Dashboard

The Varada Control Center’s main dashboard gives your data platform team a high-level picture of the activity in your cluster, together with actionable insights based on workload usage patterns.

Click the D (day), W (week), M (month), or Q (quarter) option at the top-right of the Dashboard page to change the reporting period. The reporting period is applied to all the widgets on the dashboard.

The dashboard includes the following widgets:

Top Workloads

The Top Workloads widget shows the behavior of the top five workloads by compute utilization.

In this widget:

  • The X-axis shows time.
  • The left Y-axis shows the compute utilization relative to the total available compute resources.
  • The right Y-axis shows the number of nodes, from 0 to the maximum number of nodes. The dashed yellow trendline in the graph shows the number of nodes in use.

👍

Click a workload in the legend to remove that workload from the graph, so that you can focus on other workloads. Click the workload again to add it back to the graph.

Top Users

The Top Users widget shows the top five users by compute utilization. All the remaining users are aggregated in a sixth user called Other.

In this widget:

  • The X-axis shows percentage values.
  • The Y-axis shows user names.

Two bars are displayed for each user. Queries shows the percentage of queries the user executed out of the total number of queries executed. CPU shows the user's compute utilization out of the total available compute resources.

Varada Impact

The Varada Impact widget shows the percentage of splits that utilized Varada's acceleration layer during query processing out of the total splits processed.

Top Tables

The Top Tables widget shows the top five tables by compute utilization for table scans. All the remaining tables are aggregated in a sixth table called Other.

In this widget:

  • The X-axis shows percentage values.
  • The Y-axis shows the table names.

A bar is displayed for each table, showing the compute utlization for scanning that table out of the total compute resources for scanning tables. Exploring Tables

👍

Click the title of the Top Tables widget to go to the Tables page to view additional details about the workloads hitting the table, as well as the table's column acceleration details.