Workflow and task alarms are a unified notification capability in DataBuddy. This document is the central document for alarms: The alarm configuration rules are consistent across all task types and are described in detail in this document, while other task documents (Notebook / Python / SQL / offline ingestion, nested workflows, quality monitoring, and so on) provide only brief tables and links.
Overview
Workflows and tasks can be configured with alarms independently without interfering with each other. An alarm consists of three parts: alarm conditions, receiving channels, and do-not-disturb rules.
Category | Workflow alarm | Task alarm |
Startup Notification | ✅ | ✅ |
Success Notification | ✅ | ✅ |
Failure Notification | ✅ | ✅ |
Monitoring Metric Alarms | ✅(Running time / Completion time) | ✅(Running time / Waiting time / Completion time) |
Receiving Channel | Email / Webhook / Teams / Slack | Email / Webhook / Teams / Slack |
Do Not Disturb: Manual Termination | ✅ | ✅ |
Do Not Disturb: Before Last Retry | - | ✅ |
Prerequisites
You are already the workflow owner or have Manage permission.
Receiving channels have been configured for the workspace. Except for directly entering an email address, Webhook / Teams / Slack must be configured in advance in Configure Notification Channels. Operation Steps
Step 1: Entering Alarm Configuration
Workflow Alarms
1. Open the workflow details page.
2. In the right drawer, click Alarm Configuration > Add Alarm.
Task Alarms
1. Click the target task on the workflow editor canvas.
2. In the task configuration panel, locate Alarm Configuration > Add Alarm.
Step 2: Adding a Notification Channel
Channel. | Configuration Method |
Email | Select the personal email channel and manually enter the email address in the input box. Selecting a configured email channel is not supported (email addresses added in Workspace > Notification Management are not used for alarm channel reference). |
Webhook | Select the configured Webhook channel. |
Teams | Select the configured Teams channel. |
Slack | Select the configured Slack channel. |
Adding a System Channel
If no suitable option is available in the dropdown list:
1. Click + Add Sending Method.
2. Go to Workspace Settings > Notification Management to configure a new channel.
3. After completion, return to the alarm configuration. The new channel appears in the dropdown list.
Step 3: Selecting Alarm Conditions
Select at least one item:
Condition | Description |
Starting | Send a notification when a workflow/task starts running. |
Successful | Send a notification when a workflow/task runs successfully. |
Failed | Send a notification when a workflow/task fails to run. Selected by default. |
Monitoring metrics | Send a notification when a monitoring metric triggers a warning or timeout. Configure the monitoring metric threshold before selecting this option. |
Note:
If the monitoring metric threshold is not configured first, the Monitoring Metrics option cannot be selected. Set the time threshold in Monitoring Metrics first.
Step 4: Configuring Monitoring Metrics
In the Monitoring Metrics area of a task or workflow, click Add. A metric configuration dialog appears. Only one rule can be configured for each metric. If any metric is triggered, the alarm condition is considered met.
Metric | Description | Workflow | Task |
Running duration | Duration of task running. | ✅ | ✅ |
Waiting duration | Duration of waiting for upstream or resources after a workflow is triggered. | - | ✅ |
Completion time | Specific time point when a task run is completed (periodic runs only). | ✅ | ✅ |
For each metric configuration:
Threshold | Description |
Warning threshold | An alarm is triggered when this value is exceeded (associated with the alarm condition monitoring metric). |
Timeout threshold | When this value is exceeded, the task fails or is retried, and the failure and timeout policy needs to be configured accordingly. |
Note:
The timeout threshold must be greater than or equal to the warning threshold. Otherwise, the configuration cannot be saved.
Default Thresholds and Scenarios Where They Do Not Take Effect
Metric | Default Value | Invalid Scenario |
Running Time | 00h00m00s
| The default value does not take effect. You must manually change it to a value greater than 0 to trigger monitoring. |
Waiting duration | 00h00m00s
| Same as above. |
Completion time. | Null | A null value does not take effect. You must specify a specific time point to trigger monitoring. |
Step 5: Selecting Do Not Disturb
Do Not Disturb | Description | Workflow | Task |
Do Not Disturb on Manual Termination | No notification is sent when a task is manually terminated. | ✅ | ✅ |
Do Not Disturb Before Last Retry | No notification is sent during retries. A notification is sent only when the last retry configured in the failure and timeout policy fails. | - | ✅ |
Step 6: Saving
Click Save to make the alarm configuration take effect. After the configuration is completed, the Alarm Configuration area on the workflow and task details pages displays the following:
Receiving channels (user group name and channel icon).
Alarm policy.
Parameter Description
Alarm Condition Trigger Logic
Condition | Trigger Timing |
Starting | Triggered once when a workflow / task goes from Pending to Running. |
Successful | Triggered when a workflow/task reaches the final state Succeeded. |
Failed | Triggered when a workflow/task reaches the final state Failed, including scenarios where the workflow/task is set to Failed by the timeout threshold. |
Monitoring metrics | Triggered once when a monitoring metric exceeds the warning threshold. |
Additional Rules for Receiving Channels
Item | Rule |
The same email address cannot be configured repeatedly in a single alarm pop-up window. | Validate during configuration and prompt "This email address already exists" when a duplicate is found. |
The same system channel cannot be configured repeatedly in a single alarm pop-up window. | Configured channels are hidden in the dropdown list. |
Maximum number of alarms for a single workflow / task | 50 items. |
Time Zone for Completion Time Alarms
The trigger time of the completion time alarm is consistent with the scheduling time zone of the workflow. For example, if the scheduling time zone is set to East US time and the completion time is set to 09:00, the alarm is triggered at 09:00 East US time.
Use Limits
The default monitoring metric values (00h00m00s or empty) are filtered out by the system, so the alarm does not actually take effect. Enter a valid threshold.
You must first configure the monitoring metric threshold for the Monitoring Metrics option in the alarm conditions.
The alarm limit is 50 per workflow or task. No more alarms can be added after the limit is exceeded.
FAQs
Q1: Why did the task fail but no alarm was received?
Check whether Failure is selected in the alarm configuration.
Check whether the alarm is blocked by Do Not Disturb on Manual Termination / Do Not Disturb Before Last Retry.
Check whether the receiving channel is still reachable, for example, whether the Webhook URL has expired or whether the Email has gone to the spam folder.
Q2: Monitoring metric = Completion Time. Will the alarm time change after the scheduling time zone is switched?
Yes. The completion time alarm is calculated based on the scheduling time zone. For example, if the scheduling time zone is switched from China to the US, the 9:00 alarm will be triggered at 9:00 US time.
Q3: Is alarm deduplication supported?
A monitoring metric triggers an alarm only once each time the warning threshold is reached during a task run. Different runs are independent of each other, and no cross-run deduplication is performed.
Q4: Can I temporarily disable all alarms for a workflow?
Yes. If you delete all alarm entries in the alarm configuration, the workflow will not send any notifications. We recommend that you pause individual entries instead of deleting them so that they can be restored later.
References
Other Configurations