A healthy-path load test missed the retry storm. Here is how 300 customer requests became nearly 900 downstream calls—and how we redesigned the test.
Our Load Test Passed Because It Never Triggered Retries Read More »
A PostgreSQL index scan still processed 8.7 million rows because a materialized CTE applied selective filters after an expensive join in production.
PostgreSQL Used the Index. It Still Processed Millions of Rows. Read More »
Spring Boot stayed healthy while Glass Fish exhausted its Oracle connection pool, delaying searches before SQL execution and causing timeouts.
Our Server Was Healthy. The Connection Pool Was Exhausted. Read More »
A fast SQL query loaded 100,000 Hibernate entities, exhausted the JVM heap, and exposed why previews and full reports need different retrieval boundaries.
The Database Was Fast. Our Spring Boot API Still Ran Out of Memory. Read More »
A Spring Boot outage showed how green readiness probes hid Tomcat thread exhaustion, slow dependencies, and broken timeout design.
Why Our Spring Boot Health Check Lied During an Outage Read More »
Learn why a PostgreSQL index can still be useless and how composite indexes, EXPLAIN ANALYZE, and query patterns improve reporting speed.
A Spring Boot load test where async handled requests slower than sync. Compare latency, throughput, and thread behavior.
Async Was Slower Than Sync in This Spring Boot Load Test Read More »
See how Redis changed Spring Boot API latency, throughput, and P95 response time in a practical cache benchmark.
Redis Cache Expiration Causes Hidden Latency Spikes Read More »
Tested Redis caching in a Spring Boot API and achieved up to 30x faster response times. See benchmarks, P95 impact, and key insights.
Redis Made This Spring Boot API 30x Faster in a Load Test Read More »
Spring Boot API latency investigation showing slow responses despite fast PostgreSQL queries, with real metrics and optimization insights.
API Slow? Database Fine — Real Performance Case Study Read More »