Menu
Grafana Cloud

Monitor Adaptive Traces

To ensure your Adaptive Traces configuration is working effectively and delivering real cost savings, it’s important to monitor its performance.

This topic covers the two primary ways to check the health and effectiveness of your sampling policies: the overview dashboard and metrics.

Use the overview dashboard for a quick, visual summary of your data volume, savings, and policy activity. For deeper analysis, you can query the metrics listed below.

Use the overview dashboard

To access the overview dashboard, navigate to Adaptive Traces > Overview tab.

This dashboard is the quickest way to get an at-a-glance view of your sampling strategy. It provides a high-level overview of your Adaptive Traces configuration and its effectiveness, showing how much data is being processed, stored, and sampled.

Adaptive Traces Ingest

This panel shows the total data volume of all traces received by Adaptive Traces per second before any of your sampling policies are applied. It represents the total amount of raw trace data your systems are generating.

Tempo Ingest

This panel displays the total data volume of the traces per second that were actually kept and sent to Tempo for long-term storage after your sampling policies made their decisions. This represents the high-value data you decided to retain.

By comparing Adaptive Traces Ingest with Tempo Ingest, you can see the direct impact of your policies. The difference is the volume of low-value data that was dropped, which translates into significant cost savings.

Number of Policies

This is a count of the total number of active sampling policies you currently have configured for Adaptive Traces.

Traces Sampled by Policies

This graph shows the count of individual traces (not their data volume) that are being kept by your policies over time. It helps you visualize the rate at which important traces are being sampled. Spikes on this graph often indicate noteworthy events, for example, latency issues, that your policies are correctly identifying and capturing for analysis.

Advanced monitoring with metrics

For more detailed analysis, you can directly query the following metrics.

Use these Grafana Cloud metrics in the grafanacloud-usage data source.

Ingest

These metrics help you monitor the rate of incoming trace data and the volume that is ultimately accepted for storage.

grafanacloud_traces_instance_adaptivetraces_bytes_received_per_second

Bytes received by Adaptive Traces (per second).

grafanacloud_traces_instance_bytes_received_per_second

Bytes received by Grafana Cloud Traces (per second).

grafanacloud_traces_instance_adaptivetraces_spans_received_total:rate5m

Spans received by Adaptive Traces (per second).

grafanacloud_traces_instance_spans_received_total:rate5m

Spans received by Grafana Cloud Traces (per second).

grafanacloud_traces_instance_adaptivetraces_discarded_spans_total:rate5m

Spans discarded by Adaptive Traces (per second).

Sampling decisions

These metric break down sampling outcomes by the specific policy that caused each trace to be either kept or dropped.

grafanacloud_traces_instance_adaptivetraces_global_sampled_traces_total:rate5m

Rate of traces sampled (per second).

grafanacloud_traces_instance_adaptivetraces_policy_sampled_traces_total:rate5m

Rate of traces sampled by policy (per second).

grafanacloud_traces_instance_adaptivetraces_policy_sampled_spans_total:rate5m

Rate of spans sampled by policy (per second).