Splunk O11y Cloud Explained: Real User Monitoring (RUM)

In this video, we explore Splunk Real User Monitoring (RUM) within the Observability Cloud. Starting from the application summary dashboard for the "webinar store" application, key metrics, such as page views and JavaScript errors, are showcased, which are grouped by stack traces. This allows developers to identify the exact lines of code causing issues for users. Notably, latency issues have been reported during checkout, and by drilling down into the UI metrics tab, it’s clear that the cart checkout time has increased to about three seconds, which is higher than expected. The demonstration highlights how to isolate the issue to a specific checkout workflow, examining user session traces and identifying that a "doc fetch" operation is taking excessive time.

Continuing troubleshooting, the tag spotlight feature is utilised to check for patterns among user agents or geographical locations but finds none. Focusing further, log traces are reviewed in Log Observer Connect, where an error reveals that an invalid API token is causing payment service failures. By examining different application versions, it is confirmed that version 350.10 is linked to errors, whereas version 350.9 is not. This insight allows the team to roll back to the stable version, resolving latency issues for end users. In summary, Splunk RUM enables rapid, directed troubleshooting, helping the team restore services efficiently and enhance user experience during checkout.