Widgets are automatically named based on the object, metric, or attribute selected for the first widget query. If you also choose a grouping for the widget, the "Group by" attributes are included in the automatic naming (aggregation method and metric).

Selecting a new object or grouping attribute updates the automatic name.

If you do not want to use the automatic widget name, you can simply type a new name.

All dashboard widgets can be positioned and sized according to your needs for each particular dashboard.

In dashboard Edit mode, click the menu on the widget and select Duplicate. The widget editor is launched, pre-filled with the original widget's configuration and with a “copy” suffix in the widget name. You can easily make any necessary changes and Save the new widget. The widget will be placed at the bottom of your dashboard, and you can position it as needed. Remember to Save your dashboard when all changes are complete.

Most widgets on dashboards can be displayed with or without legends. Legends in widgets can be turned on or off on a dashboard by either of the following methods:

As the data displayed in the widget changes, the legend for that widget is updated dynamically.

When legends are displayed, if the landing page of the asset indicated by the legend can be navigated to, the legend will display as a link to that asset page. If the legend displays "all", clicking the link will display a query page corresponding to the first query in the widget.

Data Infrastructure Insights provides different transform options for certain metrics in widgets (specifically, those metrics called "Custom" or Integration Metrics, such as from Kubernetes, ONTAP Advanced Data, Telegraf plugins, etc.), allowing you to display the data in a number of ways. When adding transformable metrics to a widget, you are presented with a drop-down giving the following transform choices:

Note that transforming metrics does not change the underlying data itself, but only the way that data is displayed.

In a chart widget, you can show either the Top or Bottom results for rolled up data, and choose the number of results shown from the drop-down list provided. In a table widget, you can sort by any column.

Data in a table widget can be grouped by any available attribute, allowing you to see an overview of your data, and to drill-down into it for more detail. Metrics in the table are rolled up for easy viewing in each collapsed row.

Table widgets allow you to group your data based on the attributes you set. For example, you might want your table to show total storage IOPS grouped by the data centers in which those storages live. Or you might want to display a table of virtual machines grouped according to the hypervisor that hosts them. From the list, you can expand each group to view the assets in that group.

Grouping is only available in the Table widget type.

You can select the time range for your dashboard data. Only data relevant to the selected time range will be displayed in widgets on the dashboard. You can select from the following time ranges:

You can override the main dashboard time range setting in individual widgets. These widgets will display data based on their set time frame, not the dashboard time frame.

To override the dashboard time and force a widget to use its own time frame, in the widget's edit mode choose the deisired time range, and Save the widget to the dashboard.

The widget will display its data according to the time frame set for it, regardless of the time frame you select on the dashboard itself.

The time frame you set for one widget will not affect any other widgets on the dashboard.

Different metrics use different units of measurements for the data they report in a chart. For example, when looking at IOPS, the unit of measurement is the number of I/O operations per second of time (IO/s), while Latency is purely a measure of time (milliseconds, microseconds, seconds, etc.). When charting both metrics on a single line chart using a single set a values for the Y-Axis, the latency numbers (typically a handful of milliseconds) are charted on the same scale with the IOPS (typically numbering in the thousands), and the latency line gets lost at that scale.

But it is possible to chart both sets of data on a single meaningful graph, by setting one unit of measurement on the primary (left-side) Y-axis, and the other unit of measurement on the secondary (right-side) Y-axis. Each metric is charted at its own scale.

This example illustrates the concept of Primary and Secondary axes in a chart widget.

In a dashboard, any time series widget (line, spline, area, stacked area) bar chart, column chart, pie chart, or table widget allows you to build expressions from metrics you choose, and show the result of those expressions in a single graph (or column in the case of the table widget). The following examples use expressions to solve specific problems. In the first example, we want to show Read IOPS as a percentage of Total IOPS for all storage assets in our environment. The second example gives visibility into the "system" or "overhead" IOPS that occur in your environment—​those IOPS that are not directly from reading or writing data.

You can use variables in expressions (for example, $Var1 * 100)

Variables allow you to change the data displayed in some or all widgets on a dashboard at once. By setting one or more widgets to use a common variable, changes made in one place cause the data displayed in each widget to update automatically.

Dashboard variables come in several types, can be used across different fields, and must follow rules for naming. These concepts are explained here.

The Solid and Bullet Gauge widgets allow you to set thresholds for Warning and/or Critical levels, providing clear representation of the data you specify.

To set formatting for these widgets, follow these steps:

You can optionally choose a minimum and maximum value for the gauge. Values below minimum will not display the gauge. Values above maximum will display a full gauge. If you do not choose minimum or maximum values, the widget selects optimal min and max based on the widget's value.

in the Single-Value widget, in addition to setting Warning (orange) and Critical (red) thresholds, you can choose to have "In Range" values (those below Warning level) shown with either green or white background.

