Exposing SUSE Observability outside of the cluster
SUSE Observability Self-hosted
Last updated
SUSE Observability
SUSE Observability Self-hosted
Last updated
SUSE Observability can be exposed with a Kubernetes Ingress resource. The example on this page shows how to configure an nginx-ingress controller using . This page also documents which service/port combination to expose when using a different method of configuring ingress traffic.
When observing the cluster that also hosts SUSE Observability, the agent traffic can be kept entirely within the cluster itself by during agent installation.
The SUSE Observability Helm chart exposes an ingress section in its values. This is disabled by default. The example below shows how to use the Helm chart to configure an nginx-ingress controller with TLS encryption enabled. Note that setting up the controller itself and the certificates is beyond the scope of this document.
To configure the ingress for SUSE Observability, create a file ingress_values.yaml with contents like below. Replace MY_DOMAIN with your own domain (that's linked with your ingress controller) and set the correct name for the tls-secret. Consult the documentation of your ingress controller for the correct annotations to set. All fields below are optional, for example, if no TLS will be used, omit that section but be aware that SUSE Observability also doesn't encrypt the traffic.
Note that setting up TLS is required for the use of the rancher UI extension.
What stands out in this file is the Nginx annotation to increase the allowed proxy-body-size to 50m (larger than any expected request). By default, Nginx allows body sizes of maximum 1m. SUSE Observability Agents and other data providers can sometimes send much larger requests. For this reason, you should make sure that the allowed body size is large enough, regardless of whether you are using Nginx or another ingress controller. Make sure to update the baseUrl in the values file generated during initial installation, it will be used by SUSE Observability to generate convenient installation instructions for the agent.
Include the ingress_values.yaml file when you run the helm upgrade command to deploy SUSE Observability:
The SUSE Observability Helm chart exposes an opentelemetry-collector service in its values where a dedicated ingress can be created. This is disabled by default. The ingress needed for opentelemetry-collector purposed needs to support GRPC protocol. The example below shows how to use the Helm chart to configure an nginx-ingress controller with GRPC and TLS encryption enabled. Note that setting up the controller itself and the certificates is beyond the scope of this document.
To configure the opentelemetry-collector ingress for SUSE Observability, create a file ingress_otel_values.yaml with contents like below. Replace MY_DOMAIN with your own domain (that's linked with your ingress controller) and set the correct name for the otlp-tls-secret. Consult the documentation of your ingress controller for the correct annotations to set. All fields below are optional, for example, if no TLS will be used, omit that section but be aware that SUSE Observability also doesn't encrypt the traffic.
What stands out in this file is the Nginx annotation to increase the allowed proxy-body-size to 50m (larger than any expected request). By default, Nginx allows body sizes of maximum 1m. SUSE Observability Agents and other data providers can sometimes send much larger requests. For this reason, you should make sure that the allowed body size is large enough, regardless of whether you are using Nginx or another ingress controller. Make sure to update the baseUrl in the values file generated during initial installation, it will be used by SUSE Observability to generate convenient installation instructions for the agent.
Include the ingress_otel_values.yaml file when you run the helm upgrade command to deploy SUSE Observability:
To make SUSE Observability accessible outside of the Kubernetes cluster it's installed in, it's enough to route traffic to port 8080 of the <namespace>-stackstate-k8s-router service. The UI of SUSE Observability can be accessed directly under the root path of that service (i.e. http://<namespace>-stackstate-k8s-router:8080) while agents will use the /receiver path (http://<namespace>-stackstate-k8s-router:8080/receiver).
Make sure to update the baseUrl in the values file generated during initial installation, it will be used by SUSE Observability to generate convenient installation instructions for the agent.
SUSE Observability itself doesn't use TLS encrypted traffic, TLS encryption is expected to be handled by the ingress controller or external load balancers.
The example uses the ingressClassName field to specify the instead of the deprecated kubernetes.io/ingress.class annotation. If your cluster has a default ingress class defined the ingerss class name field can be omitted.
This step assummes that was already executed.
The example uses the ingressClassName field to specify the instead of the deprecated kubernetes.io/ingress.class annotation. If your cluster has a default ingress class defined the ingerss class name field can be omitted.
This step assummes that was already executed.
Agents that are deployed to the same cluster as SUSE Observability can of course use the external URL on which SUSE Observability is exposed, but it's also possible to configure the agent to directly connect to the SUSE Observability instance via the Kubernetes internal network only. To do that replace the value of the 'stackstate.url' in the helm install command from the with the internal cluster URL for the router service (see also above): http://<namespace>-suse-observability-router.<namespace>.svc.cluster.local:8080/receiver/stsAgent (the <namespace> sections need to be replaced with the namespace of SUSE Observability).
(not using nginx)
(using nginx)