Codefresh Adds a Command Line Interface for Kubernetes Management
He says 2017 brought clarity to the container orchestration market, with Kubernetes the dominant player in the container orchestration technology.
The startup, which released its Kubernetes-based continual delivery platform a year ago, teamed up with Google last August to make the technology more accessible. For its latest update to this platform, the company has added a command-line interface to its software.
“We saw one of the things clearly lacking in our platform to fully unleash the power of Codefresh and pipelines and sophisticated deployments to Kubernetes was command line interface (CLI). It’s a personal preference, but you can find a significant part of the community that prefers command line interface over the user interface. It allows [users] to do things they could not do with the Kubernetes command line,” he said.
The CLI includes the ability to:
- Build matrix pipelines, in which you can run the same pipeline multiple times in parallel with different values, such as building and testing a release for several different architectures at once.
- Manage images from multiple registries — The CLI exposes every connected Docker registry making it possible to add annotations, quickly search and review all images.
- Install and upgrade Helm charts.
He pointed to sophisticated Kubernetes users such as GIF file database Giphy, which uses the command line at the start of the automation script to run pipelines across different clusters and to test code for multiple internal services.
The CLI is the third major enhancement to Codefresh in the past couple of months, Tabib said. In December, it added support for Helm charts as native entities. They allow users to package microservices and networking components for Kubernetes.
Using Helm charts, you can automate more complex processes such as auto-provisioning the application, run integration tests against it and make canary deployments. And Codefresh users can employ the full platform inside of continuous delivery pipelines. Instead of pushing software changes to a staging server and running tests, the automated pipeline can spin up an environment on demand and load the full application for tests.
As the Docker image is being built and tested and deployed, Codefresh is adding more metadata, so eventually anyone in DevOps seeing the image running in production, whether it’s running well or not, it has all the information one click away about what this Docker image has been through, how it tested, whether it passed a security scan and more, Tabib said.
In a survey of 200 companies released last summer by 451 Research, 71 percent they are using Kubernetes to manage their container infrastructure.
Lawrence Hecht’s research for The New Stack, however, found the complexity of implementation and maintenance remains a major hindrance to adoption. A major theme of Kubecon 2017 in December was the need to make the technology boring, ensuring stability and usability.
In addition to working with companies now more comfortable in deciding on Kubernetes, Tabib said Codefresh will be working to incorporate two other open source projects in 2018:
- Istio, a proxy and a management layer on top of Kubernetes. Used along with Helm charts, it will enable users to easily set up deployment strategies such as canary and blue/green.
- And the monitoring solution Prometheus, allows users to check the health of the application amid the move toward increased automation.
The Cloud Native Computing Foundation, which manages Kubernetes, is a sponsor for The New Stack.