Alerts

Sentry users with admin permissions or higher can create one of two types of alerts: Metric Alerts or Issue Alerts. You can use Metric Alerts for high-level monitoring of patterns, or fine-grained monitoring of individual events. An Issue Alert fires whenever any issue in the project matches specified criteria, such as a resolved issue re-appearing or an issue affecting many users. To confirm or set admin permissions, see Project Settings > Alerts > New Alert.

Metric Alerts

Metric alerts allow you to filter and set thresholds on all errors across a project. Use metric alerts for high-level monitoring of patterns or fine-grained monitoring of individual events. Metric alerts help express lightweight service-level objectives (SLOs) such as users affected by signup-page errors or volume of database errors.

Alternatively, issue alerts notify you about the impact of a specific issue.

Alert Builder

A metric alert has, at most, two triggers. The first is a critical trigger, which is required. The second is a warning trigger, which is optional. Triggers are independent of one another; however, the warning (optional) must be reached before the critical (required) trigger.

Sentry evaluates triggers approximately every minute from the highest severity to lowest. Sentry creates an alert with the severity of the matched trigger (warning or critical). If an alert is already active, its status is updated. Admins can resolve alerts manually, or automatically by setting the resolution threshold.

When an alert is created or changes status, the actions associated with the trigger are executed. The available actions are:

  • Send an email (to a member or team). If sent to a member, the member’s personal project alert opt-out settings are overridden.
  • Send a Slack notification

Alert Stream

Alerts can be accessed via the Alerts sidebar item.

Alert stream illustrates alert duration, trends, severity, time duration and percentage of users affected.

Alert Details

In this page, you can view a single instance of an alert from the time the alert is created to when the alert is resolved, or the current time if the alert is still active.

Alert details illustrates a graph showing errors.

The “subscribe” button subscribes you to workflow notifications for the alert, which include status changes and comments.

Issue Alerts

Issue alerts fire whenever any issue in the project matches the specified criteria, such as a resolved issue re-appearing or an issue affecting many users.

Conditions

Conditions are evaluated for an issue alert each time the issue receives a new event, subject to rate limits.

Available Conditions

  • Sentry detects/creates a new issue
  • An issue changes state from resolved to unresolved
  • An issue changes state from ignored to unresolved
  • An issue has happened more than {value} times in {interval}
    • value: a positive integer
    • interval: one minute, one hour, one day, one week, or 30 days
  • An issue has been seen by more than {value} users in {interval}
    • value: a positive integer
    • interval: one minute, one hour, one day, one week, or 30 days
  • Sentry detects a new event
  • Sentry detects a new event with tags matching {key} {match} {value}
    • key: any tag
    • match: equals, does not equal, starts with, ends with, contains, does not contain, is set, or is not set
    • value: any key’s value
  • Sentry detects a new event with {attribute} {match} {value}
    • attribute: messageplatformenvironmenttypeexception.typeexception.valueuser.iduser.emailuser.usernameuser.ip_addresshttp.methodhttp.urlstacktrace.codestacktrace.module, or stacktrace.filename
    • match: equals, does not equal, starts with, ends with, contains, does not contain, is set, or is not set
    • value: any attribute’s value
    • When viewing an event, if an attribute is unclear or missing, you can view its JSON payload to determine its value.
  • Sentry detects a new event with level {match} {level}
    • match: equal to, less than or equal to, greater than or equal to
    • level: fatal, error, warning, info, debug, or sample

Actions

The following actions are available:

  • Send a notification to all legacy integrations
    • Legacy integrations, also known as Plugins, are configured per project. Legacy integrations currently include email.
  • Send a notification to a single legacy integration (service)
    • “service” is the same as legacy integration
  • Send a notification to a global integration, which includes

The alert condition in this example shows the conditions are conflicting with each other.

For more details, see Legacy Integrations (via “Service” as any individual integration) that can send Notifications.

By default, issue owners receive alert emails. If an issue owner is not configured or not found, the email will either not send or send to all project members as defined in Project Settings > Issue Owners.

Issue owners can be toggled on or off.

Rate Limit

The rate limit determines how frequently an issue alerts. After an issue fires an alert, Sentry won’t check the conditions and won’t execute the actions for that issue until the rate limit period passes. The limit is set to perform the action according to one of these intervals:

  • minutes: 5, 10, 30, 60
  • hours: 3, 12, 24
  • one week or 30 days

For example, if an issue violates alert conditions multiple times in a 30-min period, but your frequency threshold is 30 minutes, you’ll only get one alert.

A dropdown that adjusts the frequency of an alert.

Environment

Environment control allows you to specify which environment qualifies for your issue alert.

  • All Environments
  • Any one of your defined (and not hidden) environments

A dropdown allowing you to choose an environment to attach to an alert.

If you select “All Environments”, Sentry checks individually for each environment, rather than combined events across environments.

Digests

The digests feature works only for issue alert emails and limits alerts across projects. This project-level setting allows you to batch issue alerts to limit the total number of emails you receive for that project. Use the sliders to control the frequency.

A sliding adjustment scale for the frequency of alert emails.

Alert Listing

This is a listing of all your project’s alerts.

Alert rules are divided by projects, their conditions/triggers, and actions(s).

Alert Subscription

Issue Owners

By default, issue owners receive alert emails. If an issue owner is not configured or not found, the email will either not send or send to all project members as defined in Project Settings > Issue Owners.

A toggle indicating if all users are issue owners or not.

Ignore

When an issue is set to ignore, Sentry mutes the alerts for that issue. You can also ignore an issue until certain conditions are met —- for example, “ignore for 30 minutes.” Keep in mind; an ignored issue will still count towards your quota.

Personal Alert Settings

Review your personal alert settings in User Settings > Account > Notifications.

Use the “default project alerts” setting to set your default preference across all projects — subscribed or unsubscribed.

Note that this setting does not affect alerts you’ve configured to send to your email explicitly.

A toggle for turning on or off all project alerts.

After selecting the appropriate alert setting, selectively change it by project in User Settings > Account > Fine tune alerts by project.

Each project has three options: Default, On, or Off. Selecting default uses your default preference from the previous step.

Dropdown indicating a choice between default, on, or off.

FAQs

  • Can I copy an alert rule to another project?
    • This feature will be available in a future release.
  • Can I set different default alert rules?
    • This feature will be available in a future release.
  • Are there issue-level filters other than environment?
    • No, all filters are event-based. For example, configurations don’t exist for alerting only if an issue is X days old, or assigned to Y, or alerted-on-before, etc.
  • What is the difference between Delete, Delete & Discard, and Ignore?
    • Delete - Deleting an issue deletes all data associated with it, and creates a new issue if an event with the same fingerprint happens again. Alerts and workflow notifications for this new issue behave just like notifications for any new issue.
    • Delete & Discard - when you delete and discard an issue, all notifications for the issue will stop
    • Ignore - While an issue is ignored, all alerts for that issue are muted. You can also ignore an issue until certain conditions are met — for example, “ignore for 30 minutes.” Keep in mind; an ignored issue will still count towards your quota.