Grafana Cloud

Analyze User Sessions

Explore User Sessions

Browser, user-journey, and UX related events recorded by the Frontend Observability Web SDK are associated with a unique session id to enable deeper insights into erroneous conditions and why they occurred.

You can find all recorded user sessions on the Sessions page. The sessions list can be filtered by the session properties.

Sort and filter sessions

You can filter sessions by properties and meta data in the session panel.

Filter User Sessions

For example, filter on the following meta data to get quick insights into users’ behavior:

  • page_url - the full url where the event was recorded
  • browser_* - meta attributes about the browser and the operating system
  • event_* - meta attributes related to a recorded event

Click on the column headings to sort the data. Subsequent clicks will cycle through the sort order: descending, ascending, and not sorted. For example, click the “Errors” heading to order the sessions by error count in descending order.

Sort User Sessions

Analyze a single user session

Click on a Session Id to view the session details.

navigate to a single session

On the Session Details view, you can learn more about: the user session attributes, user journey events, and backend traces that were created as part of the user session.

Inspecting user journey events

The user journey table contains all events that were automatically recorded and manually created:

  • session_start - This event signals that a user session was started
  • view_changed - This event indicates that the user transitioned into another view
  • exception - This event represents an error that occurred, use it to navigate into error awareness and analyze the occurrence

Analyze a single session

Click on a single event to view the raw event log or a single error view, depending on the event kind.

Inspecting traces produced as part of a user journey

If the tracing instrumentation is activated, the Web SDK will record interactions with backend services as trace data. Those interactions will have an additional session_id attribute and allow for easy discovery of slow backend transactions.

Analyze traces in scope of a single session