The Alerts page displays your alert policies, and allows you to create customized alert policies for anything lumigo monitors, including event-based alert (e.g., timeout) and key metrics (e.g., error rate).

What is an Alert Policy

An Alert Policy defines the conditions under which an alert is triggered for a resource and, optionally, specifies whether to generate notifications to the relevant team members, by email or other channels such as Slack or PagerDuty.

Lumigo comes with a number of predefined Alert Policies, and you can define (and edit) additional Alert Policies from the Alert Policies page, which you can access by clicking the Alerts button on the navigation bar or Manage Alerts on Issues page.

📘

Quick Start

To set up alert policy, click "Create New Alert" and follow these steps or

• Select the service you want to set up the alert for
• Select the resources that you wish to monitor for alerts
• Select the alert type and define the condition
• Define the notification preferences
• Click “Save Alert”

Configure Alert Policies

The Alert Policy page displays the Alert Policy you are editing, or if you are creating a new Alert Policy, it displays an empty policy.

1. Under Description, enter the following parameters:

   • Alert Policy Name: The policy’s name. If you are editing an existing policy, you can change its name. If you are creating a new policy, then enter a descriptive meaningful name.
   • Service: Select a service from the dropdown menu.

2. Under Resources, select one of the following:

   • All Functions: The Alert Policy will be applied to all functions, including functions that will be defined in the future.
   • Functions with Tags: The Alert Policy will be applied to all functions with the Tags you select from the dropdown, including functions that will be defined in the future.
   • Selected Functions: The Alert Policy will be applied only to the functions you select from the dropdown. If you later define a new function to which you would like to apply this Alert Policy, you will have to add it to this list.

3. Under Condition, select one of the following:

   • Event: The Alert Policy will be applied only if one of the events you then select from the Select events dropdowns occurs. You can select any number of events.
     • Select all: The Alert Policy will be applied to all the events in the dropdown.
     • Any error: The Alert Policy will be applied to all error events.
     • Otherwise, select individual events.
   • Metric: the Alert Policy will be applied only if the condition you define next is fulfilled. Select a condition (Percentage of or Sum of) from the dropdown, an operator (>=, <=, >, or <), and specify an operand (a number).

4. Under Notification, you can specify which of the available notifications will be sent when the Alert Policy is applied, and how often.

   • You can enable or disable the notifications displayed by checking or unchecking them. The available notifications are those defined in the Notifications page.
   • Notification Frequency: Select a frequency from the dropdown.
     • Every event means a notification will be sent every time the Alert Policy is applied.
     • The other choices specify the length of time after the first time the Alert Policy is applied for a specific Condition that must pass before a notification is sent. For example, if you select 1 Hour, then notifications will be sent 1 hour after the Alert Policy is first applied, no matter how many times it is applied during that hour.

📘

Ignore issues triggered by HTTP 4XX errors

You can configure your alert policies to ignore HTTP 4XX errors and focus on your system health, by hover over the event "Any error" and check "Ignore issues triggered by 4xx error responses".

Adding Alerts with AWS Tags

If you are using auto-tracing for your Lambda functions, alerts will already be added automatically.

If you're not using auto-tracing, the fastest way of adding alerts to your functions automatically is by flagging them using AWS Tags. To do so:

• Open the configuration for your AWS Tags
• Create the tag lumigo:auto-alert and set the value to basic
• Apply the tag to your Lambda function
• Wait 15-20 minutes for the changes to propagate.

Lumigo will now automatically report statistics on your Lambda function!

• Using the auto-alert tag will override any alert configurations that were previously configured with the UI.

📘

Note

We recommend limiting the notification frequency for policies based on single events in your system, such as Lambda Errors.

• One function tends to fail in batches, with significant periods of inactivity between them. Throttling this function's alerts to 24 Hours will limit notifications to once a day, preventing you from being bombarded with notifications.
• For functions where all failures are critical, you can disable throttling entirely by selecting "Never" from the dropdown. With this option selected, you will receive a notification each time the function fails.