Clicking the link in either a single-value widget or a gauge widget will display a query page corresponding to the first query in the widget.

Like single-value and gauge widgets, you can set conditional formatting in table widgets, allowing you to highlight data with colors and/or special icons.

Conditional Formatting allows you to set and highlight Warning-level and Critical-level thresholds in table widgets, bringing instant visibility to outliers and exceptional data points.

Conditional formatting is set separately for each column in a table. For example, you can choose one set of thresholds for a capacity column, and another set for a throughput column.

If you change the Unit Display for a column, the conditional formatting remains and reflects the change in values. The images below show the same conditional formatting even though the display unit is different.

You can choose whether to display condition formatting as color, icons, or both.

Most widgets on a dashboard allow you to specify the Units in which to display values, for example Megabytes, Thousands, Percentage, Milliseconds (ms), etc. In many cases, Data Infrastructure Insights knows the best format for the data being acquired. In cases where the best format is not known, you can set the format you want.

In the line chart example below, the data selected for the widget is known to be in bytes (the base IEC Data unit: see the table below), so the Base Unit is automatically selected as 'byte (B)'. However, the data values are large enough to be presented as gibibytes (GiB), so Data Infrastructure Insights by default auto-formats the values as GiB. The Y-axis on the graph shows 'GiB' as the display unit, and all values are displayed in terms of that unit.

If you want to display the graph in a different unit, you can choose another format in which to display the values. Since the base unit in this example is byte, you can choose from among the supported "byte-based" formats: bit (b), byte (B), kibibyte (KiB), mebibyte (MiB), gibibyte (GiB). The Y-Axis label and values change according to the format you choose.

In cases where the base unit is not known, you can assign a unit from among the available units, or type in your own. Once you assign a base unit, you can then select to display the data in one of the appropriate supported formats.

To clear out your settings and start again, click on Reset Defaults.

Data in widgets on Dashboards and Asset Landing Pages auto-refresh according a refresh interval determined by the Dashboard Time Range selected. The refresh interval is based on whether the widget is time-series (line, spline, area, stacked area chart) or non-time-series (all other charts).

Each widget displays its auto-refresh interval in the upper-right corner of the widget.

Auto-refresh is not available for Custom dashboard time range.

When combined with TV Mode, auto-refresh allows for near-real-time display of data on a dashboard or asset page. TV Mode provides an uncluttered display; the navigation menu is hidden, providing more screen real estate for your data display, as is the Edit button. TV Mode ignores typical Data Infrastructure Insights timeouts, leaving the display live until logged out manually or automatically by authorization security protocols.

You can temporarily suspend auto-refresh by clicking the Pause button in the upper right corner. While paused, the dashboard time range field will display the paused data's active time range. Your data is still being acquired and updated while auto-refresh is paused. Click the Resume button to continue auto-refreshing of data.

Grouping allows you to view and manage related dashboards. For example, you can have a dashboard group dedicated to the storage in your environment. Dashboard groups are managed on the Dashboards > Show All Dashboards page.

Two groups are shown by default:

The number of dashboards contained in each group is shown next to the group name.

To create a new group, click the "+" Create New Dashboard Group button. Enter a name for the group and click Create Group. An empty group is created with that name.

To add dashboards to the group, click the All Dashboards group to show all dashboards in your environment, of click My Dashboards if you only want to see the dashboards you own, and do one of the following:

Remove dashboards from the current group in the same manner by selecting Remove From Group. You can not remove dashboards from the All Dashboards or My Dashboards group.

You can further manage your dashboards by pinning favorite ones to the top of your dashboard list. To pin a dashboard, simply click the thumbtack button displayed when you hover over a dashboard in any list.

Dashboard pin/unpin is an individual user preference and independent of the group (or groups) to which the dashboard belongs.

You can choose to display Data Infrastructure Insights using either a light theme (the default), which displays most screens using a light background with dark text, or a dark theme which displays most screens using a dark background with light text.

To switch between light and dark themes, click the username button in the upper right corner of the screen and choose the desired theme.

Dark Theme Dashboard view:

Light Theme Dashboard view:

Different data collectors often poll their data at different intervals. For example, data collector A may poll every 15 minutes while data collector B polls every five minutes. When a line chart widget (also spline, area, and stacked area charts) is aggregating this data from multiple data collectors into a single line (for example, when the widget is grouping by "all"), and refreshing the line every five minutes, data from collector B may be shown accurately while data from collector A may have gaps, thus affecting the aggregate until collector A polls again.

To alleviate this, Data Infrastructure Insights interpolates data when aggregating, using the surrounding data points to take a "best guess" at data until data collectors poll again. You can always view each data collector's object data individually by adjusting the widget's grouping.