Monitoring Watson OpenScale

Important: IBM Cloud Pak® for Data Version 4.7 will reach end of support (EOS) on 31 July, 2025. For more information, see the Discontinuance of service announcement for IBM Cloud Pak for Data Version 4.X.

Upgrade to IBM Software Hub Version 5.1 before IBM Cloud Pak for Data Version 4.7 reaches end of support. For more information, see Upgrading IBM Software Hub in the IBM Software Hub Version 5.1 documentation.

To monitor and allocate resources for Watson OpenScale for IBM Cloud Pak for Data, use the Red Hat® OpenShift® grafana dashboard.

Procedure

  1. Log in.
    1. To log in, from the Red Hat OpenShift Container Platform Cluster Console, click Monitoring > Dashboards .
    2. Then, select the dashboard named K8s / Compute Resources / Namespace.
    3. In the dashboard, change the "namespace" selector to namespace1, which is the default namespace for Cloud Pak for Data, or whatever your namespace for Cloud Pak for Data is.
      Watson OpenScale containers can be found with the prefix aiopenscale.
  2. Interpret the dashboard.
    1. Use the grafana dashboard to monitor cluster total usage, CPU, and memory use:
      • CPU Usage

        The CPU Usage section provides a quick overview of the total IBM Cloud Pak for Data cluster resource utilization trends. To understand the resource utilization trends for the pods running in Watson OpenScale, use the dashboard to check for the current usage and recent trends for CPU and memory.

      • CPU Quota

        The CPU Quota detail section enables you to look more closely at how Watson OpenScale services are using CPU. Zoom into the trend for the Bias Service, where you can see the split between the actual CPU usage for the service and the currently-configured maximum CPU usage allowed for the Bias Service. An administrator can watch the trends of CPU usage, observe whether the actual production usage is trending toward the maximum, and can use these trends decide whether and when to increase the maximum CPU allocation for the service. The same method can be use for trends in memory usage for the Watson OpenScale services. For example, you can monitor usage to ensure that it stays in the healthy range.

      • Memory Usage

        The Memory Usage section provides a quick overview of the total IBM Cloud Pak for Data memory use. To understand the memory usage trends for the pods running in Watson OpenScale, use the dashboard to check for the current usage and recent trends for memory.

      • Memory Quota

        The Memory Quota detail section enables you to look more closely at how Watson OpenScale services are using memory. Zoom into the trend for the Bias Service, where you can see the split between the actual memory usage for the service and the currently-configured maximum memory usage allowed for the Bias Service. An administrator can watch the trends of memory usage, observe whether the actual production usage is trending toward the maximum, and can use these trends decide whether and when to increase the maximum memory allocation for the service.

Results

The operational data that you gather from the grafana dashboard can be especially useful in scaling your instance of Watson OpenScale.

What to do next

For full details on how to set alerts, see the grafana documentation.