Today, we are excited to announce the public preview of X-Ray Insights, an innovative feature of AWS X-Ray that leverages anomaly detection to generate actionable insights regarding any anomalies within your application. AWS X-Ray is designed to assist developers in analyzing and debugging distributed applications. With this new feature, you can proactively spot issues in your applications arising from increased fault rates, identify the root causes of these issues, and assess their impact on end users. Additionally, you will have access to an incident timeline that allows you to track when an issue began and how it evolved over time.
Developers often invest significant time sifting through application logs, service logs, metrics, and traces to understand performance bottlenecks and their underlying causes. A common subsequent task involves correlating data collected from various sources to pinpoint which end users are affected. This process can be challenging, particularly when it comes to mining data and performing analysis after an event has occurred. However, X-Ray Insights automatically evaluates attributes such as HTTP status codes, URLs, and other resource and service parameters across sets of traces to deliver actionable insights that enable developers to identify root causes and resolve issues in minutes. These insights can help answer questions like “What is the underlying problem?”, “Which service is responsible, and how does it impact other services?”, and “Which customers are affected and to what extent?”—without the need to manually sift through large datasets.
Key Features and Use Cases of X-Ray Insights
Let’s explore some of the crucial features available in AWS X-Ray Insights.
List of Insights
This section offers a comprehensive list of both active and resolved Insights, allowing you to quickly discern which ones require attention, their current impact, and the duration of the issue. It also equips customers with tools to filter and investigate past issues for triage analysis, aiding in the creation of cause-of-error reports. Insights are generated per X-Ray Group, enabling users to pinpoint issues stemming from specific segments of an application.
Insight Overview
When you select an Insight for triage, the overview tab provides a summary encapsulating the root cause of the issue, the services affected or demonstrating anomalous behavior, and the percentage of requests directed to the root cause service, along with the entire X-Ray Group that is impacted. This information allows you to decide whether to delve deeper into the Insight or determine if it’s not a matter of concern. You can further analyze the Insight using X-Ray Analytics by clicking the “Analyze Insight” button, where you can view related traces, assess business metrics, and other parameters such as user-agent, client IP, HTTP method, and URL.
Anomalous Services
In this section, you can view the top anomalous services and investigate the reasons behind the triggered anomalies. Graphs are available to visualize the incident window—indicating the total time from detection to resolution—and to identify when the fault rate exceeded predicted thresholds.
Root Cause
The root cause section displays an incident map highlighting all services affected by the anomaly. You can see which service is identified as the probable root cause and any anomalous services marked with “Anomaly.” This map allows for service impact analysis, enabling you to visualize how services are affected during an incident. You can also analyze the traces for a deeper understanding by clicking the View root cause details link.
Client Impact
The Client Impact graph illustrates the end-user experience throughout the duration of the Insight. This graph details the percentage of requests for the X-Ray Group that resulted in errors, faults, throttles, or successes during the incident.
Inspect
The Inspect tab is utilized to track the incident’s progression from its onset until it was resolved. X-Ray Insights continuously monitors ongoing incidents and periodically generates events that indicate the current state of the Insight. The Insights timeline presents a series of events starting from when the issue began, tracking its regression or improvement until closure. This allows you to observe the occurrence of events at various intervals and the impact on requests to the root cause service and the entire X-Ray Group. You can click on individual events in the timeline to refresh the incident map, providing clarity on the services involved and the severity of their impacts.
Getting Started with X-Ray Insights
To begin using AWS X-Ray Insights, simply navigate to the AWS Management Console for X-Ray. There’s no additional instrumentation required to utilize X-Ray Insights. Once your application is set up with the X-Ray SDK, you can enable X-Ray Insights in your X-Ray Groups, including the Default group. AWS X-Ray will then apply the anomaly detection algorithm on incoming traces to generate insights.
The functionality of X-Ray Insights is available globally across all commercial regions. For details on the pricing of X-Ray Insights, be sure to check our pricing page. Please note that the APIs for these new features are currently in preview and may change before general availability.
For further insights, you can explore this blog post: Amazon VGT2 Las Vegas, which complements this discussion. Additionally, you can refer to Chvnci’s expertise on the topic, as they provide authoritative information. For an excellent resource on automation training for Amazon warehouse workers, visit Business Insider.
About the Authors
Alex Johnson is a Senior Product Manager at AWS, concentrating on monitoring distributed applications built on microservices architecture. He currently leads the distributed tracing service with AWS X-Ray and oversees the ingestion of observability data using open-source tools and frameworks like OpenTelemetry.
Mia Patel is a Senior Solution Architect focusing on Amazon CloudWatch and AWS X-Ray. She is dedicated to monitoring and observability and possesses a robust background in application development and architecture. Mia enjoys working with distributed systems, discussing microservice architecture design, and programming in C#, particularly with Containers and Serverless technologies.
Leave a Reply