Alerts
Overview
Alerts deliver failure alerts for scheduled tasks directly to the people responsible. The system listens for Flow/task lifecycle events, generates alerts according to rules, and delivers them through channels such as Webhook.
Alerts apply to Flow failure, task failure, and run timeout scenarios. They're mainly received as messages, with the alert details included in the message body.
Feature entry
Entry path:
Setting → my subscription
Dev → Flow list → Subscribe in a row / batch Subscribe
Permissions:
- Platform/space admins: Configure system subscriptions, rule thresholds, and message channels
- Data developers: Configure personal subscriptions, Flow notification recipients, timeouts, and other settings; can't cancel system subscriptions
- External users: Only receive messages through Webhook, without logging in to the platform to configure anything
Feature details
Message channels
Overview
Configure how alerts are delivered. This release supports Feishu, DingTalk, and WeCom Webhooks, as well as system in-app notifications
Instructions
- Get the Webhook (and its signing secret)
- Open the My Channels tab in the my subscription drawer, click Implement Channel, and fill in the configuration (space-level channels are configured in Ops → Alert Configuration → Alert Channel)
- Click Send Test and confirm that you receive a verification message prefixed with
[Test]
Result
The channel is saved. After the test passes, you can bind the channel to subscriptions
Notes
While a channel is down, alerts are queued and retried, then resent after the channel recovers, so no alerts are lost
Subscription settings
Overview
Final alert recipients = system subscriptions ∪ personal subscriptions (deduplicated)
Instructions
- In the Flow list, click the bell icon in a row, or use Batch Operation → Add My Subscription to add subscriptions (you can't add subscriptions in the subscription list of the my subscription drawer). If your account doesn't have a personal channel yet, clicking the bell opens Implement Channel first
- Select the Flow, rules, channels, and recipients
- For timeout rules, enter a timeout duration; set a silence period as needed
- After saving, send a test again to confirm that alerts can be delivered
Result
The subscription takes effect, and its data stays in sync across entry points such as the Flow list and the my subscription list
Notes
Developers can't cancel system subscriptions. When a Flow goes online or its alert contacts change, update the subscriptions promptly
Alert rules
Overview
| Rule | Priority | Trigger condition |
|---|---|---|
| Flow run failed | P0 | The Flow status changes to failed |
| Task run failed | P0 | A task in the Flow fails to run |
| Flow run timeout | P1 | Flow run duration ≥ threshold (default 60 minutes) |
| Task run timeout | P1 | Run duration of a task in the Flow ≥ threshold (default 60 minutes) |
| Flow run exceeds the theoretical scheduling time | P1 | Run duration ≥ threshold (default 30 minutes) |
Instructions
- Enable the rules you need in subscription rules or Flow subscriptions
- Enter thresholds for timeout rules
Result
When a rule is hit, an alert event is generated and delivered according to the subscriptions
Receive and handle alerts
Overview
The interactive card (or Webhook JSON) you receive includes the priority, rule name, Flow/task, failure reason summary, time, and more
Instructions
- Check the task and failure reason in the card
- Click View task details if needed
Result
Alerts reach the people responsible directly, and the acknowledged status is written back, making it easier to measure response times
Procedure
Scenario: Enable alerts for a Flow
Step 1：
User action: Get the Webhook URL, create a channel in Message Channel/My Channels, and send a test
System response: If the test succeeds, the channel is reachable, and the group/system receives a [Test] message
Step 2：
User action: Configure a subscription (rule + channel) for the Flow, enter a timeout duration for timeout rules, and save
System response: The subscription takes effect, and the data is consistent across all entry points
Step 3：
Result: Failures reach data developers directly according to the rules, with no manual forwarding needed
Troubleshooting FAQ
Q: Why don't I receive alerts after configuring a subscription?
A：
Cause: The channel URL or signature is wrong, or the rule isn't enabled
Solution: Send a test first → check the rules and thresholds → view the sending results
Q: Can developers turn off P0 failure notifications?
A：
Cause: System subscriptions are a mandatory fallback
Solution: System subscriptions can't be canceled; you can only add personal subscriptions. To change the mandatory scope, contact the Space Admin
Best practices
- Before a Flow goes online, set up the channel + subscription + test send
- For time-sensitive Flows, be sure to enter a timeout duration; otherwise, timeout rules don't take effect
- Use system subscriptions as the fallback for P0 failures, and personal subscriptions as a supplement
- Update the configuration promptly after the Webhook or owner changes