SAS Viya Monitoring for Kubernetes has introduced two features in version 1.2.36 (15th April 2025) aimed at simplifying deployment; auto-generating Ingress definitions and automatically specifying storage class references. These enhancements reduce the need for manual configuration, making setup easier and more efficient.
How Ingress is Used in SAS Viya Monitoring
Ingress resources in Kubernetes manage external access to services within the cluster, typically HTTP and HTTPS traffic. In SAS Viya Monitoring, Ingress facilitates access to web-based monitoring tools like Grafana and OpenSearch Dashboards (OSD). Properly configured Ingress ensures that users can securely and efficiently interact with these interfaces.
There are two primary approaches to Ingress configuration:
Host-Based Ingress: Each application is accessed via a unique subdomain.
Example:
- Grafana: https://grafana.gelcorp.com/
- OpenSearch Dashboards (OSD): https://osd.gelcorp.com/
Path-Based Ingress: A single domain is used, with applications differentiated by URL paths.
Example:
- Grafana: https://monitoring.gelcorp.com/grafana
- OpenSearch Dashboards (OSD): https://monitoring.gelcorp.com/osd
The choice between host-based and path-based Ingress depends on your organisation's policies. Detailed guidance on configuring Ingress for SAS Viya Monitoring is available in the official documentation.
Auto-Generated Ingress Definitions
Previously, configuring Ingress resources required manual editing of YAML files. With this new feature, users can automate the creation of these definitions by setting environment variables, preferably by including them in the $USER_DIR/user.env file. See the Pre-Deployment topic in the Help Center documentation for information on the $USER_DIR directory and more information about the SAS Viya Monitoring customisation process. The system supports both host-based and path-based Ingress configurations, aligning with common deployment scenarios.
To enable auto-generated Ingress definitions:
- Set the environment variable AUTOGENERATE_INGRESS=true.
- Specify your Ingress type by setting ROUTING to either host or path.
- Specify the hostname to be used in the URLs by setting the environment variable BASE_DOMAIN. This is required.
- Apply your configuration as usual, and the system will generate the appropriate Ingress definitions automatically.
If you provide your own TLS certificate files, it's also possible for the autogeneration process automatically create corresponding Kubernetes secrets needed by the Ingress resources. Two further variables are required for this:
- Set INGRESS_CERT to point to the location of the TLS certificate file.
- Set INGRESS_KEY to the TLS key file.
How Storage Classes Are Used in SAS Viya Monitoring
Persistent storage is crucial for retaining monitoring data, logs, and metrics across pod restarts and deployments. Kubernetes uses Persistent Volume Claims (PVCs) to request storage resources, and storage classes define the types of storage available for these PVCs.
In SAS Viya Monitoring, various components require persistent storage to function correctly. For instance, Prometheus needs storage to retain metric data, and OpenSearch requires storage for log data. Configuring appropriate storage classes ensures that these components have access to reliable storage resources.
Key considerations for storage configuration include:
- Durability: Ensure that data persists beyond the lifecycle of individual pods by using persistent volumes external to the Kubernetes cluster. This approach safeguards data integrity during scaling or pod failures.
- Performance: Select storage classes that meet the performance requirements of your monitoring workloads, considering factors like input/output operations per second (IOPS) and latency.
- Capacity Planning: Allocate sufficient storage capacity to accommodate the volume of monitoring data generated, with appropriate provisions for future growth.
For comprehensive information on storage requirements and configuration in SAS Viya Monitoring, refer to the official documentation.
Automatic Storage Class References
Specifying storage classes for PVCs traditionally involved modifying YAML files manually. The latest update introduces an automated approach, generating the necessary storage class references based on predefined environment variables, preferably by including them in the $USER_DIR/user.env file. See the Customization topic in the Help Center documentation for information on the customisation process. While this simplifies deployment for standard configurations, manual customisation remains an option for more complex setups.
To enable automatic storage class references:
- Set the environment variable AUTOGENERATE_STORAGECLASS=true.
- Define your preferred storage class by setting STORAGECLASS to the appropriate value for your environment. It is important to note that the storageClass specified must exist; SAS Viya Monitoring will not create any storageClass resources.
- The system will apply this storage class to all relevant PVCs at deployment time without requiring manual edits.
Next Steps
These new capabilities provide a streamlined way to configure Ingress and storage in SAS Viya Monitoring. By reducing manual effort and automating common setup tasks, they make deployment more efficient while still allowing for customisation where needed. Refer to the the Help Center for more information about customising the process to suit your needs.
To get started, ensure you have yq version 4.32.2 or later (which will be a requirement for future releases of SAS Viya Monitoring for Kubernetes). Detailed instructions, including configuration examples, can be found in the documentation.
My thanks to Greg Smith for his contributions to this post.
Find more articles from SAS Global Enablement and Learning here.
