Announcing Unified Dashboards - Simplifying and Consolidating monitoring across all your telemetry
by Jake Knostman
Your team is eager to push the changes you have been working on all week so your customers can take advantage of the great new features you’ve been developing. All of the testing pipelines are green and you’re feeling confident that a new deployment will go off without a hitch.
Based on past experiences, however, you have a healthy sense of skepticism around any new push to production and want to make sure you’re prepared. You want to proactively monitor the health of this particular service to ensure your deployment doesn’t cause any trouble. Keeping an eye on your key health indicators, spread across telemetry types (metrics, traces/spans, and possibly even logs) and even potentially different observability tools, are going to tell you whether this release is successful, or you’re ordering takeout for the team while you troubleshoot. Since each telemetry type provides a different lens into understanding the health of your system, you have to keep tabs on all of them.
To cover your bases, you kick off the “multiple tabs of observability(™).” The first tab contains your dashboard of metrics charts that tell you the current machine/resource health. The second tab is used to visualize your key Service Level Indicators (SLIs - latency, error rate, and throughput) which are based on aggregated spans/traces to better understand the health of the customer experience related to your service. You may even have a third tab open to view logs just in case. Once you have your tabs up and you have a sense for what the current baseline levels are for your service, you are ready to click deploy. For the next 10-15 minutes after the deployment is live you continuously cycle through those tabs to keep an eye on the health of your service and the customer experience with your service, and ensure none of your charts suddenly change. Do you really need three tabs for simply viewing the health of your software across telemetry types? Does it need to be so difficult?
The answer is no. Lightstep Observability’s new Unified Dashboards solve this exact problem. Unified Dashboards provide you with a simple and consistent way to interact with and visualize your data across both metrics and spans, giving you a single location to monitor the health of your service. Instead of clicking back and forth between tabs to see the current health of your software, you can have a single dashboard open with all of your key indicators charted across telemetry types. You can now proactively monitor your applications and visualize the relationship between your telemetry types, from a single tab.
We are very excited to bring you a truly unified observability experience to proactively monitor the applications you care about. Simply visualizing span and metric data side-by-side in a single location is an improvement from the multiple tabs of observability experience many users have today. Beyond that, Unified Dashboards provide added simplicity and consistency to how you interact with and visualize your data, regardless of the telemetry type you want to use:
- The Unified Query Builder provides a fast and highly expressive way to search across metrics and span data to make it easy to find the right indicators to monitor.
- The UQB makes it easier to interact with your data across metrics and spans to ensure you are monitoring the entirety of your service.
- You now have a consistent UI across telemetry types with easy access to direct traces on span based queries.
- Users shouldn’t have to learn different workflows to interact with the different types of telemetry.
Proactively monitoring your applications with a Unified Dashboard allows you to see the health of your software at a given moment and can help you catch anomalies or changes in your key indicators - but it is only a part of the story. If you detect a change in the health of your software, it is important to be able to investigate and try to find the root cause. Through Unified Dashboards you can pivot back and forth between metric and span charts and even get directly to a trace from span based charts. On top of that, the dashboards serve as a jumping off point to investigate issues further within a notebook or even directly with Change Intelligence from a metrics chart.
Workflows should not be siloed - your monitoring workflow should lead you into your investigative workflow. This is especially important when something goes wrong (and something will inevitably go wrong).
Unified Dashboards can not only be useful in any monitoring scenario where you want to visualize and interact with metric and span data side by side, but essential to get a holistic view of the health of the software you care about - whether it is monitoring the health of a deployment, or keeping an eye on the performance of critical applications during a busy period, or simply monitoring the current health of your system. With Lightstep’s Unified Dashboards, we’re simplifying the monitoring workflow within Lightstep, and setting the foundation for future improvements to be made for both spans and metrics together – as it should be.