Datadog announced Deployment Tracking, a new feature for Datadog APM. This feature enables engineering teams to identify when new code deployments are the root cause of performance issues.
With the rise in adoption of continuous integration and continuous delivery (CI/CD) practices, DevOps teams are increasingly using modern code deployment strategies such as Canary, Blue-Green, and Shadow deployments to test new versions with limited impact to end-users.
While this agility comes with an increased risk of failure, traditional APM vendors offer hard-to-setup solutions that do not monitor these deployments in real time or the impact they have on specific endpoints.
To prevent such risky visibility gaps, Datadog Deployment Tracking visualizes key performance metrics such as requests per second and error rate, identifying new error types for specific endpoints during every code deployment. This allows developers to detect and contain the impact of changes as they happen, as well as respond to incidents more quickly.
“Our customers build and ship applications with multiple types of deployment practices, increasing efficiency but often with the risk of impacting overall performance or introducing errors,” said Renaud Boutet, Vice President of Product, Datadog.
“Deployment Tracking will visualize and compare key data related to various version deployments, helping our customers efficiently prevent outages related to bad code deploys, so they can rapidly iterate their applications in a more organized way.”
“Our customers depend on Cvent’s flexible and scalable platform for a seamless, impactful virtual event experience, which is essential in our current environment,” said Brent Montague, Site Reliability Architect at Cvent.
“To ensure a world-class experience for all users, we lean on Deployment Tracking within Datadog APM, which easily pinpoints new errors down to specific endpoints, improving our MTTR and allowing for continuous shipping of features and fixes to our customers. By leveraging Datadog’s Deployment Tracking, our engineering teams can adopt Canary deployments with confidence.”
Datadog Deployment Tracking is available for all languages supported by Datadog APM and works in both containerized and non-containerized environments. Deployment Tracking extends existing APM capabilities by using a unified version tag to analyze recent deployments. Functionalities include:
- Easily comparing performance between versions: quickly identifying bad deployments by comparing high-level performance and error data between releases.
- Ensuring efficiency of targeted fixes: viewing granular performance data down to a single endpoint to ensure a hotfix is actually resolving the issue.
- Starting troubleshooting in one-click: leveraging seamless correlation between version performance metrics and the associated hosts, traces, logs, code profiles, and processes to detect the root-cause faster.