August 2018 Summaries

Dynatrace OneAgent is a comprehensive tool for full-stack monitoring that provides alerting on over 100 problem types, including resource contentions and availability issues. The recent update to Dynatrace SaaS version 1.151 enhances its functionality by allowing third-party API clients to trigger all types of severity events through the Dynatrace Events API, which was previously limited to informational events. This update enables the integration of external alerts into Dynatrace's intelligent problem correlation system, allowing for more nuanced alerting and problem management. For instance, alerts from external systems such as power control monitoring can be pushed into Dynatrace to correlate with existing alerts, offering valuable root-cause insights. The new functionality also includes the ability to refresh and extend the duration of alerts, providing flexibility in incident management. This enhancement greatly improves the integration of external event sources and expands the options for using third-party tools and cloud deployment platforms with Dynatrace.

Atlassian provides a comprehensive suite of tools that enhance collaboration and continuous feedback in the software delivery cycle, which, when used efficiently, can significantly boost engineering productivity and optimize business metrics such as Time to Market and Time to Adoption. Addteq, an Atlassian partner, has demonstrated how to implement DevOps best practices using Atlassian and Dynatrace tools to ensure optimal performance and productivity. Key use cases include monitoring and addressing performance issues in Jira, Confluence, Bamboo, and Bitbucket, where Dynatrace's AI capabilities are employed to detect slowness, errors, and failures, providing root cause analysis and aiding in proactive error prevention. Addteq has developed a Jira Server plugin to enhance incident management, allowing seamless integration between Jira and Dynatrace, thereby closing the gap between development and operations teams. The collaboration and continuous improvement efforts by Addteq and Dynatrace aim to maintain and scale the efficiency of Atlassian tools as they are adopted more widely across enterprises.

Dynatrace has introduced LoadRunner request tagging to enhance performance engineering within DevOps practices, allowing users of LoadRunner and other load testing tools to tag simulated HTTP requests with context information through HTTP headers. This integration with Dynatrace enables advanced filtering and charting options by extracting test-transaction context as request attributes, which are not mandatory but beneficial for targeted load test diagnostics. The LoadRunner Request Tagging tool simplifies this process by inserting a method into LoadRunner scripts to populate the x-dynatrace-test header with key/value pairs like test step name, load script name, and more. These attributes help in analyzing load test data by filtering it in Dynatrace for performance issues such as crashes or scalability problems. The tool, based on the AppMon LoadRunner Converter and updated for JavaScript support, is open source and encourages community contributions. Users can start a 15-day free trial of Dynatrace to explore these features.

The text outlines a detailed process for transitioning a monolithic application, TicketMonster, to a microservices architecture, focusing specifically on the OrdersService microservice. It describes the use of feature flags to gradually incorporate the microservice into the existing monolith, allowing controlled traffic management through synthetic and live requests. The implementation involves using Apigee Edge for API proxy management, enabling dark launches and canary releases to ensure a smooth transition and minimize risk. The deployment strategy includes monitoring and adjustments informed by Dynatrace analytics to ensure the service operates effectively. The text also highlights the importance of addressing technical debt and decoupling databases for fully independent microservices in future stages of the migration journey, encouraging readers to apply these strategies to their own projects.

Dynatrace's engineering teams utilize Jenkins as their preferred build server to manage the complexities of a growing code base and diverse technology stacks while striving to optimize build times for their OneAgent across multiple platforms. The Engineering Productivity Team aims to reduce full build and test execution time from two hours to 45 minutes by leveraging automatically scaled Jenkins Nodes and Executors through Kubernetes clusters. They focus on effective capacity planning and configuration management to ensure efficient build processes, monitoring Jenkins with Dynatrace for comprehensive visibility into metrics like CPU usage, memory, and garbage collection. By addressing issues such as CPU spikes from merging test results and optimizing garbage collection settings, they enhance both system performance and user experience. Additionally, Dynatrace's Real User Monitoring (RUM) helps identify user interaction bottlenecks, allowing for further improvements in the Jenkins web interface. The team's strategic approach includes breaking down monolithic codebases, evaluating resource utilization, and potentially deploying multiple Jenkins servers to maintain efficiency and scalability.