Notes from running CleverQA in production
Engineering posts written from real production experience — what broke, what we learned, and how we fixed it.
Your cron job said success. 8% of it failed.
November 11, 2026 · 8 min read · The CleverQA TeamA notification job exited 0, pinged its heartbeat and stayed green while 8 of 100 emails never went out. Why batch jobs fail a little instead of all at once, why the usual workarounds backfire, and how to let the job report its own partial failures.
Read more →Load testing without writing a script
October 21, 2026 · 14 min read · The CleverQA Team“How many users can we handle?” deserves a number, per flow, and the breakpoint rather than a pass at expected traffic. Why most teams never load test, and why over-provisioning is the expensive way to avoid asking.
Read more →What is a monitoring system actually worth?
October 21, 2026 · 18 min read · The CleverQA TeamA demo died on a background job nobody was watching. For a startup, downtime costs credibility, not revenue-per-minute, and detection is most of every outage. What monitoring actually buys, and what it doesn't.
Read more →Your monitor says the site is up. Your server is dying.
October 7, 2026 · 8 min read · The CleverQA TeamEvery uptime check passed for seven hours while a memory leak took the server down. Why outside-in monitoring misses a dying box, and the three questions you actually need answered: which service, since when, how fast.
Read more →Why we never deploy on Fridays
October 7, 2026 · 7 min read · The CleverQA TeamA Friday release broke one quiet flow and nobody noticed for 36 hours. Why 'no Friday deploys' treats the symptom, and what it takes to actually find out when something breaks.
Read more →