Monitoring IBM LinuxONE LPAR and Peer Pods

Edit online

For the IBM Confidential Computing Containers for Red Hat OpenShift Container Platform (CCCO) peerpod deployments, you can monitor the health, performance, and security of IBM LinuxONE Logical Partitions (LPAR) and the peerpods. This capability applies only to LPARs running Red Hat Enterprise Linux (RHEL) operating system.

This procedure is intended for users with the role Environment Operator.

Prerequisites

  • Podman or Docker
  • Grafana instance must be accessible from the LPAR.
  • Node exporter and Libvirt exporter container images is built for s390x.

Procedure

  1. Create a directory Monitoring for monitoring purposes.
    mkdir Monitoring
  2. Change the directory to Monitoring.
    cd Monitoring
  3. Run the node exporter to fetch the metrics from the LPARs.
    • If you are using docker, run the following command:
      docker run -d --name <docker_container_name> -p 9100:9100 prom/node-exporter
    • If you are using podman, run the following command:
      sudo podman run -d --name <podman_container_name> -p 9100:9100 prom/node-exporter
  4. Run the libvirt exporter to fetch the metrics from the libvirt peer pod VMs.
    • If you are using docker, run the following command:
      docker run -d --name <libvirt_container_name> -p 9177:9177 -v /var/run/libvirt:/var/run/libvirt libvirt_exporter:latest
    • If you are using podman, run the following command:
      sudo podman run -d --name <libvirt_container_name> -p 9177:9177 --security-opt label=disable -v /var/run/libvirt:/var/run/libvirt:z libvirt_exporter:latest
  5. Create prometheus.yaml according to the following example:
    cat << EOF > prometheus.yaml
global:
scrape_interval: 15s

scrape_configs:
- job_name: 'linuxone-lpar'
    static_configs:
    - targets: ['<node_exporter_target_IP>:<port>',<Libvirt_exporter_target_IP>:<port>] (1)

EOF
    1. Specify the IP address and port where the data must be stored.
  6. Run prometheus container.
    • If you are using docker, run the following command:
      docker run -d --name <prometheus_container_name> -p 9091:9090 -v ./prometheus.yml:/etc/prometheus/prometheus.yml prom/prometheus
    • If you are using podman, run the following command:
      sudo podman run -d --name <prometheus_container_name> -p 9091:9090 -v $(pwd)/prometheus.yml:/etc/prometheus/prometheus.yml:z prom/prometheus
  7. Run Grafana.

    You can either use an existing Grafana instance or create a new instance by performing the following steps:

    1. Ensure to use the Grafana image that is compatible with s390x.
    2. Create a Grafana instance.
      • If you are using docker, run the following command:
        docker run -d --name <grafana_container_name> -p 3000:3000 -v grafana-storage:/var/lib/grafana prom/grafana
      • If you are using podman, run the following command:
        sudo podman run -d --name <grafana_container_name> -p 3000:3000 -v grafana-storage:/var/lib/grafana prom/grafana
  8. Download the following JSON files:
  9. Access Grafana at <IP_address>:3000.

    Specify the IP address where Grafana instance is created.

  10. From the GrafanaLabs page, go to Dashboards > New > Import.
  11. Create the dashboards for Node exporter and Libvirt exporter by importing the JSON files respectively.
  12. Once the dashboards are created, you can view the metrics for LPARs and peer pod VMs.

    Example dashboards for the Node exporter and the Libvirt exporters are shown below:

    Figure: Node exporter dashboard

    Figure: Libvirt exporter dashboard