@sureshbabu
Backend Developer
Shipping SaaS features weekly. Always learning system design. (tamil/kannada)
We ran into this around data exports during a traffic spike. Retries arrived out of order and our fixtures only modeled the happy path. Sleeping in tests hid the races and slowed CI. Replaying a small set of recorded payloads found the idempotency bug. We asserted on stored keys, not on elapsed time. The suite is shorter and finally recreates the incident we cared about.
We ran into this around team onboarding while rolling back payments. The work was mostly unblocking other people, not closing my own tickets. Clear noes protected the roadmap more than extra hours. Writing the doc nobody wanted still changed how fast the team moved. Impact was hard to see until a project stalled without that context. I am still learning to describe that work without vanity metrics.
I used to think more indexes always meant faster queries. Production taught me about write amplification and table bloat instead. We had three indexes that nothing queried, slowing every insert. EXPLAIN ANALYZE finally showed which plans actually used which indexes. After dropping the dead ones, writes got healthier without hurting reads. Now I review unused indexes in the same ritual as reviewing slow queries.
No spike in CPU. Error budgets looked fine at a glance. Users still reported blank pages in a thin slice of traffic. Logs only showed upstream resets with no clear application exception. The culprit was a stale keep-alive timeout between nginx and the app. Aligning idle timeouts stopped the intermittent 502s within an hour. We also added a dashboard for upstream reset reasons so the next page is faster.
We extracted one high-churn billing endpoint behind a strangler facade. Dual-writes ran for two weeks while we compared totals nightly. A feature flag controlled read traffic so we could roll back instantly. The hardest part was matching edge-case rounding in legacy invoices. Cutover finished with no customer-facing downtime and a smaller blast radius. We kept the facade until three more endpoints followed the same path.
We needed locks so queue workers did not process the same job twice. Redis SET NX was faster under load and easy to expire automatically. Postgres advisory locks were simpler operationally for our small team. Failover behavior mattered more than raw latency in our case. We chose Postgres first, then moved hot paths to Redis later. Pick the lock store you can operate confidently at 3am.