Prometheus API for SLO time series
The Prometheus API for SLO time series is a Prometheus-compatible API that exposes your Nobl9 SLO data as time series. You query it with PromQL against a public metric schema, and consume the results in Grafana, any Prometheus client SDK, or your own programs.
For the full endpoint, parameter, and response reference — including a downloadable OpenAPI specification — see the Prometheus API reference.
The Prometheus API for SLO time series is currently in Beta. The metric schema and the set of implemented endpoints can change. For the current status of each endpoint, see the Prometheus API reference.
What you can do with it
Every value Nobl9 computes for your SLOs is available as a metric: reliability, remaining error budget, burn rate, good and total counts, the raw SLI values, and the per-component metrics of composite SLOs. Because the data is exposed through a standard Prometheus API, you can:
- Build custom Grafana dashboards on top of your SLO data.
- Answer open-ended questions across many SLOs at once with PromQL (for example, "which objectives in this project have less than 20% budget left?").
- Pull SLO values into your own scripts, alerting, or reporting pipelines using a Prometheus client SDK.
Availability
The Prometheus API is available on all Nobl9 production environments. It is a Beta feature; if you don't have access, contact Nobl9 support.
Limitations
A time window is evaluated as a subquery, not read from the stored samples. A window such as
[7d] around a Nobl9 metric doesn't read the samples inside it. The expression is re-evaluated
across the window at a resolution taken from the request rather than from the data — so a range
query's answer changes when its step changes, and in Grafana the width of the panel sets that
step. Write the resolution yourself, as in [7d:1m], to tie the result to your data. See
How a time window is evaluated.
