https://wid-blog.github.io/en/tags/slo/Recent content in SLO on wid's blogHugoenMon, 15 Sep 2025 00:00:00 +0000https://wid-blog.github.io/en/posts/tech/architecture/low-latency-sla/Mon, 15 Sep 2025 00:00:00 +0000https://wid-blog.github.io/en/posts/tech/architecture/low-latency-sla/An SLA isn’t kept by monitoring after the fact — it’s built by design and held by operation. Once the SLI for a low-latency service becomes p99 latency rather than availability, timeout budgets, caching, degradation, isolation, and load shedding build the SLA, while p99 SLOs, burn-rate alerting, headroom, deploy gates, and the review cycle hold it.