Control your telemetry data

Your app generates a lot of data, but not all of it's useful. Storing every byte is costly, and can quickly overwhelm your capacity limit, license limit, or both. Aside from financial cost, there's computing cost associated with generating, processing, and temporarily storing your telemetry data.

Reducing telemetry data volume keeps costs predictable and manageable as your business scales, and helps teams manage their data effectively by reducing the amount of available information. When you optimize your telemetry data, on-call engineers can find the data they need to solve problems faster.

To take control of your telemetry data, Chronosphere Observability Platform provides multiple tools you can use to identify and eliminate data that isn't used and doesn't deliver value to your business. See Chronosphere control concepts to learn about the mechanisms you can use to control your telemetry data.

Understand consumption and limits

Manage license consumption by reviewing licensing dashboards to identify usage trends and avoid exceeding your organization's licensing limits.

Understand telemetry data limits for each telemetry type to ensure a performant experience when ingesting, querying, and creating resources.

Shape metric data

Use the Chronosphere Control Plane to shape your metric data and reduce the amount of telemetry data you persist. Manage your telemetry data with tools that help you shape that traffic, such as creating drop rules, quotas, and aggregation rules.

Sample trace data

Create trace datasets that map tracing license consumption to relevant business units in your organization. Datasets let you track and measure trace data usage over time before you implement sampling strategies, whereas head and tail sampling alone require you to measure and implement together.

By parsing out the ability to measure and track trace data volume, you can create datasets, learn about how trace data is processed and persisted, and then apply behaviors to set sampling rules for your datasets without having to write, coordinate, and manage head and tail sampling rules individually.