6 updates shipped to Dash0 in October 2024: Extended support for grouping in built-in metrics, Powerful Filtering, Pie charts, tree maps and more in dashboarding and more.
Dash0 synthetic metrics are now available under the dash0.* namespace, and we have expanded the grouping capabilities!
We have improved the consistency of synthetic metrics. The metrics are now all grouped under the dash0 namespace. You can use dash0.spans in the query builder to get the span count, dash0.spans.duration for the span duration and dash0.logs for the log count. For greater compatibility, we have also made it so that dash0.spans.duration is interpreted as a native Prometheus histogram!
We have extended grouping capabilities for these metrics. Now can now group them by the dash0.*, otel.span.* and otel.log.* attributes, e.g., dash0.operation.name and otel.log.severity_range. Giving you more control than ever before!
Query builder showing a 90th percentile span duration, grouped by service name operation name.
Searching and filtering for the data you are about is central to your observability experience. Now, we have made the experience even better.
The filter control is now a permanently visible input field. This notably improves the experience, aids discoverability, allows you to see already applied filters, and makes for a much more natural keyboard interaction.
We have also made the content of the filter popup much more intelligent. You can now see values immediately on the first page, guiding you and allowing you to avoid having to know all attribute keys by heart. Also go straight into search made when your input doesn't full match an existing value!
Easy search and filter – with suggestions for keys, values and direct access to thecommonly used filter operation.
Dashboards are flexible and allow easy tailoring to your needs. They are now improved with tree maps, pie charts, and more!
We have been improving the dashboarding experience. Now, you can experience faster and more flexible dashboarding than before!
Starting with two new widget types, you can now visualize distributions using tree maps and pie charts. Both are handy for determining which pod is using the most memory or which service is generating the most errors.
Next, we have revised how we are showing unformatted series names. You know, those my_metric{service_name="shop", k8s_pod_name="shop-abcd"} selectors that are used as fallbacks when no better name is available. These are now color-coded for easier reading, and the attributes are sorted in the same as within the filter dialog.
Last, we highlight the chart cursor position across charts for much easier visual correlation!
Dashboard showing a tree map, pie chart and a time series chart.
Managing dashboard configurations and check rules manually can become a huge burden. Now you can take back control over both of these with the Dash0 API or with the Dash0 Kubernetes operator.
Dashboards are a very powerful tool to see all the telemetry for your specific use cases on one screen. Check rules are indispensable to get alerted when conditions in your system require human intervention.
Editing a dashboard or a check rule via Dash0's UI is probably the best way to get fast feedback, and iterate quickly. However, if you have a lot of dashboards or checks, at some point you might want to go with a more systematic approach. Configuration as code is the way to go.
The Dash0 Kubernetes operator now supports managing dashboards as well as check rules. Simply deploy them as Kubernetes resources, the Dash0 operator will automatically pick them up and synchronize them with your dashboards and check rules in Dash0. Additionally, you can also directly manage them via Dash0's API, if you want to do so from your CI system or via scripting.
Last but not least, Dash0 is committed to open standards. For dashboards, we use the emerging CNCF standard Perses, and for check rules we use the Prometheus rules custom resource definition. This enables you to build your dashboards and check rules with zero vendor lock-in.
Resources are a cornerstone of OpenTelemetry, and we have revised how we show information about them.
Resources are everywhere within Dash0: prominently within the resource map and table and explicitly called out within tracing, logging, metrics, and check rules. Starting today, resources are even more helpful.
We have thoroughly revised the sidebar and removed our modes concept from the resource map and table. The latter added unnecessary complexity and confusion. We also moved the table and map switcher to the main navigation, resulting in an overall much cleaner look without any information loss!
The hovercard now gained quick references to the reported signals in the footer, while we have reorganized the sidebar for clarity and visual hierarchy.
Within the sidebar, you can now always find the resource's health—with a dedicated tab for more details. Similarly, we have merged the requests, errors, and duration tabs into one, allowing us to switch between aggregations for these RED metrics!
The OpenTelemetry demo adservice within the resource table. Showing the sidebar RED-metric tab at the side.
Resources are also present whenever you look at spans, logs, and metrics. We also provide the same powerful hovercard-context experience within these locations. See a slow gRPC span? Quickly look up which service or Kubernetes pod is involved in the call!
The powerful resource hovercard is also available within all other areas where we reference resources.
The resource telemetry tab is now complete: Showing reported metrics and allowing navigation.
The resource sidebar's telemetry tab has been showing information about tracing and logging data for some time. Now, we are extending it with details about metrics reported by a resource!
Resource sidebar's telemetry tab showing a metrics card.
You can quickly navigate to metrics available for a service or pod through the new metrics card. This card makes identifying whether a resource emits metrics easier and facilitates jumping into the metric explorer.
