Alert channels
CronRadar sends alerts by email, the Slack app, webhook, PagerDuty, and Discord. Every plan includes every channel. Configure channels in the console; there is nothing to change in your code.
Where channels are configured
Channels belong to a team or to an application:
- Team-managed notifications. When a team manages notifications, every application in the team uses the team's channels. Configure them under Teams → your team → Notifications. The application's Notifications tab then links to the team settings.
- Application-managed notifications. Otherwise each application keeps its own channels under Applications → your application → Notifications.
Changing channels, or switching between team and application management, requires the Admin or Owner role. Saved webhook URLs and keys are never shown again; the console only shows that a channel is configured, and you can enter a replacement.
Choosing destinations
Each job, application, and team can name the destinations its alerts use: email, slack, webhook, pagerduty, or discord. CronRadar uses the job's list first, then the application's, then the team's. When none is set, alerts go to every configured channel. A channel sends only when it is both a destination and configured.
Email is on by default (Enable Email Alerts). Alerts go to every active team member with an email address, whatever their role, except members who turned off alert email for that application in their notification preferences.
Emails come from CronRadar Alerts. The subject is [CronRadar] <application>: <job> and stays the same for an incident, and reminders and the recovery reply to the first email, so one incident reads as one thread. Each email links to the incident in the console.
Slack
CronRadar connects to Slack as an app:
- Open the Notifications settings where the channels are configured (the team's for team-managed notifications, otherwise the application's) and choose Connect Slack. Slack asks you to approve the CronRadar app in a popup; it requests the
chat:writeandchannels:readscopes. - In Slack, invite the bot to a public channel with
/invite @CronRadar. - Back in CronRadar, choose Refresh channels, pick the channel, and save. Only public, non-archived channels the bot has joined are listed; private channels are not supported.
- Use Send test message to confirm the bot can post.
Alerts appear as a message with the application, job, schedule, last check-in, owner, severity, environment, why CronRadar alerted, recent activity, and a View Dashboard button. The team and each application have their own connection, and reconnecting replaces that connection. A legacy Slack incoming-webhook URL can only be set through the API; when the app is connected, the app is used instead.
Webhook
- Turn on Enable Webhook Alerts.
- Enter the
https://URL that should receive alerts and save.
CronRadar sends a POST with a JSON body for every alert. Any 2xx response counts as delivered. Requests carry no signature, so use a hard-to-guess URL and treat the payload as untrusted input. See Webhook payload for every field and a sample.
PagerDuty
- In PagerDuty, add an Events API v2 integration to the service that should page, and copy its integration key.
- In CronRadar, turn on PagerDuty, paste the key as Integration Key, and save.
CronRadar triggers a PagerDuty event for each failure alert and resolves it when the job recovers. A Vanished alert also triggers an event, which nothing resolves later: resolve it in PagerDuty once you have confirmed whether the removal was intended. Events for one job share the dedup key cronradar-<job id>, so reminders update the open incident and a recovery resolves it. A recovery is sent with severity info; other events use the job's configured severity (warning, error, or critical).
A test alert sent to PagerDuty opens a real incident with its own dedup key. CronRadar never resolves it, so resolve it in PagerDuty after the test.
Discord
- In Discord, open the channel's Integrations settings, create a webhook, and copy its URL (
https://discord.com/api/webhooks/...). - In CronRadar, turn on Discord, paste the URL as Discord Webhook URL, and save.
Alerts post as a message with an embed colored red for failures, orange for warnings, and green for recoveries, plus the same details as Slack and a View Dashboard link.
What each channel receives
Every configured destination receives every alert type:
| Alert type | Sent when |
|---|---|
InitialFailure | A job misses its run, reports a failure, or starts and does not finish in time |
ContinuedFailure | Daily reminder while an incident stays open and unacknowledged |
Recovered | The job reports success again |
NeverExecuted | A new job never sends its first report |
OutcomeFailure | A run completes but breaks an outcome expectation |
Vanished | Jobs leave their framework registry; one alert per sync, however many jobs left |
Acknowledging an incident in the console stops its reminders. Paused jobs send no reminders, and Vanished has none. See Missed, vanished, and never-ran jobs for when each alert is raised.
Delivery and retries
CronRadar delivers pending alerts every minute. Each channel request times out after 10 seconds. When a channel fails, CronRadar retries that channel up to three more times, at least five minutes apart; channels that already succeeded are not sent again. Because a retry after an interruption can repeat a delivery, webhook requests carry an Idempotency-Key header and Slack and Discord messages show the alert ID.
Test your channels
Choose Send test alert on the application's Notifications tab, or send one from a job's page. The test uses an example InitialFailure for one of the application's jobs, so the application needs at least one job. Test alerts are marked as tests ([TEST] in email and PagerDuty, "test": true in webhooks) and are not saved to alert history.