kubesense-apm

You can analyze Kubernetes distributed traces by searching raw span records and computing latency percentiles across clusters and namespaces. This process pinpoints exact latency bottlenecks and error patterns within specific time windows.

Field discovery in Kubernetes APM automatically identifies available trace fields prior to querying. You need it to avoid guessing which fields exist, ensuring your trace search filters match valid data structures before execution.

You search raw Kubernetes traces by applying built-in filters to inspect span records across workloads. This surfaces relevant error events and service relationships, allowing you to diagnose specific application performance issues.

Yes, you can compute latency percentiles and error rates across multiple Kubernetes namespaces. Aggregate trace analysis evaluates workload relationships and service performance to identify systematic errors over defined time windows.

The best way to diagnose service performance issues using distributed tracing is to combine field discovery with aggregate trace analysis. This approach identifies latency patterns and service relationships across clusters without blind querying.

Yes, you need to specify time windows and workloads when analyzing Kubernetes traces. Defining these parameters filters the distributed trace analysis, ensuring you accurately identify latency patterns and error rates for targeted workloads.