Create alert rules

By creating alert rules, you and your team can proactively monitor your app's performance. This page provides a step-by-step guide on how to create an alert rule for your app, ensuring you stay informed about potential issues.

Best practices

Follow these best practices to create useful alert rules:

• Focus on critical metrics: Create at least one alert rule based on the invocation success rate metric to ensure that customers have low downtime and your app is reliable.
• Categorize alerts: Use different severity levels, including Minor, Major, and Critical, to prioritize responses.
• Avoid overly sensitive thresholds: Set realistic and meaningful thresholds to avoid triggering too many alerts. You can also consider configuring the alert tolerance of your alert rules to control sensitivity.
• Base thresholds on historical data: Use past performance data to establish baseline metrics and determine appropriate thresholds.
• Reduce alert noise: Avoid creating multiple alerts for the same condition. Disable alert rules during ongoing incidents or planned maintenance to prevent redundant alerts.
• Assign relevant ownership: Ensure each alert has responders who are responsible for taking action.
• Seek continuous improvement: Periodically review and adjust alert rules based on new insights, system performance changes, or team feedback.
• Include documentation: Ensure alerts provide enough context and guidance on the required actions by including a link to documentation in the provided field when creating the alert rule.

To begin creating an alert rule:

1. Access the developer console.

2. Select the app for which you want to create an alert rule.

3. In the side navigation panel, select Alerts and then select Alert rules.

4. Select the Create alert rule button in the top right-hand corner. This will open a screen that breaks down the process of creating an alert rule into four steps:

   • Step 1: Select alert source
   • Step 2: Configure conditions
   • Step 3: Add responders
   • Step 4: Specify alert details

Step 1: Select alert source

In this step, you will select the metric and filters that define the source of the alert.

Select source metric

Select the metric you want to monitor. The alert will be based on the production data of the selected metric. The available metrics are:

• Invocation count: The total number of invocations, regardless of success or failure.
• Invocation errors: The number of invocations that failed with an error. Errors can be due to out of memory issues, timeouts, or unhandled exceptions.
• Invocation success rate: The percentage of successful vs. failed invocations.

Add source filters (optional)

For more specific alerts, you can add filters to refine the source. You can filter by:

• Error type (only for invocation errors): The specific types of errors that will trigger the alert.
• Site: The specific sites that will be monitored.
• Version: The app versions that will be monitored.
• Function: The specific functions within your app.

Step 2: Configure conditions

In this step, you will define the conditions that will trigger the alert. This includes selecting the time period over which data is collected, setting the severity levels and thresholds, and configuring advanced tolerance settings as needed.

Select time period

Select the time period during which data will be collected and evaluated. Data is aggregated over contiguous fixed intervals.

The available options include 5 m, 10 m, 15 m, 30 m, 1 h, 2 h, 3 h, 6 h, 12 h, and 1 day.

Define trigger conditions

Specify the conditions that will trigger the alert. These conditions have three main properties:

1. Severity: The level of importance of the alert.

   • Minor
   • Major
   • Critical

2. When: The trigger logic for the alert.

   • Above
   • Below
   • Above or equal to
   • Below or equal to

3. Threshold value: The numeric value that the metric must reach to trigger the alert. For invocation errors and invocation count, this is an integer. For invocation success rate, this is a percentage and the value must be between 0 and 100.

You can add several conditions to create an alert with multiple severities. This allows you to differentiate between varying levels of business impact and also to detect issues early. For instance, a Minor alert could serve as an early warning, prompting preventive measures before the issue escalates to a Critical level.

Select alert tolerance (optional)

To help reduce alert noise from temporary spikes or dips in data, you can define the alert tolerance as an optional advanced configuration. Alert tolerance specifies how many consecutive time periods the threshold must be breached before the alert is triggered.

For example, if you set the alert tolerance to 3, the threshold must be breached for three consecutive time periods before the alert is triggered.

If alert tolerance is not configured, it defaults to 1. This means that the threshold must only be breached for one time period before the alert is triggered.

Step 3: Add responders

In this step, you will specify who should be notified when the alert is triggered. Responders will be notified by email when the alert is opened, when the alert changes severities, and when the alert is closed.

Add responders

From the list of contributors associated with the app, choose the individuals who should receive email notifications when the alert is triggered. You can select multiple responders to ensure that all relevant team members are notified.

Once you’ve added the responders, you’re ready to finalize the alert rule in the next step.

Step 4: Specify alert details

In this final step, you will provide additional details about the alert to help your team identify and manage it effectively.

Enter alert name

Enter a clear and descriptive name for the alert. This name should help you quickly identify the alert in the developer console and in notifications.

Add description (optional)

Provide a brief description of the alert. Include relevant details, such as what the alert monitors and why it’s important. This will help your team understand the alert’s purpose at a glance.

Add documentation (optional)

If available, enter a URL that links to related documentation or a runbook. This could be a guide to troubleshoot the alert or additional context on how to resolve the issue. Providing a documentation link can make it easier for the team to take action when the alert is triggered.

You can review the alert rule configuration by using the Back and Next buttons. Once you've completed all the steps, select Create alert rule to save the alert rule.