Skip to main content
Some Datadog dashboard widgets have a close match in Chronosphere Observability Platform standard dashboards. Recreate them with panels on a standard dashboard, not a classic dashboard. Before you rebuild visuals, translate Datadog queries to PromQL. For more information, see Querying metrics, Mapping, and Migration considerations.

​
Widget overview

Datadog widgetObservability Platform panelNotes
Time series (stacked bars)Time series chartPlots values across the dashboard time range
Top listBar chartSet Query type to Instant in Options
Top list (tabular)TableUse when you need columns or multiple calculations
Query valueStat chartSingle aggregated value
Observability Platform doesn’t directly import Datadog dashboard JSON. Rebuild panels manually or with your migration plan and tooling.

​
Stacked bar graph

Datadog time series widgets with stacked bars show values over time. Match them with a Time series chart panel.
  1. Create a panel of type Time series chart, or edit an existing panel and change its type to Time series chart.
  2. In the Query tab, enter a PromQL query for the series you want to stack.
  3. In the Settings tab, in the Visual section:
    • Set Display to Bar.
    • Set Stack series to All for stacked bars, or to Percent for percent-stacked bars.
  4. Optional: In the Settings tab, in the Tooltip section, set Sort to Descending instead of Auto or None.
  5. In the Edit panel interface, click Apply.
  6. On the dashboard, click Save.
After you convert the Datadog query, the following example shows one pattern for stacked values over time. 
sum by (LABEL_NAME) (rate(METRIC_NAME[5m]))
Replace the following:
  • METRIC_NAME: Your Prometheus metric name.
  • LABEL_NAME: The label you group by.

​
Top list

Datadog top list widgets rank series by a single value at a point in time. The closest match is a Bar chart panel.
  1. Create a panel of type Bar chart, or edit an existing panel and change its type to Bar chart.
  2. In the Query tab:
    • In Options, set Query type to Instant.
    • Enter a PromQL query with one value per series.
  3. In the Settings tab, in the Misc section, set Mode to Value for absolute values, or to Percentage for values as a percent of the total.
  4. Optional: In the Settings tab, in the Misc section:
    • Set Sort to Descending (value) if it isn’t already. New bar chart panels default to descending sort.
    • Set Limit to cap how many bars display. If your query uses topk, set Limit to the same value as the topk argument, or omit topk from the query and use Limit only.
  5. In the Edit panel interface, click Apply.
  6. On the dashboard, click Save.
Use an instant query that returns one value per series. The topk argument limits how many series the panel ranks, similar to a Datadog top list limit. The following example shows one pattern for a ranked instant query. 
topk(10, sum by (LABEL_NAME) (rate(METRIC_NAME[5m])))
Replace the following:
  • METRIC_NAME: Your Prometheus metric name.
  • LABEL_NAME: The label you group by.

​
Top list as a table

Use a Table panel when you need label columns, multiple calculations, or a sortable grid instead of horizontal bars.
  1. Create a panel of type Table, or edit an existing panel and change its type to Table.
  2. In the Query tab:
    • In Options, set Query type to Instant.
    • Enter a PromQL query with one value per series.
  3. In the Settings tab:
    • In the Data Format section, set Time series format to Series To Rows.
    • Under Calculations, select Total.
    • In the Sorting section, set Sort by to Total and Sort order to Descending.
  4. Optional: In the Settings tab, in the Data Format section, enable Label-based columns to show each series’s labels.
  5. In the Edit panel interface, click Apply.
  6. On the dashboard, click Save.
Use an instant query that returns one value per series. The topk argument limits how many rows display, similar to a Datadog top list limit. The following example shows one pattern for a ranked instant query. 
topk(10, sum by (LABEL_NAME) (rate(METRIC_NAME[5m])))
Replace the following:
  • METRIC_NAME: Your Prometheus metric name.
  • LABEL_NAME: The label you group by.