Incident Report: PWA Path Change Caused High Latency
After introducing API Gateway and separating backend services from the frontend, I started to experience real slowness in Goalixa.
Before this, latency was mostly something I saw in SRE dashboards. Now I can feel it directly while using the app, and that gave me much better understanding of what latency means in real product experience.
Incident Timeline
| Time | Event |
|---|---|
| 2026-02-24 | Changed the path/routing setup in the PWA |
| Shortly after | Website became noticeably slow |
| Detection | Monitoring showed latency increase in the PWA service |
Impact
- Slower page response in the PWA
- Higher latency visible in monitoring graphs
- Network instability symptoms between frontend and backend paths
Root Cause
Even in monolithic systems, latency is critical. In gateway-based and service-separated systems, latency becomes even more sensitive because every extra network hop and misconfiguration adds delay.
The path change introduced inefficient routing that caused multiple unnecessary hops between services.
Actions Taken
Fixed major bugs in:
- PWA - Corrected path routing configuration
- API Gateway - Optimized request forwarding
- Core API - Reviewed endpoint response times
Reviewed route/path behavior to remove high-latency flow.
Next Optimization Steps
I want to enable caching mechanisms after stabilizing the bug fixes:
- Gateway-level caching - Where safe for idempotent requests
- Static and edge caching - For PWA assets
- API response caching - For read-heavy endpoints
Lessons Learned
Seeing latency in charts is useful, but feeling it in real usage is different.
This incident helped me connect SRE metrics to actual user experience and prioritize bottleneck analysis in both code and network paths.