit_warning
alertable_type	The resource type (cron or queue)
alertable_id	The resource ID
failure_reason	The categorized failure reason (see Runs)
message	Human-readable description
is_read	Whether the alert has been dismissed

Managing alerts

From the Alerts page in the dashboard:

• Filter alerts by type or failure reason
• Mark individual alerts as read
• Mark all alerts as read
• Delete individual alerts

Maintenance windows

To suppress alerts during planned downtime, create a maintenance window from Maintenance Windows in the dashboard. Specify:

• Which resource to suppress (a specific cron or queue, or all resources)
• A reason (optional)
• Start and end time

Alerts are suppressed for the specified resource during the maintenance window.

Next steps

• Setting Up Alerts — Configure alert thresholds and notification channels
• Cron Schedules — Alert threshold configuration
• Queues — Queue-level alert settings
• Alert Configuration Best Practices — Avoid alert fatigue