Traffic Metrics for Laravel Cloud Migration
Summary of data shared with Ryan (Google Doc). This doc is the engineering reference.
info
All values in req/sec. Metrics sourced from AWS CloudWatch Lambda invocations (per-minute granularity, converted to per-second). Peak values represent max within 5-15 minute windows. All projects run on Laravel Vapor with Docker runtime (except farfalla-https-guard and farfalla-integrations which use PHP runtime). Queue workers are separate Lambda functions from HTTP handlers.
Current Baseline​
Period: Dec 31, 2025 - Jan 15, 2026 (production)
| Project | HTTP avg | HTTP peak | Queue avg | Queue peak |
|---|---|---|---|---|
| farfalla | 23.5 | 177 | 0.5 | 37 |
| coniglio | 7.0 | 27 | 6.5 | 16 |
| castoro | 0.4 | 42 | - | - |
| farfalla-https-guard | 0.3 | 28 | - | - |
| farfalla-integrations | 0.05 | 17 | - | - |
| medusa | 0.03 | 20 | 0.6 | 6 |
| Total | ~31 | ~310 | ~8 | ~60 |
Traffic Patterns​
- Peak hours: 01:00-03:00 UTC (Americas evening)
- Spikes can reach 7-8x average
- Weekday/weekend: only 3.6% drop
Target (Post Customer Migration)​
info
Timeline: April 2026 - Oct 2027 (~18 months gradual onboarding)
Reference: K6 Load Test (Nov 25, 2025) - 3,600 VUs, 12min
| Project | HTTP avg | HTTP peak | Queue avg | Queue peak |
|---|---|---|---|---|
| farfalla | ~350 | ~1,200 | ~1 | ~5 |
| coniglio | ~100 | ~400 | ~20 | ~100 |
| castoro | ~5 | ~50 | - | - |
| others | ~5 | ~10 | ~1 | ~5 |
| Total | ~450 | ~1,500 | ~22 | ~110 |
Load Test Results​
- Peak RPS achieved: 1,199 req/sec
- Target capacity: ~1,500 req/sec
- P95 response time: 311ms
- Error rate: 0.004%