Operations with Layers Example
This example demonstrates using the three-layer model (code, infra, runtime) to organize operations metrics.
Document Overview
The example includes:
- Layer definitions with golden signals
- Metrics tagged with layer assignments
- Coverage across code, infrastructure, and runtime layers
Key Concepts
Layer Definitions
Each layer defines golden signals pointing to specific metrics:
{
"layers": [
{
"id": "runtime",
"name": "Runtime",
"signals": {
"latency": "metric-runtime-p99",
"traffic": "metric-runtime-rps",
"errors": "metric-runtime-errors",
"saturation": "metric-runtime-cpu"
}
}
]
}
Layer-Tagged Metrics
Metrics include a layer field for classification:
{
"id": "metric-code-lint-errors",
"name": "Linting Errors",
"domain": "operations",
"stage": "build",
"layer": "code"
}
Metrics by Layer
| Layer | Metrics |
|---|---|
| Code | Linting errors, outdated dependencies |
| Infra | Compliance rate, CPU utilization |
| Runtime | P99 latency, RPS, error rate, CPU |
Download
See Also
- Layers Concept - Understanding the three-layer model
- prism layer - CLI commands for layers