Spark Thrift Chargeback Reports

The Spark Thrift Chargeback Reports page enables you to monitor and predict the costs of using shared resources on a cluster over a period of time. You can monitor the costs using the data of reports you create.

note

The default time range is Last 24 hrs. To view statistics from a custom date range, click the icon and select a time frame and timezone of your choice.

Creating Reports

You can create one or more than one report in chargeback reports. To create a report, do the following.

  1. Click Create Report.
  2. Enter details of the following fields.
    • Name: The name of the chargeback report.
    • Compute Cost/Hr: The computing cost of resources per hour (in $).
    • Memory Cost/Hr: The cost of memory used per hour (in $).
    • Groups: Select a group from the following for which you want to monitor the cost. Note: You can select more than one group.
      • User
      • Status
  3. (Optional) Click Apply to see the results in the table below Create Report panel.
  4. Click Save to create a report.
note

You can view your created reports in the left pane of the Chargeback Reports page under Reports panel.

The chargeback report is created.

Chargeback Reports Table

After creating a chargeback report, you can monitor the following metrics.

note

If you choose Status while creating a report, the metrics are grouped by Status.

MetricDescription
StatusThe current status of the queries in Spark Thrift server.
UserThe name of the user executing queries.
Total ApplicationsThe total number of applications used by a query.
VCore SecondsCalculated as the aggregated number of vcores Spark Thrift application allocates x the number of seconds the application has been running.
VCore CostThe cost of vcore consumption (in $).
Memory SecondsCalculated as the aggregated amount of memory (in megabytes) Spark Thrift application has allocates x the number of seconds the application has been running.
Memory CostThe cost of memory consumption (in $).
Total CostThe total combined cost of VCore Cost and Memory Cost.