AppsFlyer reduced log volume by 66% without compromising production visibility by building an offline validation framework. Instead of manually tuning log levels across 1,800 sources, they partnered with Auditty to deploy a host-level agent that suppresses repetitive logs. The validation approach involved replaying peak production traffic through S3 and Athena, testing the suppression engine offline before live deployment. The solution required zero code changes, maintained security boundaries, and gave service owners full control through Kubernetes-native configuration. Rolling out to Kafka clusters achieved 50% log reduction across 1,000+ servers while preserving critical debugging signals.

7m read timeFrom medium.com
Post cover image
Table of contents
The Signal vs. The BillThe Challenge:“Just Tune It” Wasn’t RealisticOur Solution: The Offline Validation FrameworkThe Peak Hour ReplaySlicing the DataThe “Offline” Agent (CLI)The Cross-Check

Sort: