Types of widgets
Types of widgets
Learn here about the different types of widgets that you can use on your dashboards for visually displaying data related to the computation of metrics. Use these types of widgets on dashboards within Basic modules, where you can choose the visualizations of your metrics individually.
You find similar widgets on dashboards within Service Monitoring modules, but the types of widgets and the layout are fixed on these dashboards. On their side, Software Metering modules display only one kind of widget on their dashboards, which is completely different from the widgets on the dashboards of both Basic and Service Monitoring modules.
In this article, we focus on the types of widgets for Basic modules, whose purpose is to display the values of metrics.
A KPI (Key Performance Indicator) widget is mainly a single numerical figure that reflects the inner value of the represented metric (or metrics).
The figure in a KPI widget is either an absolute or a relative number (a percentage). It can be a percentage only if the definition of the associated metric includes a ratio computation; that is, a comparison between the value of the metric itself and the value returned by an additionally supplied investigation (usually a less constrained version of the investigation that computes the value of the metric).
In addition to the main figure, a KPI widget may also include a secondary figure that displays the variation in the value of the metric with respect to its previous value; where the meaning of previous depends on the time frame selected in the dashboard. Moreover, when specified in the definition the associated metric, a KPI widget indicates whether the increase or a decrease in the value of the metric is good or bad by coloring the arrow next to the variation figure in green or in red, respectively. If you decided that a variation in the value of the metric is not necessarily good or bad, the arrow is painted blue.
If the definition of a metric includes thresholds that limit the ranges of values that can be considered normal, worrying (optional), or bad, you can configure the KPI widget to display threshold information in the form of a colored dot that precedes the main KPI figure. The dot is green if the value lies within the normal range, yellow if in the worrying range (which exists only if you define two thresholds), and red if in the bad range.
The following example shows a KPI widget related to a metric that counts the number of devices with CPU issues. The definition of the associated metric includes a ratio (devices with CPU issues compared to the total number of devices) which is the main figure in the widget. An increase in the value of the metric is obviously bad (it implies more devices with CPU issues), so the increase is displayed with a red arrow. Two thresholds are defined: when up to 10% of the total number of devices have CPU issues, the situation is considered normal; from 10% to 20% the situation becomes worrying; more than 20% devices with CPU issues is considered bad.
A single KPI widget can display the values of more than one metric. Arrange the individual visualizations of each metric vertically or horizontally within the KPI widget.
The KPI widget is compatible with count and quantity metrics. Top metrics are not suitable for being displayed as a single figure.
A table widget arranges the value of a metric (or metrics) in a grid. Add a table widget to your dashboards to display either a top metric or a group of count and quantity metrics.
Displaying a top metric
When displaying a top metric in a table widget, the table shows the list of top objects as rows, while the columns are the display fields chosen in the configuration of the top metric. You can limit the number of fields displayed in the configuration of the table widget: you are allowed to turn off only those which are not essential to the definition of the metric.
A single table widget may display at most one top metric, which takes the whole of the widget. Therefore, it is neither possible to combine a top metric with another top metric nor with any other kind of metric in the same table widget.
Count and quantity metrics
The display of count and quantity metrics in a table widget is very flexible. Add up to 50 metrics of these types to a single table widget. For each metric, choose to display either the value of the metric itself or a computed ratio value (when specified in the definition of the metric). If the metric defines thresholds, choose among displaying the status (a green, yellow, or red dot) next to the value, the status alone, or not to display threshold information at all and display only the value. You can also choose to display the variation of the metric with respect to its previous value. These choices are intendedly very similar to those that you can find in the KPI widget.
Arrange the values into the rows and columns of the table according to the hierarchy, the metric names, or the grouping criteria of the metrics that you have added to the widget (if specified in the definition of the metrics). Beware that, depending on the particular metrics that you choose, not all combinations are allowed.
A line chart graphically displays the historical evolution of the value of a metric (or metrics) with time. Line charts let you visually find out significant events in the history of metrics, compare values of different metrics along time and discover trends.
Add up to 5 metrics to a single line chart. All the metrics in the same line chart must be expressed in the same units, where the word unit must be understood in a broad sense. Note, for instance, that you can add count metrics of different objects, such as devices and binaries, to the same line chart, because the units are compatible (they are always a number of objects). You can even mix count metrics with those quantity metrics that measure a number of events. However, you cannot mix a metric that counts devices with a metric that measures the average boot time of devices, since a number of devices is not compatible with time units.
