OpsMx just released a new Delivery Intelligence for Argo module that enables you to accelerate your application delivery by automating and shifting left deployment analysis and release verification. It reduces the operational complexity of managing multiple Argo instances and provides a fast and secure way to scale Argo and Kubernetes in your enterprise.
Deploying and releasing software is a complex process constrained by manual deployment and release verification interventions. IT organizations have been pretty good at automating the deployment of artifacts but verifying that those artifacts have been securely and safely released and deployed is still a tedious and largely manual process. This is a significant constraint impacting the business.
The Argo project is an open-source CD platform that provides robust capabilities to implement contemporary software delivery practices such as GitOps and Progressive Delivery. However, identifying and understanding the risk of a change is still a manual and frustrating exercise that can lead to unnecessary delays and impact service availability.
Understanding the impact of a change
What is the impact of a software change when released into production, and did you deliver what you said you would deliver?
Today, DevOps and SRE teams spend many hours/days trying to estimate the risk of a change and verifying that a change has been successfully released into production. Deployment strategies help mitigate this risk, but it usually involves a manual decision to move forward regarding security and governance policies.
A software change can impact directly or indirectly many components or services. Determining the impact of the change involves aggregating and analyzing data from many different processes and tools. The earlier you determine the impact, the lower the risk and the faster you can fix and deliver the change. You need to be able to “Shift Left” the risk of deployment and release verification before releasing it to your customers.
Here are the three things you need to implement to “Shift Left” your deployment risk.
The Need for Continuous Verification
You are too late if you only verify your software changes once they have been released into production! You must have a process that continuously identifies and mitigates risks at all stages of the software development process. It should provide a continuous risk assessment of software builds, automated test cases, and deployment releases from a single pane of glass and inform and notify stakeholders of the confidence level for promoting a release to the next stage.
Delivery Intelligence for Argo is an intelligent continuous verification platform that enhances and maximizes analysis, performance, and quality measurement at every CI/CD pipeline stage.
Leveraging the Power of AI/ML Deployment Analytics
Powered by advanced AI/ML algorithms, the Delivery Intelligence module enables large-scale acquisition and analysis of logs and metrics from monitoring tools to understand production health post-deployment and allow further diagnosis and triage.
OpsMx’s automated verification feature is intelligent-driven verification, whether you are deploying to test or deploying a canary or progressive release. The Delivery Intelligence module looks at the statistical data to compare and contrast with past data; it uses machine learning analysis to identify relevant and significant metrics for risk evaluation automatically. For logs, the Delivery Intelligence module leverages natural language processing to automatically classify and report on errors or warnings, saving considerable time.
In a nutshell, the OpsMx Delivery Intelligence for Argo module observes, learns, and applies these statistical analyses and machine learning-based models to identify risk identification providing a confidence score to automate verifications and approvals.
Automate Decisions and Actions
Transform your Argo CD and Argo Rollouts environments into a fully autonomous pipeline by verifying new application deployments based on errors, exceptions, and performance and automating rollbacks or deployments based on machine learning algorithms.