Which Jaeger Sampling Strategy Best Supports Accurate Bottleneck Detection?

Goal

Identify a latency bottleneck in a microservice architecture before applying costly optimizations.

Constraints

• High‑throughput traffic forces the use of adaptive sampling to avoid collector overload.
• Clock skew between services can distort span timestamps in the Jaeger UI.
• Jaeger’s UI visualizes only the sampled traces, potentially hiding short‑duration spikes.

Unresolved Decision

It is unclear whether the current sampling configuration preserves the fidelity needed to detect subtle latency gaps, especially when clock skew is present.

Questions

• Does adaptive sampling at a 1% rate still expose the critical path gaps that lead to service‑level latency spikes?
• Can the Jaeger UI timeline reliably reflect true span durations when services have a ±50 ms clock skew?