What is trace-based visualization? In microservices, distributed tracing is a method for aggregating all the operations that occur in your distributed systems that were triggered by a specific request. If these traces are visualized, developers can gain insights into how their service behaves when it’s run with other services, which helps them understand why errors occur.
In a new video, Aviv Kerbel, our developer advocate, shows how to visualize traces and then use the results to investigate issues, reproduce scenarios and generate tests for your cloud-native applications with Helios, which is a free developer tool. This blog post covers the main steps. You can also watch the entire video here.
How to Visualize Traces in Helios
1. First, sign up for Helios. It has a free tier and you can use it to visualize your traces and investigate issues.
When you open Helios, you will see all the services Helios is installed on and the entry points for each specific service.
2. To see all the traces for a specific entry point, select the preferred entry point.
3. You can expand the trace to get more details about the trace. You will be able to see errors, attributes, payloads, logs, how the data flows in the system, and more.
In the example, you can see the request attributes. There was a call to Postgres and a Kafka message broker. We can also see there is an error.
4. To understand why the error occurred, you can drill down into the error logs. As you can see, there was a Stripe request that failed. Now, you can go back to your code and resolve the issue!
Once this issue is resolved, we want to visualize the trace once again, just to make sure.
5. As you can see, the same Stripe request that had previously failed, now succeeded.
Instead of combing through logs, Helios helped us quickly identify the error so we can fix it.
Additional capabilities trace visualization capabilities that are supported by Helios include:
- Seeing the duration of each span:
Seeing the commit hash of each service:
Ready to get started? Start today!
