Flash back to five years ago. The infrastructure space looks drastically different. Some companies, like the one I worked for at the time, were figuring out how to successfully orchestrate containers at scale. Yet for the majority of businesses, containerization wasn't even on their radar. Given how quick-moving and young containerization was, many of the critical cloud-native ecosystem projects we rely on today simply did not exist, and nor did Containership.
Scratching our own Itch
Being part of a small operations team, my cofounder Phil and I were frustrated with the bugginess of orchestration systems at the time, the overhead of running systems like Mesos and ZooKeeper, needing to build our own service mesh and loadbalancing, and getting this to work consistently across regions / clouds. Naturally, we decided to build a container scheduler to solve the problems we were facing day-to-day. The goal of the project was to provide organizations a batteries-included, but removable experience when running containers at scale on any cloud provider. There were a few principles we kept in mind when architecting Containership:
- The software should ship in a single package to reduce operational overhead and ongoing management. We want the pieces to fit like a glove.
- User experience is paramount. Simplicity is key when dealing with lots of information and complex systems.
- Everything we build should be cloud-agnostic. Facilitating a move to multicloud is our mission.
The Launch of Containership Cloud
In early 2015, after countless hours of night and weekend work, we decided to apply to TechCrunch Disrupt Startup Battlefield with an idea for a software as a service (SaaS) product to deploy and run our open source containership orchestration project across clouds. To be clear, we did not yet have the SaaS product built, nor a formal business around the idea. To our surprise, we were accepted, and had a month to build a product around our orchestrator before we were to present it on stage in front of an audience of thousands. We called the product Containership Cloud. Not super original, but the name stuck.
With the launch of Containership Cloud, we supported automatically deploying containership clusters to DigitalOcean and Amazon Web Services, as well as moving infrastructure between the providers using an innovative snapshot and restore system. We wanted to demonstrate that the future of infrastructure was multicloud, and workload portability was a reality. In a world where AWS was the big winner, our goal was to commoditize the cloud. The launch at TechCrunch Disrupt signified traction for Containership, helping us raise a round of seed investment less than a year later.
In the months that followed, we hired a stellar team to further build out the Containership Cloud Platform technology, as well as partnership and marketing channels. We expanded the platform to support 12 cloud providers, as well as any attached open source containership installations, accommodating virtually the entire cloud and on-premises market. We built out and maintained over 50 marketplace stacks which could be deployed to containership infrastructure with a single click, from simple blog software to complex systems such as Cassandra. Coupled with log aggregation, metric collection, full infrastructure backups, IAM, and SSH key management, we felt Containership Cloud was the most comprehensive solution to deploy and manage multicloud containerized infrastructure in the market.Unfortunately, while user adoption was good, we struggled to monetize the business in a meaningful way. We were constantly experimenting with per-user, per-node, and per-application pricing strategies, without profound success. During this period we started to notice a common theme from our more well-established users and prospects. They wanted to know what differentiated containership from Kubernetes.
The Rise of Kubernetes
I won’t bore you with the Kubernetes backstory, but it is important to note that the project is the third fastest growing open source project of all time, and as I’m sure you’re aware, has fundamentally changed the infrastructure landscape in the past few years. In 2017 we had to take a hard look at ourselves and what we were building. We had to ask ourselves the same questions our customers were asking: “What makes us different?” We quickly realized our greatest value was not necessarily at the orchestration layer, but rather it was the intuitive multicloud control plane we were building (Containership Cloud). Though difficult, we made the choice to abandon our work on open source containership, and pivot to delivering customers the same sort of experience using Kubernetes. The choice to refocus on the next layer up was beneficial in that we no longer had to split our small team across core orchestration work and product work.
In 2018, at KubeCon EU in Copenhagen, we officially launched our first version of Containership Cloud which included support for deploying Containership Kubernetes Engine (CKE) clusters. While our launch only included support for running on AKS, GKE, and CKE on DigitalOcean, we were able to quickly add support for over 70% of the cloud market. Since that time, we have:
- Interfaced with the Kubernetes and cloud-native community through participation in the ClusterLifecycle and MultiCluster SIGs.
- Released Cerebral, a pluggable cluster autoscaling solution.
- Released Konstellate, a GUI to build and visualize Kubernetes applications.
- Shipped six versions of our certified Kubernetes distribution to end users, providing incredibly simple, in-place cluster upgrades with the click of a button.
- Unlocked IAM by providing centralized user, team and RBAC management in Containership Cloud Platform, including comprehensive centralized audit logging.
- Built the most usable dashboard for managing multicloud infrastructure, bar none.
I’m incredibly proud of the work our team has done on the Containership Cloud Platform, CKE, and in the community. Over the past four years, we’ve had the fortune of operating infrastructure for thousands of organizations who have valued our tools, and in turn, we have learned a great deal.
Ultimately, we came up short. It has been nearly a year and a half since the Containership Cloud replatform, and we have failed to monetize Containership Cloud in such a way that we could build a sustainable business. As the founder and CTO, this is on me.
It saddens me that as of this week, Containership has ceased operations. The Containership Cloud Platform will stay operational through October 31st, 2019, at which point we will be unable to provide further services to users, and the platform will be permanently shut down. It is important that your organization moves its infrastructure off of Containership Cloud in order to prevent unwanted disruption to your Kubernetes clusters. Please see our documentation for more information.
For those who have supported us over the years, we are genuinely thankful. We understand this may be a painful transition for your organization. If you have any questions, please contact email@example.com which will remain active through the end of the business’ shutdown process.
Norman Joyner, Founder & CTO