Dashboard

Daria Barysheva

The Dashboard page in the ADB Control web interface allows you to display charts for tracking the dynamic changes of various metrics over the selected time period in the ADB clusters connected for monitoring.

Each tab on the page corresponds to a separate chart:

  • Commands — the number of SQL commands in the Running status.

  • Transactions — the number of transactions in the Active status.

  • Nodes — the number of active instances of plan nodes (Broadcast motion, Redistribute motion, and Dynamic seq scan).

  • CPU — total CPU consumption in the ADB cluster (percentage).

  • Memory — the total amount of allocated virtual memory in the ADB cluster (in bytes).

  • Read — the total amount of data read in the ADB cluster (in bytes).

  • Write — the total amount of data written in the ADB cluster (in bytes).

For all charts except Nodes, metrics are collected every 5 seconds, and the smallest time interval shown is 5 seconds. For the Nodes chart, the number of active node instances is collected once per minute. If your query runs for less than a minute and its execution node is not active at the moment of the metric collection, information about that node will not appear on the Nodes chart.

To view any chart in full-screen mode, click the cron dark cron light icon to the right of the cluster name at the top of the chart.

The Dashboard page
The Dashboard page
The Dashboard page
The Dashboard page

To configure charts, you can use the following fields at the top of the page:

  • Cluster — the ADB cluster for which all charts should be drawn. Initially, the default cluster is selected.

  • Frequency — defines how often to update charts.

  • Time period — the time range to display on the chart. In the window that opens after clicking the field, you can select an interval from the available options or set a custom time range.

  • Type — defines execution nodes to display on the Nodes chart: Broadcast motion, Redistribute motion, or Dynamic seq scan. Select All types to show all three node types.

  • View — defines how to display charts. The field is available on the CPU, Memory, Read, and Write tabs. Possible values:

    • Summary — display a summary chart for the whole ADB cluster.

    • Hosts — display information for each cluster host as a separate chart.

  • Hosts — the drop-down list that is used to select cluster hosts. Becomes enabled if the Hosts value is selected in the View field.

The figure below shows an example of separate charts displayed for individual cluster hosts.

Display a chart in the context of hosts
Display a chart in the context of hosts
Display a chart in the context of hosts
Display a chart in the context of hosts

On each tab, you can see the following information under charts:

  • Current — the last metric value received over the selected time period. If there is a difference in the metric value between the last two points on the time axis, this difference is shown to the right of the Current value.

  • Maximum — the maximum metric value received over the selected time period. By clicking View, you can navigate to the maximum value on the chart.

  • Minimum — the minimum metric value received over the selected time period. By clicking View, you can navigate to the minimum value on the chart.

  • Average — the average value, which is calculated based on all metric values received over the selected time period.

Aggregated metric information
Aggregated metric information
Aggregated metric information
Aggregated metric information

Below, you can see how to highlight the maximum or minimum metric value on the chart by clicking View.

View the maximum metric value on the chart
View the maximum metric value on the chart
View the maximum metric value on the chart
View the maximum metric value on the chart

You can also see a metric value for any specific time point by hovering over the chart.

Found a mistake? Seleсt text and press Ctrl+Enter to report it