Make each line in the chart represent either the value of a metric or a computed ratio (when specified in the definition of the metric). When a metric defines thresholds, optionally display them in the line chart as horizontal lines. If you defined just one threshold, the chart displays a horizontal red line. When two thresholds are defined, the chart displays both a yellow and a red horizontal lines at the level specified by the thresholds. Points of the line above the thresholds are painted with the corresponding color. Even when you decide to show the thresholds, they may not be visible as horizontal lines in the chart if all the values of the metric are under the specified limits. Besides, displaying thresholds is only available when you display just one metric in the chart, since having thresholds in the same line chart for more than one metric would be too confusing.
Line charts do not span to the time frame selected in a dashboard, but further to the past. When you select a time frame for a dashboard (day, week, month, or quarter), each point in the line of a line chart represents the value of the metric aggregated for that period. The last point in the line, corresponds to the selected date in the dashboard, while the values to the left are past values. How far the line chart goes into the past depends on the selected time frame and on the amount of available historical data. The minimum span gradually grows into the maximum span as the Portal computes more and more data:
|Time frame||Line chart min span||Line chart max span|
|Day||30 days (1 month)||60 days (2 months)|
|Week||12 weeks (3 months)||52 weeks (1 year)|
|Month||12 months (1 year)||24 months (2 years)|
|Quarter||8 quarters (2 years)||16 quarters (4 years)|
The line chart widget is compatible with count and quantity metrics, but it is not adapted to display top metrics.
A bar chart graphically displays the values of count or quantity metrics in horizontal bars, letting you compare results visually with just a glimpse of the eye.
Depending on the metrics displayed on a bar chart, the bars that represent the measured values may be arranged differently:
- If the bar chart displays just one metric, arrange the bars either:
- By the grouping criterion selected in the definition of the count or quantity metric.
- By hierarchy. If you chose two grouping criteria (or none at all) when creating the metric, this is the only method available for arranging the bars.
- If the bar chart displays more than one metric, bars are arranged necessarily by metric.
Sort the results in the bar chart either by value or by name (the names of the metrics, the nodes in the hierarchy or the labels of the group by option). To see the most important values first, it is usually recommended to sort the bar chart by value in descending order as best practice.
If the bar chart holds a count metric, choose between displaying its value or one of the two following ratios:
- Ratio defined in the metric
- Because the ratio defined in the metric compares arbitrary groups (those determined by the conditions in both the metric and ratio definitions), the value of each bar is independently calculated and the sum of all the bars in the widget does not add up to 100% in general. This option is only valid if the count metric actually defines a ratio.
- Ratio of total
- In bar charts, you are seldom interested in displaying the ratio defined in the metric (if any), but rather the distribution of each grouping option with respect to the total number of objects seen in the widget. The bars in a widget that displays the ratio of total do add up to 100% and the related count metric does not need to define a ratio for this option to be valid.
If the bar chart displays a metric that defines thresholds, tick the box Threshold for the chart to paint the bars in a specific color depending on the value of each bar crossing the threshold or not: green, yellow (only for metrics with two thresholds), or red. If the metric does not define any threshold or you do not tick the Threshold box, bars are depicted in blue. If multiple metrics are displayed in the same bar chart, each one will be colored independently.
Choose the minimum number of bars that you wish to see in the chart. More bars may be visible if space in the dashboard allows it. Otherwise, when more bars than those visible are available, a small down arrow appears at the bottom right corner of the chart. Hover the mouse cursor over the bar chart and the small arrow turns into a slider. Use the slider to scroll through all the values in the bar chart. If you leave the bar chart after scrolling, small arrows appear again to indicate the availability of more results either to the top or to the bottom of the chart, or both, when applicable.
The width of the bar chart is automatically scaled to accommodate enough space for the highest value to display. For bar charts with only one metric, it is possible to fix the scale to a maximum value that is meaningful for you. in the configuration of the bar chart, choose between Automatic or Fixed. In the latter case, specify the maximum value that the chart should show. If this maximum fixed value is exceeded by a bar in the chart, the widget is automatically redimensioned to fit the new maximum.
A title widget lets you group several other widgets under the same title. The title widget does not hold metrics by itself, but acts as an umbrella for other widgets that do display metric results.
Title widgets are particularly useful for organizing widgets of different types into logical sets, offering you more flexibility in the design of layouts for your dashboards.
Not every type of widget is suitable for displaying any kind of metric. In the table below, find the compatibility matrix among the types of metrics and the widgets to display them.
|Metric Type||KPI||Table||Line chart||Bar chart|
|Count metric||Ok||Ok||Same units||Same units|
|Quantity metric||Ok||Ok||Same units||Same units|
|Top metric||-||Only one||-||-|
- Same units
- If you add more than one metric to the chart, the metrics must share the same units.
- Only one
- The widget accepts at most one metric.