Table

The Table panel displays your queries in tabular form. Use Table panels for quantitative analysis where you want to see actual values instead of visual representations. Similar to a spreadsheet, you can look at a combination of metric values and segments. This is useful when you don’t necessarily care about the change of a metric over time, or want to run reports to download as CSV/JSON for offline analysis.

Use Table Panel

The PromQL panel supports the configuration of dynamic links. You can generate an external or internal URL using the label values returned by the PromQL query.

To add a link, open the Links tab and enter:

  • the URL template. In the URL you can reference a label value with {{label_name}}.
  • the link title.

If your URL contains the { and } character, escape it with \.

Use Cases

Build an Aggregate View of Your Workloads Per Cluster and Namespace

This example shows how to build a Form table panel to view a concatenated list of workloads per cluster and namespace.

When configuring the panel options, choose:

  • Segmentation: kube_cluster_name and kube_namespace_name
  • Metrics/Labels: kube_workload_name
  • Time aggregation: Concat

You can then export the results as CSV for further exploration.

View Resource Utilization of Your Kubernetes Workload Using a PromQL Table Panel

This example shows how to build a PromQL table panel to view resource utilization of your Kubernetes workload.

For demonstration, CPU and memory utilization is evaluated against the given quota in each workload, cluster, and namespace. The query displays top 10 segments that consume highest number of CPU cores and highest amount of memory.

  • CPU Used Percent

    topk(10,avg(avg_over_time(sysdig_container_cpu_used_percent{$__scope}[$__interval])) by (kube_cluster_name, kube_namespace_name, kube_workload_name))

  • CPU Quota

    topk(10,avg(avg_over_time(sysdig_container_cpu_cores_quota_limit{$__scope}[$__interval])) by (kube_cluster_name, kube_namespace_name, kube_workload_name))

  • Memory Used Percent

    topk(10,avg(avg_over_time(sysdig_container_memory_used_percent{$__scope}[$__interval])) by (kube_cluster_name, kube_namespace_name, kube_workload_name))

  • Memory Quota

    topk(10,avg(avg_over_time(sysdig_container_memory_limit_bytes{$__scope}[$__interval])) by (kube_cluster_name, kube_namespace_name, kube_workload_name))

Select an appropriate unit of value for each query to return meaningful result. For example, use number for CPU quota, data for memory quota, and percent resource utilization metrics.

To scope down your selection, use the scope variable in combination with Team scope.

In this example, the panel scope is limited to two clusters (netsec-load-data, test-k8s-data), one namespace (kube-system), and one workload (coredns).

Explore Kubernetes Objects from a PromQL Table

This example shows how to configure the table links to explore a Kubernetes object on Metrics Explorer:

  1. Use the URL that gives the Metrics Explorer default All Workloads:
#/explore/metrics?last=3600&scope=kubernetes.cluster.name%20%3D%20%3F%20and%20kubernetes.namespace.name%20%3D%20%3F%20and%20kubernetes.workload.name%20%3D%20%3F%20and%20kubernetes.pod.name%20%3D%20%3F
  1. Apply the labels to the URL:

  • Show cluster in Metrics Explorer:

    #/explore/metrics?last=3600&scope=kubernetes.cluster.name+%3D+"{{kube_cluster_name}}"+and+kubernetes.namespace.name+%3D+%3F+and+kubernetes.workload.name+%3D+%3F+and+kubernetes.pod.name+%3D+%3F

  • Show namespace in Metrics Explorer:

    #/explore/metrics?last=3600&scope=kubernetes.cluster.name+%3D+"{{kube_cluster_name}}"+and+kubernetes.namespace.name+%3D+"{{kube_namespace_name}}"+and+kubernetes.workload.name+%3D+%3F+and+kubernetes.pod.name+%3D+%3F

  • Show workload in Metrics Explorer:

    #/explore/metrics?last=3600&scope=kubernetes.cluster.name+%3D+"{{kube_cluster_name}}"+and+kubernetes.namespace.name+%3D+"{{kube_namespace_name}}"+and+kubernetes.workload.name+%3D+"{{kube_workload_name}}"+and+kubernetes.pod.name+%3D+%3F

Show Host Logs on Mezmo

This example shows how to configure the table links to show the host logs on Mezmo:

  1. Use the default URL of the company named Mezmo, where w52qwe12vv is the account id:

    https://app.mezmo.com/w52qwe12vv/logs/view

  2. Apply the host label:

    https://app.mezmo.com/w52qwe12vv/logs/view?hosts={{host_hostname}}

Known Limitations

The PromQL panel column order doesn’t honor the query segmentation order

Features in Technical Preview

The PromQL panel show a maximum of 250 rows. This limit can be changed with a beta flag in the User settings page: