Upgrade from Dashboards Classic to Dashboards

After the upgrade, the existing layout is preserved, ensuring a smooth upgrade without immediate visual changes, and also accommodating wide dashboard layouts, which otherwise could not be upgraded without breaking the layout.

When you open the upgraded document, a message box offers you the following choices about your dashboard layout:

Switch to the new responsive layout

When you switch to the responsive layout, you can benefit from a reflowing layout, supporting all screen sizes, even mobile ones, enhancing usability and accessibility.

Switch to responsive layout

Switching back

When you switch to the responsive layout, you're notified that "Your dashboard has been successfully switched to Responsive Layout."

If the conversion to the new layout causes any issues, you can select Revert changes in the message box to display the previous version of the dashboard. From there, select Restore to revert to the previous version (with the canvas layout).

Apply responsive layout in a later stage

If you choose to keep the canvas-based layout, you can still switch to the responsive layout later from your dashboard settings.

Switch to the new responsive layout via settings

The upgrade supports a limited set of tile types from Dashboards Classic for recreation in the new Dashboards app. Supported tile types are:

Because Grail provides better scale and cardinality for metrics, we didn't replicate metrics one-to-one. Instead, we created new metrics that are either extended or a merged version of multiple previous metrics.

See Built-in Metrics on Grail for a full list of metrics available on Grail, with a mapping from their Dashboards Classic counterparts, so you can select the best metrics whenever the upgrade can't automatically map them.

The following settings are upgraded. These mainly apply to Data Explorer tiles.

The upgrade process does not support all tiles available in Dashboards Classic. Built-in tiles (for example, USQL, Service, or Database) not described in Supported tile types. If a tile does not convert, in new dashboard (in Dashboards), Dynatrace replaces the built-in tile with a Markdown tile explaining that the built-in tile type can't be converted automatically and that you need to recreate it manually in Dashboards.

Built-in tiles that are not upgraded:

The following visualization settings are not supported for dashboard upgrade. These mainly apply to Data Explorer tiles.

The following dashboard settings are not supported for dashboard upgrade. These mainly apply to Data Explorer tiles.

If you upgrade Kubernetes-based dashboards, the new Kubernetes metrics are available only for DPS-enabled customers, so you might not see data, even though the upgrade was successful.

Not all data types are available on Grail:

Be aware that some data types, as well as some metrics, are only available when using DPS. For example, if you upgrade Kubernetes-based dashboards, the new Kubernetes metrics are available only for DPS-enabled customers, so you might not see data, even though the upgrade itself was successful.

The following features are not supported for dashboard upgrade:

Be aware that not every data point is available on the latest Dynatrace. Before starting to upgrade, ensure that all relevant data types that you need are available for and stored in Grail.

Grail currently supports the following data types:

If you rely on RUM data (see above), and are part of the current preview, ensure that you have enabled the new RUM agent. You can enable it in the settings of a frontend application: Enablement and cost control > Enable new Real User Monitoring Experience.

You can only use the full scope of the latest Dynatrace if your license agreement includes Dynatrace Platform Subscription (DPS).

Dashboards Classic allows tiles with a limited set of predetermined data sources, such as metric-based tiles on Data Explorer, health states for applications and services, digital experience signals like Apdex on a world map, and other built-in tiles.

With the new Dashboards app, you can visualize all data stored in Grail and external data.

With latest Dynatrace we have implemented new concepts for filtering data—segments and variables—to help reduce visual clutter and make it easier find relevant data. Previously, you had to define many static management zones to achieve the same results.

Segments

Segments logically structure and filter observability data across apps on the Dynatrace platform. Use cases include filtering your data based on logical structure.

Unlike management zones, segments aren’t selected globally. However, when you set segments for one app, they are carried over to other apps when via Open with.

For instance, if you filter a notebook’s data by segments and then use Open with > Dashboards to create a new dashboard, the selected segments are automatically applied to the corresponding tile.

Variables

Variables allow you to filter your data across dashboard tiles but in a more versatile way. Variables let you fully control:

Variables go far beyond just being filters on top of your dashboard. They are increasingly used to optimize performance and costs for repeating parts of queries, as well as to increase efficiency and maintainability.

For instance, you can also add variables to your markdown, tile title, or tile descriptions to make your dashboards and tiles more dynamic. In the future, variables will also be increasingly used in visualization settings.

There are two types of dashboard layouts available in Dynatrace:

The canvas-based layout led to issues in day-to-day work, such as screen size mismatches causing partial visibility or unused whitespace on different screens, and required scrolling on smaller devices, making tiles difficult to interact with.

The new responsive layout offers much more flexibility, and ensures dashboards are fully visible and easier to navigate on smaller devices through dynamic reflow, while also making tile swapping and repositioning easier, enhancing user experience and flexibility.

With the latest Dynatrace, we have introduced new ways to define alerts and get notified.

In Dashboards Classic, Dynamic filters allow you to filter your data across dashboard tiles based on entities, tags, or custom metric dimensions. You can add multiple dynamic filters to a dashboard—for example, to filter infrastructure metrics by operating system—and only apply them once a filter value is entered. However, how these filters are applied to tiles is fully pre-determined by Dynatrace.

Basic filters (for example, when you filter a service metric by a service name or tag) can be recreated as a variable.

Relationship filters, however, should not be replicated immediately, because, in a subsequent step, you have to edit every tile referencing the newly created variable. In our experience, these referencing efforts are greater than rethinking your filtering using variables. For example, for Kubernetes metrics, unlike in Dashboards Classic, where using relationships were often times required to filter by related fields—like pods, namespaces, and clusters—are now automatically enriched on most metrics. This means you can rely on these enriched fields directly, instead of needing to define and use relationships in your filter statements manually. However, as mentioned earlier, this requires adding the filter per tile manually.

In Dashboards Classic, you could use management zones to filter individual tiles or dashboards. Management zones no longer exist in the latest Dynatrace, however, so there is no automatic upgrade available.

Instead of management zones, use segments in the latest Dynatrace for data segmentation. Once your segments are defined, you can set them as individual filters on a tile or use them as dashboard-wide filters, similar to what you know from Dashboards Classic.

With Dashboards, we have introduced a lot of new capabilities, providing deeper insights and adding new concepts and use cases.

Review the available visualizations to see if there are better ways to visualize your data.

Use the power of Dynatrace Intelligence by giving your dashboards analyzers for forecast analysis and anomaly detection (thresholds or seasonal baselining).

Use code tiles to

Use Open with from your dashboard to drill down into other applications to gain further insights or start workflows.

In cases where a tile couldn't be upgraded, you can recreate it manually if all of the following conditions are met:

Let's look at an example where the upgrade doesn't automatically upgrade your tile, but where the above conditions are met and you can manually recreate them.

Example: Problem tile

To recreate the tile

The upgrade supports a limited set of tile types from Dashboards Classic for recreation in the new Dashboards.

Supported tile types are:

Please reach out to your Dynatrace contact if you need further information about upgrading dashboards.