Skip to content

metering.vcpu_consumption

The metering.vcpu_consumption view contains information about how much vCPU has been consumed by the running clusters.

The amount of vCPU consumed every second by an instance is determined by:

  • The number of clusters in that instance.
  • The number of nodes in each cluster.
  • The number of actual cpu cores that the hardware instance type exposes to each node.

When the user executes any query, vCPU is consumed only when the cluster is resumed. Active resuming turns cluster into a running cluster. There are two events that will cause the consumption to stop counting:

  1. Resume and Suspend: In this case the cluster stays around.
  2. Resume and dropped: In this case the cluster is deleted.

This view contains information about how much vCPU has been consumed by the running clusters.

Column NameData TypeDescriptionExample
metering_hourTIMESTAMPTZThe hour of the metering event. All metering is aggregated to the hour level.2023-12-19T08:00:00Z
interval_startTIMESTAMPTZThe start of the hour in metering_hour.2023-12-19T08:00:00Z
interval_endTIMESTAMPTZThe end of the hour in metering_hour.2023-12-19T09:00:00Z
usernameVARCHARThe user who caused the metering event. Either by starting or altering a cluster.testuser@yellowbrick.com
cluster_nameVARCHARName of the cluster that is running at this metering_hour.default_cluster
cluster_idUUIDThe unique ID number of the cluster that was running at this metering_hour.fb96868a-cbea-422a-99b4-af485e9b43a9
yb_instance_nameVARCHARThe yellowbrick name of the cloud instance that the cluster was created with.small-v1
regionVARCHARCloud region where the cluster is located.east-us
cloud_providerVARCHARThe cloud provider that hosts the cluster.azure
cloud_provider_instance_nameVARCHARThe cloud providers name of the instance used to create the cluster.Standard_L16s_v3
vcpu_secondsINTEGERNumber of vcpu_seconds consumed in the metering_hour.55200