Investigate a trace

A trace typically consists of multiple spans, where each span represents a distinct unit of work, such as a specific operation within a service. Spans can also have child spans, representing nested operations or tasks.

In the following image, shop-backend article-to-cart is a trace that completes in 1.19 seconds. The trace contains a number of spans, for example, authenticate, shop-backend place-articles, and shop-backend get article. Each of those spans contain child spans.

Viewing the details of a trace allows you to gain insights into the performance and behavior of your system. For example:

• Traces help pinpoint specific operations or services causing delays, such as slow database queries or external API calls.
• You can visualize how different services and components interact, making it easier to identify problematic dependencies.
• You can divide the total latency of a trace into individual spans so that you can better understand where time is spent.
• Traces often include error information, helping you quickly identify and resolve issues.
• By understanding the flow of requests, you can optimize resource usage and improve overall system performance.

How to view trace details

To view the details of a trace, click the trace name on the Slow traces tab.

Insights from viewing trace details

The image below provides the following insights:

• The trace shows how the shop-backend service interacts with other services like auth-service, cart-service, and article-service. This helps identify which service contributes the most to the overall latency.
• The get-article operation within the shop-backend service takes the longest time (1.02s), with significant delays in the select-articles (579.69ms) and postgres (503.49ms) spans. This indicates potential database query inefficiencies.
• The trace highlights nested spans, such as persist-cart (353.04ms) within cart-service, showing how operations depend on each other.
• The authenticate operation takes 71.77ms, with 38.59ms spent in the auth-service. This suggests a moderate overhead for authentication.
• The visualization helps identify the critical path of the request, which is essential for optimizing the most impactful spans.

In the next milestone, you’ll investigate the details of a trace span to identify the root cause of latency.