#benchmarking

Performance & Scaling Bug Report

Our in-process cache made p99 latency worse: an optimization we had to roll back

Ingrid Solberg · Nov 4, 2026 · 7 min read

We added a large in-memory cache to a Java service to save Redis round trips. Average latency improved and p99 doubled. The cache pushed the heap into long GC pauses. How we found it in GC logs, why our benchmark missed it, and the measurement method we use now.

Performance & Scaling

p95, p99 or p99.9? Choosing latency percentiles for SLOs, benchmarks and dashboards

Elena Kovač · May 25, 2025 · 6 min read

Every team argues about whether to track p95 or p99. The answer depends on traffic volume, fan-out and what the number is for. A practical guide to choosing percentiles for SLOs, alerts, benchmarks and capacity planning, with the sample-size maths, coordinated omission and how to report results honestly.