Virtualization Technology News and Information
Codefresh 2019 Predictions: Infrastructure as Code, Deployment Metrics, and GitOPS

Industry executives and experts share their predictions for 2019.  Read them in this 11th annual series exclusive.

Contributed by Kostis Kapelonis, Developer Advocate at Codefresh

Infrastructure as Code, Deployment Metrics, and GitOPS

Terraform finally faces a competitor with Pulumi

Many major organizations with multi-cloud strategies have adopted Infrastructure as Code (IaC) to provide more flexibility and scalability. IaC enables companies to manage their stacks by describing platforms in formal specification files.  So far, Terraform has been uncontested in this sector, and enjoys universal adoption without any other competitors. Pulumi is a new company that is challenging this space with its innovative solution that, unlike Terraform, uses actual programming languages such as Python, Go, and JavaScript.

With Pulumi, developers and operators can now use the same languages for applications and their infrastructures. This groundbreaking capability is going to be favored by organizations who feel restricted by the custom syntax of Terraform files.

Advanced deployment methods (blue/green/canaries) become the norm

Gone are the days where downtime because of deployments are acceptable. Today people are researching advanced deployment strategies assuming that this is something extra or a "nice to have" feature.

In practice, deploying without breaking anything (e.g. with canary rollouts) will become the "only" proper way of deploying. Companies that don't adopt gradual rollouts will soon be left behind.

Automated metrics will also become an integral part of deployments. Deployment systems will use these metrics as a basis for automatic rollbacks. No longer will engineers be awoken in the middle of the night to manually roll back a build. You can see Canary in action in this webinar

Service meshes in the form of Istio will explode in popularity

Because of advanced deployments, service meshes will also become an integral part of any cluster. Today service meshes are seen as something that only large organizations require, and most people are unaware of their existence. One of the most powerful characteristics of service meshes is dynamic traffic routing within a cluster which is a prerequisite for performing rolling canary updates or any other gradual deployment pattern.

Service meshes will soon become essential in companies of all sizes as more and more services are moved into clusters. Security constraints and traffic limitations will be some of the first requirements that will make service meshes popular.

GitOps will become the new buzzword

GitOps will become the new buzzword and nobody will agree on its specific meaning. If you deploy using git then it is GitOps, but thousands of articles will be written that will attempt to explain what GitOps is or isn't (you can start with a great webinar on GitOps with Serverless). Each company will have its own definition, which is bound to confuse decision makers even more.


About the Author


Kostis Kapelonis is a Developer Advocate at Codefresh, a continuous delivery platform built for Kubernetes and containers. Formerly a Software Engineer, Kostis has years of experience containerizing applications, building CI/CD pipelines, and developing Java applications. He lives in Greece and loves roller skating.

Published Thursday, January 31, 2019 10:01 AM by David Marshall
There are no comments for this post.
To post a comment, you must be a registered user. Registration is free and easy! Sign up now!
<January 2019>