This document establishes the official Service Level Objectives (SLOs) categorized into Product SLOs, Runtime SLOs, and Developer SLOs.
1. Product SLOs (Customer First-Run Experience)
| Metric |
Target |
Measurement Method |
CI Regression Gate |
| Install to First Deploy |
< 5 minutes |
End-to-end VPS onboarding walkthrough |
Manual / E2E Gate |
| Runtime Detection |
< 250 ms |
Indicator file scan & detector pipeline |
Hard Gate |
Config Generation (aegis.toml) |
< 100 ms |
Zero-boilerplate config synthesis duration |
Hard Gate |
2. Runtime SLOs (Production Supervision Engine)
| Metric |
Target |
Measurement Method |
CI Regression Gate |
| Daemon Cold Startup |
< 500 ms |
Time from binary launch to gRPC server bind |
Hard Gate |
| Process Crash Restart |
< 2.0 s |
Time from process exit to supervisor re-spawn |
Hard Gate |
| Health Probe Execution |
< 100 ms |
HTTP/TCP probe latency |
Hard Gate |
| Event Propagation Latency |
< 10 ms |
Time from EventBus.publish() to receiver |
Hard Gate |
| Metric |
Target |
Measurement Method |
CI Regression Gate |
| CLI Command Response |
< 100 ms |
Execution latency of aegis status / list / events |
Hard Gate |
aegis doctor Execution |
< 500 ms |
Full environment & daemon check duration |
Hard Gate |
aegis validate Execution |
< 500 ms |
Pipeline readiness & config check duration |
Hard Gate |
| TUI Dashboard Refresh Rate |
60 FPS (~16.6 ms/frame) |
Render frame draw duration |
Soft Gate |
4. Benchmark Integration
Performance benchmarks located in benchmarks/benches/event_throughput.rs are executed in CI to track latency drift over time.