Diagnosing service issues in distributed infrastructure is a complex multi-layered problem. Component oriented performance management tools often report a good state even while customers are complaining about poor service. This causes many wasted hours in operations war rooms casting blame and spinning wheels.
Often problems are caused by the complex interaction between different service components. Miss-configuration or ‘sick but not dead’ components can be incredibly difficult to isolate. Performance diagnostics centered around standalone applications, middleware or database components can often misdirect engineering teams to look in the wrong place.
When all other monitoring tools prove blind to session level problems, packet capture is typically the last resort. But reactive human analysis of raw packets, often at multiple sample points, is a difficult & time consuming task, especially under pressure during a business impacting problem.
What’s missing is a real-time customer experience data source that provides both a deep and wide view of transaction route & execution across enterprise infrastructure. A monitoring solution that can provide granular service quality insight by decoding every message, query, protocol and relating them together in an end to end transaction lifecycle.