Apcera Platform 3.0 is Here
The Apcera Platform 3.0 release is now available! We’ve bumped the major version number to reflect the significant changes in our platform and its architecture, some of which you’ll see in this release and more to come in our follow-up releases.
The new features and changes in Apcera Platform 3.0 reflect the needs of our customers and our response to the latest dynamics and changes in the Container Management Platform (CMP) space. We’re excited to deliver solutions that help our customers solve their business challenges and continue to remain competitive in this emerging space. I've highlighted a few of the most exciting features in this release below.
First up is Job Autoscaling. Our goals for Job Autoscaling is to go beyond parity with other platforms and provide additional capability that is more relevant in a container environment. In additional to the usual metrics such as CPU usage and requests per second, our job autoscaling also supports custom metrics via a HTTP endpoint. For example, you can build a service to watch the queue of your billing system. You can configure Job Autoscaling for your jobs in the Apcera Platform to call the queue watching service and autoscale based on thresholds you define for this service.
Our implementation of Job Autoscaling also provides an alternative to using threshold based autoscaling; you can alternatively choose to define autoscaling based on Proportional-Integral-Derivative (PID) control theory, which is better suited to autoscaling jobs with long lived requests. With PID autoscaling configured for your jobs, autoscaling is done as a correction to achieve a setpoint metric. Definitely an interesting topic for a deep dive in a future blog post.
Another exciting new feature in the 3.0 release is that routes have been upgraded to “first class citizens” in the platform (previously they were an attribute of jobs). This gives you much more control over routes, being able to add, delete, etc. routes to jobs at a specified port. Both HTTP and TCP protocols are supported. By making routes independent of jobs, routes are easier to administer; including creating routes, mapping them to jobs, and balancing the weighting between jobs. Another awesome feature for routes in the 3.0 release is the use of wildcards for the route path!
Encryption At Rest
In the Apcera Platform 3.0 release, your application data is even more secure when you use the new Encryption at Rest feature. Job application data can be securely saved to local (ephemeral) or persistent (NFS or SMB volumes) storage using standards based encryption (LUKS and EncFS). Policy can optionally be used to automate the enforcement of encryption and supported service providers.
Cluster Configuration Storage
The platform now leverages Consul as a high availability store for cluster configuration key/value data. Vault technology has been incorporated into the platform to securely store passwords, certificates, and keys for both application secrets and cluster components. After encrypting secrets, Vault stores them within Consul.
Certificate and Key Management
You can install and manage SSL/TLS certificates and private keys for use with routes, giving you the ability to have with custom domains (outside the domain of your cluster) with secure endpoints. The platform’s secure secret storage is used to store the private keys and secrets.
And there’s more
This gives you a taste of some of the new features in 3.0. A few of the other new features include OpenAPI, Configurable Virtual Networks, and Storage Improvements. The 3.0 release also includes a sneak preview of our new installer. The installer, currently in beta, tremendously improves the user experience for creating and setting up clusters, and will be made available in the near future.
What’s your favorite new feature in the 3.0 release?
I'm interested in your feedback, email me directly at firstname.lastname@example.org