How We Optimized Our Product Without Paid Solutions
summary
A team replaced third-party DevOps services with their own operations, audited infrastructure, and added production monitoring and profiling. They built response algorithms using Graylog, Grafana, InfluxDB, Pyroscope, and Prometheus, halved resource utilization, and sped up key API responses, explaining the problems found, metric improvements, and the move to Swoole.
AI-written summary · about
chapters (10)
- 0:00Optimization context
- 3:55Service response metrics
- 7:13Server caching
- 10:04Memory leak fixes
- 14:19Business logic auto-tytles
- 20:43Redis connection pool
- 24:00PostgreSQL tuning
- 27:11Async data processing Open Swoole
- 32:33Optimization benefits
- 36:33Q&A
related talks
20mUsing Monitoring to Speed Up Your DevOps Journey
27mBenchmarking and Optimizing Software Delivery Performance
25mOpen source observability for private cloud: mission impossible or not?
45mDecentralizing SaaS Monitoring Using Open-Source Tools and Standards
54mDevOps in Real Life, A How-To
55mDevOps @Scale (Greek Tragedy in 3 Acts)
45mPerformance-by-Design: Embedding Intelligent Scaling and Guardrails Into Platform Engineering
31mHow We Went From Two Major Outages to 99.98% Reliability in 6 Months