Centralised notification
Overview
Harness centralised notifications allow you to send notifications based on rules and conditions across pipelines, delegates, and GitOps applications. You can configure and manage alerts to stay informed about critical events in your Harness environment.
You can configure centralised notifications for:
- Pipelines: Get notified about pipeline events like success, failure, or stage completion.
- Delegates: Receive alerts when delegates disconnect, expire, or are about to expire.
- GitOps Applications: Get notified about application sync and health events, such as sync success/failure, out-of-sync drift, and health degradation.
Pipeline Notifications
Before you begin, ensure you have the following feature flags enabled. Contact Harness Support to enable them.
PIPE_CENTRALISED_NOTIFICATION: Required for centralized pipeline notifications.PIPE_PIPELINE_RESUME_NOTIFICATION: Required for the Pipeline Resumed event.
Centralised Notifications for pipelines allow you to configure notifications for multiple pipelines at once instead of setting them up individually. You can create rules at different scopes—Account, Organisation, or Project—with each scope providing appropriate targeting options for the pipelines within its hierarchy.
These notifications are beneficial for tracking pipeline start, success, and failure events, as well as stage-level events (start, success, failure). This helps you monitor pipeline health and quickly identify issues across your deployment processes.
You can also configure the Waiting for User Action event in centralized notification rules. This event fires whenever a pipeline pauses for user input, such as an Approval step, Manual Intervention, or runtime execution input. When you add this event as a condition in a centralized notification rule, alerts are sent across all pipelines that match the rule's scope—without requiring per-pipeline configuration.
Configuration Steps
- Interactive
- Manual
-
Access Notification Management
- Navigate to Account Settings → General → Notifications Management
- Click + New Notification to create a new rule
-
Set up Pipeline Notification Rule
- Provide a Notification Name and click Continue
- Select Pipeline as the Resource Type
- Choose your scope: All Organizations or Specified Organizations
- Click Continue
-
Set Conditions
- Click + Add Condition to specify when notifications trigger
- Enter a Condition Name
- Select relevant Pipeline Events (start, success, failure, etc.)
- (Optional) Add a custom Notification Template
- Click Continue
-
Configure Channels
- Select notification channels.
- Choose existing channels or create new ones
- Click Submit to save your configuration
Delegate Notifications
Central Delegate notifications provide proactive monitoring of Harness delegates, ensuring that you are immediately aware of connectivity issues and expiration. Rules configured at higher scopes automatically apply to delegates at lower scopes, based on your rule configuration.
These notifications help track delegate disconnection, expiration, and expiring soon events, allowing you to monitor delegate health and quickly identify connectivity issues that could impact your deployments.
Configuration Steps
- Interactive
- Manual
-
Access Notification Management
- Navigate to Account Settings → General → Notifications Management
- Click + New Notification to create a new rule
-
Set up Delegate Notification Rule
- Provide a Notification Name and click Continue
- Select Delegate as the Resource Type
- Choose your scope: All Delegates or Specific Tags
- If using tags, specify up to five delegate tags
- Click Continue
-
Define Conditions
- Click + Add Condition to specify when notifications trigger
- Enter a Condition Name
- Select Delegate Events:
- Delegate Disconnected: Immediate notification when disconnection occurs
- Delegate Expired: Daily notification for expired delegates
- Delegate Expires in [X] Weeks: Advance warning (1-4 weeks)
- Click Continue
-
Configure Channels
- Select notification channels.
- Choose existing channels or create new ones
- Click Submit to save your configuration.
GitOps application notifications
Centralised notifications for GitOps allow you to receive alerts for application sync and health events. This gives you visibility into your GitOps environment without configuring notifications on individual applications or pipelines. All five notification channels (Slack, Microsoft Teams, email, webhooks, and PagerDuty) are supported.
Supported events
The following events are available for GitOps application notification rules:
| Event | Description |
|---|---|
| Application Sync Succeeded | Fires when a sync operation completes successfully. |
| Application Sync Failed | Fires when a sync operation fails. |
| Application Out Of Sync | Fires when an application's live cluster state drifts from the desired Git state. |
| Application Health Degraded | Fires when an application's health status changes to degraded. |
Custom notification templates are not supported for GitOps application notifications. Notifications use the default template.
Configure GitOps application notifications
-
Access Notification Management
- Go to Account Settings, then General, then Notifications Management.
- Select + New Notification to create a new rule.
-
Set up GitOps Notification Rule
- Provide a Notification Name and select Continue.
- Select GitOps Application as the Resource Type.
- Choose your scope: All Organizations or Specified Organizations.
- Select Continue.
-
Set conditions
- Select + Add Condition to specify when notifications trigger.
- Enter a Condition Name.
- Select the relevant GitOps Events (Sync Succeeded, Sync Failed, Out Of Sync, Health Degraded).
- Select Continue.
-
Configure channels
- Select notification channels (Slack, Microsoft Teams, email, webhooks, or PagerDuty).
- Choose existing channels or create new ones.
- Select Submit to save your configuration.
Service account (SA) token notifications
Service account token notifications alert you when tokens are created, updated, rotated, deleted, or nearing expiration. Since expired tokens can lead to pipeline and automation failures, it is important to stay ahead of their lifecycle.
SA token expiry notifications provide advance warnings, allowing you to rotate or renew tokens before they disrupt your workflows.
This feature requires the PL_SERVICE_ACCOUNT_NOTIFICATION feature flag. Contact Harness support to enable it.
Before you begin
Make sure you have access to a Harness account with the following permissions:
- View permission for Service accounts
- View permission for Notification channels
- View, Create/Edit permission for Notification rules
Configure service account token notifications
- When a service account is deleted, its notification rules are updated asynchronously. If the deleted service account was the only one in a rule, that rule is automatically removed.
- All timestamps shown in notification messages are in UTC.
You can configure notifications for an existing service account or create a new one. To create a new service account, see the Create a service account guide.
- Interactive
- Manual
-
Go to Account Settings → Notifications (or Organization Settings / Project Settings, depending on the desired scope).
-
Select + New Notification Rule.
-
Enter a name for the rule (for example, SA Token Expiration Alerts).
-
In Resources, choose Service Accounts.
-
Choose which service accounts to monitor:
- All service accounts to monitor every account in the selected scope.
- Specific service accounts to select individual service accounts by name. You can only select service accounts at the present scope.
-
Select Add Condition to choose the events that will trigger notifications.
-
Give the condition a clear name (for example, Critical Expiration Warnings). You can create multiple conditions for one rule.
-
Select the events you want notifications for:
- Token Created: A new service account token is generated.
- Token Edited: An existing token or its settings are updated.
- Token Deleted: A token is permanently removed.
- Token Expired: A token has reached its expiration date and can no longer be used.
- Token Rotated: A token is replaced with a new one.
- Token Expires in 1–4 weeks: Choose a timeframe between 1 and 4 weeks before expiration. Notifications are sent daily during the selected week. For example, if you select 4 weeks, you will receive a daily notification during the fourth week before the token expires.
- Token Expires in 1 day: A final reminder sent one day before expiration.
-
Select an existing channel or create a new channel to receive alerts.
-
Turn on Enable on Save to activate the rule immediately after you click Submit.