Contents x
    See current filesystem utilization - Kubernetes
    • 2 Minutes to read
    • Print
    • PDF
    Contents

    See current filesystem utilization - Kubernetes

    • 2 Minutes to read
    • Print
    • PDF
    Article summary

    This topic describes various ways to monitor PVs Autoscaling.

    Click the link for more information about the desired monitoring method:

    Monitor from the Zesty platform

    Prerequisites

    To monitor Zesty Disk from the Zesty platform:

    1. Log in to the Zesty platform at https://app.zesty.co

    2. From the main menu, select Zesty Disk.

    3. From the Managed filesystems tab, review the filesystems that are managed by Zesty Disk, marked as

    4. Select a filesystem and review filesystem data:

      • In the Metrics tab, you can see the capacity, IOPS, and throughput changes.

      • In the Tags tab, you can see instance and volume tags.

        Note the

        “kubernetes.io/created-for/pvc/name” tag whose value is the Zesty PVC name.

    Monitor from Prometheus

    To monitor using Prometheus you need to install the PVs Autoscaling with the Prometheus exporter option. Once installed, Zesty automatically adds the annotations below to the Zesty Storage Agent so it can be discovered by Prometheus.

    For more details, see Deploy Zesty Disk for K8s.

    prometheus.io/port=<PORT_FROM_HELM_VALUES>
prometheus.io/scrape=true

    Available queries

    You can query the following utilization parameters:

    • Total PVC capacity
      label_replace(sum(kubelet_volume_stats_capacity_bytes) / 1024 / 1024 / 1024, "metric", "Total PVC's capacity", "", "")

    • Total PVC used size
      label_replace(sum(kubelet_volume_stats_used_bytes) / 1024 / 1024 / 1024, "metric", "Total PVC's used size", "", "")

    • PVC average utilization
      label_replace(avg(kubelet_volume_stats_used_bytes / kubelet_volume_stats_capacity_bytes) * 100, "metric", "PVC's avg utilization", "", "")

    • Amount of PV in cluster
      label_replace(count(kube_persistentvolume_info{csi_driver="ebs.csi.aws.com", job="kubernetes-services"}), "metric", "Amount of PV's in cluster", "", "")

    • Average PV size
      label_replace(avg(kubelet_volume_stats_capacity_bytes) / 1024 / 1024 / 1024, "metric", "Average PV Size", "", "")

    • Amount of StatefulSets in cluster
      label_replace(count(kube_statefulset_persistentvolumeclaim_retention_policy{job="kubernetes-services"}), "metric", "Amount of StatefulSets in cluster", "", "")

    Monitor from Grafana

    Zesty provides a custom Grafana dashboard that shows PVs Autoscaling PVCs metrics.

    To import the dashboard to your Grafana, use the following link: https://static.zesty.co/ZestyDiskPVC/zesty-pvc-grafana-dashboard.json

    Monitor with the Kubectl plugin

    You can use the Kubectl plugin to generate a CSV with relevant workload and storage information for a Kubernetes cluster.

    The CSV provides insights into workload types, storage capacities, your storage footprint and more. The CSV includes information about the workload name and type, the Pod name, PV name and size, usage, and utilization.

    The CSV output is sorted by PV size (descending). PVs Autoscaling PVs are not included.

    To install Krew, see Windows or MacOS.

    To use the plugin, you need to install it first.

    Prerequisites

    • Kubectl is installed

    To install the Zesty Kubectl plugin:

    1. Add the Zesty repository to Krew index:

      kubectl krew index add zestyIndex https://github.com/zesty-co/kubectl-plugin.git

    2. Verify the repo was added successfully:

      kubectl krew index list

    3. Install the plugin:

      kubectl krew install zestyIndex/zesty

    4. Verify that the plugin was installed successfully:

      kubectl krew list

    5. Validate the Kubectl context:

      kubectl config current-context

      The current cluster’s name will be displayed.

    To use the kubectl plugin:

    The following table describes the plugin commands to export information about cluster PVs to a CSV:

    To export this

    Run this command

    Example

    For PVs in all namespaces, with the default output location

    kubectl zesty generate pv-summary

    kubectl zesty generate pv-summary

    For PVs in a specific namespace

    kubectl zesty generate pv-summary --namespace=<name>

    kubectl zesty generate pv-summary --namespace=default

    To specify an output path

    kubectl zesty generate pv-summary --output=<path>.csv

    kubectl zesty generate pv-summary --output=/custom/path/report.csv

    For PVs in a specific namespace and output path

    kubectl zesty generate pv-summary --namespace=<name> --output=<path>.csv

    kubectl zesty generate pv-summary --namespace=default --output=/custom/path/report.csv

    Additional filters

    --workload-type=<type> \

    --min-size=<X>Gi \

    --max-size=<X>Gii \

    --workload-type=StatefulSet \

    --min-size=1Gi \

    --max-size=100Gi \

    Was this article helpful?
    What's Next