Creating Visualizations

Use Visualize to create and run a variety of charts and graphs for your ChaosSearch indexes.

Kibana offers a powerful interface for turning your search queries into visualizations to show important information in picture or tabular format. Creating visualizations can take some time for new users to learn, especially for developing the analytics that support them. It takes practice to become familiar with the steps and controls to create visualizations for the data of interest.

Each type of visualization shows information in different ways for summary or comparisons. Each visualization (chart) should be designed to answer a question, which can help you to identify the information that you want to show, decisions about the key metrics of interest, and different aggregation types (like sums, averages, counts, min/max, and so on) needed to create the groupings that summarize the important data.

For example, the following visualization is a bar chart that shows a breakdown of the order priorities for orders completed on a monthly basis. You can see how the total count of orders is distributed across priority.

14291429

Saving a visualization of interest allows permitted users to open and use the visualization to see the same summary for a different time range. This can be helpful for detecting trends or changes in patterns, or changes in overall totals. You can export and import visualizations to save them for possible use in another environment, such as to bring favorite ones from development areas to production or to import custom ChaosSearch Customer Success projects.

Display a Visualization

To display a visualization:

  1. In the ChaosSearch console, click Analytics > Visualize. The Visualizations page appears.
24222422
  1. In the Visualizations list, click the title of a visualization to run it.

📘

Visualizations Filter

If the list is very long, you can use the search field to reduce the list. Note that the search value is case-sensitive and must match from the beginning of the visualization name. Strings will not be matched in the middle of the visualization name.

A sample visualization follows.

33163316

❗️

"No results displayed because all values equal 0."

In the case where a visualization is not displayed, check to make sure your time filter is property configured. By default the time range is set to the last 15 minutes.

See setting the time filter for more information.

Within the visualization, you can click within the chart to take other actions like display more information about a data point, change the default time range, and drill down into a closer look at the data.

Aggregations—Kibana's (and Elastic's) Version of GROUP BY

When using SQL queries to analyze data, the GROUP BY syntax is an important and powerful way to summarize data into categories. For example, analysts might want to see these types of summaries:

  • Outages grouped by service to identify most problematic services
  • Sales grouped by representative or by region over time to trend revenue forecasts
  • Service tickets grouped by resolution time buckets to determine service level agreement adherence

In SQL, GROUP BY is the way to collapse data into the requested buckets to enable these types of summaries.

In Kibana, aggregations are the way to collapse the data into requested buckets to enable visualization of these types of summaries.

In the Kibana Visualizations page, you can create and use aggregations to specify how to combine records in the ChaosSearch index data to answer analysis questions for yourself or your users.

For example, the TPC-H orders table information contains details about orders for various items, associated with various clerks, with information about order keys, order priorities, ship priorities, costs, and so on. A sample TPC-H record follows:

_id:1173 o_clerk:Clerk#000000175 o_comment:side of the bold platelets detect slyly blithely ironic e o_custkey:88 o_orderdate:Aug 1, 1998 @ 20:00:00.000 o_orderkey:4,678 o_orderpriority:4-NOT SPECIFIED o_orderstatus:O o_shippriority:0 o_totalprice:131,752.07 _type:doc _index:orders-view _score:1

One possible orders-related summary might be:

What is the distribution of order priorities for the orders in a specific time range?

You can create a visualization of that data by aggregating records from the indexed data by order priority.

Create a Visualization

To create a visualization:

  1. On the Analytics page, click Visualize.
12231223
  1. In the Visualizations area, click Create visualization.
  2. In the New Visualization window, select a visualization type. This example uses a Data Table.
16441644

👍

Chart Types

Kibana offers many charts and graphs for visualizations. Hover over any of the types int he type chooserfor more information.

  1. In the New Data Table / Choose a source page, select the Refinery view that you want to use as the source for your visualization. This example uses orders-view.
883883
  1. In the Visualize/Create window, select a time range for your visualization.
  2. In the right menu under Metrics, use the default Metric Count because this summary will use a total count of the orders by priority. (There are other metric aggregation types like max, min, average, and sum.)
  3. Under Buckets, click Add and then click Split rows. (This aggregation will split the total count of orders by rows for each order priority.)
12281228
  1. In the Aggregation menu, select Terms to select a specific column/field value.
  2. In the Field list, select the column value (in this case, o_orderpriority).
12431243
  1. Choose options for the data display. In this example, select Order by Alphabetical, Order Ascending, and type a Column label of Priority.
  2. Click Update to display the visualization.
13131313

The resulting data table shows the orders for the time range grouped by order priority. There is a count of the number of hits for each priority. If this is a visualization that you would want to review periodically, you can save the visualization. You can then re-open it and set time ranges as desired to see the data. You could also use the saved visualization in a dashboard.

If you wanted an even more granular grouping—such as orders by priority broken down into order status value —you can easily add another bucket to split the rows by that additional summary. The split row process is identical.

  1. In the Buckets area, click Add at the bottom to add another grouping/aggregation bucket.
  2. For this case, add a Split rows bucket for the field o_orderstatus, similar to the following:
12991299
  1. Click Update to render the new visualization. The data table now shows each order priority broken down by the order status code (O,F,P) for the time range.

Always save your visualization if you or other users would want to review it periodically, or to use it in a dashboard.

Split Rows Versus Split Visualization

Kibana aggregations support the ability to split by rows (as in the table example above), and the ability to split the visualization into separate displays by summary using the Split option. When you split by the visualization type, you create separate visualizations for each summary. For example, the Split table creates a table for each sub-bucket that you define, for example:

13661366

Likewise, if you add a Split rows definition to split priority rows by order status, you would see separate data for each status code in each priority table:

13711371

View Detailed Information

Hover over any of the visualization chart components (such as pie slices in a pie chart) to see detailed information including field value, term count, as well as overall percentage. In this example, the detailed information shows that there are 118 instances of S3, making up 17.1% of total occurrences.

887887

Apply Filters to a Visualization

There are two ways to add filters to your visualization:

From the Legend or Chart

Click a part of the chart (such as a pie slice) or a legend for some charts to add that option to the filter. Click x to remove that filter if needed.

From the Add Filter Pop-up

Click Add filter to add filters to the visualization. Enter the field and operator criteria, and optionally a label, and click Save.

834834

Filter Actions

When a filter is defined, it is displayed beneath the search bar.

10221022

Click on the filter for the following available options:

12541254
  1. Pin across all apps – Click to "pin" the filter. Pinning a filter keeps the filter across various screens (Visualize, Discover, and so on).
  2. Edit filter – Click to display the Edit Filter window and change the filter.
  3. Exclude results – Click to switch between a positive and negative filter. For example, if you are filtering for a certain field, this setting reverses the action to show all fields except for the filter.
  4. Temporarily disable – Click to toggle between enabled and disabled. A quick way to see the visualization without the filter but without deleting the filter.
  5. Delete – Click to remove the filter.

👍

More information

See the Kibana documentation for more information about configuring visualization types through the Visualize screen.


What’s Next

Navigate to the Dashboard screen

Did this page help you?