1
0
mirror of https://github.com/Unleash/unleash.git synced 2024-10-18 20:09:08 +02:00
unleash.unleash/website/docs/api/internal/internal-backstage-api.md
Pål-Magnus Slåtto 3edca3efe8
[docs]: Update Prometheus docs with security info (#2246)
## About the changes

Prometheus metrics should not be exposed to the public. Added a note about this to inform people that internal endpoints should be dropped on external access.

## Discussion points

https://unleash-community.slack.com/archives/CGP2MCHPF/p1666688295670459

## Commits

* [docs]: Update Prometheus docs with security info

* Apply suggestions from code review

Co-authored-by: Thomas Heartman <thomas@getunleash.ai>

* Add info about Prometheus security

Co-authored-by: Thomas Heartman <thomas@getunleash.ai>
2022-10-28 09:58:26 +00:00

27 lines
1.3 KiB
Markdown

---
id: internal
title: /internal-backstage/prometheus
---
# Internal Backstage API
`GET http://unleash.host.com/internal-backstage/prometheus`
Unleash uses Prometheus internally to collect metrics. By default, the metrics are available at `/internal-backstage/prometheus`. You can disable this endpoint by setting the `serverMetrics` option to `false`.
Note that it's not recommended to expose Prometheus metrics to the public as of the [Prometheus pentest-report](https://prometheus.io/assets/downloads/2018-06-11--cure53_security_audit.pdf) issue PRM-01-002. Thus, if you want to keep metrics enabled, you should block all external access to `/internal-backstage/*` on the network layer to keep your instance secure.
[Read more about Prometheus](https://prometheus.io/)
## Annotations {#annotations}
Unleash will automatically count all updates for all toggles under the metric name `feature_toggle_update_total`, and the toggle name is will be set as a label value. This information can be used to create annotations in grafana for everytime a feature toggle is changed.
You can use this query in grafana to achieve this:
```
delta(feature_toggle_update_total{toggle="Demo"}[1m]) != bool 0
```
Another useful counter is the `feature_toggle_usage_total` which will give you the numbers for how many times a feature toggle has been evaluated to `active` or not.