# K8s metrics endpoint not working

**URL:** <https://dragonfly.discourse.group/t/k8s-metrics-endpoint-not-working/209>\
**Category:** General\
**Created:** [September 11, 2024, 9:18pm UTC](https://dragonfly.discourse.group/t/k8s-metrics-endpoint-not-working/209 "2024-09-11T21:18:36Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![intj9839](https://avatars.discourse-cdn.com/v4/letter/i/ed655f/32.png) [@intj9839](https://dragonfly.discourse.group/u/intj9839)\
**Post date:** [September 11, 2024, 9:18pm UTC](https://dragonfly.discourse.group/t/k8s-metrics-endpoint-not-working/209/1 "2024-09-11T21:18:37Z")

</div>

kubectl port-forward svc/dragonflydb 6379:6379

curl -v [http://localhost:6379/metrics](http://localhost:6379/metrics)

- Trying [::1]:6379…
- Connected to localhost (::1) port 6379

> GET /metrics HTTP/1.1  
> Host: localhost:6379  
> User-Agent: curl/8.4.0  
> Accept: _/_

- Received HTTP/0.9 when not allowed
- Closing connection  
curl: (1) Received HTTP/0.9 when not allowed

it’s possible port can accept both protocols within container, but for some reason doesn’t allow it when exposed as pod/service port.

Also tried from pod in same namespace:  
root@redis-pod:/data# curl [http://production-raptor-cache:6379/metrics](http://production-raptor-cache:6379/metrics)  
curl: (1) Received HTTP/0.9 when not allowed

---

<div class="post-metadata">

**Author:** ![abhradeep](https://avatars.discourse-cdn.com/v4/letter/a/bc79bd/32.png) [@abhradeep](https://dragonfly.discourse.group/u/abhradeep)\
**Post date:** [September 13, 2024, 8:52am UTC](https://dragonfly.discourse.group/t/k8s-metrics-endpoint-not-working/209/2 "2024-09-13T08:52:08Z")

</div>

Metrics are available on admin port 9999. If you want to setup monitoring, we have some documentation [Integrate Prometheus with the Dragonfly Operator | Dragonfly](https://www.dragonflydb.io/docs/managing-dragonfly/operator/prometheus-guide). You can follow that.

---

<div class="post-metadata">

**Author:** ![intj9839](https://avatars.discourse-cdn.com/v4/letter/i/ed655f/32.png) [@intj9839](https://dragonfly.discourse.group/u/intj9839)\
**Post date:** [September 13, 2024, 10:56pm UTC](https://dragonfly.discourse.group/t/k8s-metrics-endpoint-not-working/209/3 "2024-09-13T22:56:01Z")

</div>

Thank you - I got it working with port 9999 on kubernetes clusterIP service.

> **[Monitoring | Dragonfly](https://www.dragonflydb.io/docs/managing-dragonfly/monitoring)**
>
> By default, Dragonfly allows HTTP access via its main TCP port (i.e, 6379) and it exposes Prometheus compatible metrics on :6379/metrics.

- this page might need edit because **6379** port for svc doesn’t work for /metrics (but the port works on docker container port forward) - must be due to how kubernetes handles ports.