Overview
Monitoring shows graphs of what runs: CPU, memory, requests, queue lengths. Each machine has its own dashboard, and you can build dashboards of your own.
When to use it
- Is an app busy or idle?
- What uses the memory?
- Keep the graphs you watch on one page.
How it works
- Built-in sampler
- The control plane samples every resource every 30 seconds from the engine and QEMU — no extra service to run.
- SQLite
- Samples are kept in the control plane’s own store.
Containers report one-shot stats; counters become rates. Dashboards are resources: graphs of one metric across up to eight resources.
Architecture
- Containers and VMs to Sampler: stats
- Sampler to SQLite
- SQLite to Dashboards
Flows
A dashboard
- Create a dashboard with a graph per metric.
- Choose the resources for each graph.
- Pick a time range, from 1 hour to 30 days.
Set it up
In the console
- Open Monitoring and choose This machine.
- Choose New dashboard to build your own.
With the CLI
enclave dashboard create health --widget "title=API CPU,metric=cpu,resources=app/api" --waitWith OpenTofu
OpenTofu configuration
resource "enclave_dashboard" "health" {
name = "health"
widgets = [
{ title = "API CPU", metric = "cpu", resources = ["app/api"] },
]
}Settings reference
Every setting and output of the OpenTofu resource; the console and the CLI use the same names in their own words. Generated from platform 4d30582.
| Name | Type | Use | What it does |
|---|---|---|---|
name | text | required | Name of the dashboard: 3–40 lowercase letters, digits or hyphens, starting with a letter. Changing it replaces the dashboard. |
widgets | list of objects | required | 1–24 graphs, in order. |
project | text | optional | The project to create it in. Defaults to the provider's project (ENCLAVE_PROJECT, else default). Changing it replaces the resource. |
id | text | output | Resource ID. |
Limits and security
Good to know
- Up to eight resources per graph.
Security
- Metrics follow project roles like everything else.