There’s no question that Amazon, Microsoft and the other cloud vendors have changed application development. The immediate availability of development tools and computing resources enables faster creation and deployment of applications that are critical to your business. Developers can grab a VM and leverage a host of cloud-specific service like Identity and Access Management, or storage, or even something more exotic like geospatial API services. And applications that are built using a single cloud provider’s resources are easy to deploy to and run well on that cloud.
At the same time, most of us in mid-size and large enterprises face requirements for regulatory conformance, international requirements for geographic data storage, legacy application requirements, and customer-specific requirements that lead us to hybrid and multi-cloud strategies. Our IT organizations feel that they don’t have enough control over their cloud usage and spend. And this problem is only getting worse.
Do we really need to accept cloud vendor lock-in as the price for speed? And if so, does that mean abandoning hybrid cloud and multi-cloud strategies?
As we seek to gain control over where, when and how we utilize cloud provided environments newer development and deployment technologies are helping us to gain control. Development teams are increasingly using Linux or Docker containers because of their isolation and portability benefits. Containers “wrap” code and isolate it from software and hardware dependencies making application development and testing faster and easier. Kubernetes on-premise deployments are also much easier today than they used to. Likewise, the growth in the use of Microservices lends itself to containerization for rapid deployments and upgrades of these “micro” services. And the defacto choice for container management, Kubernetes, enables developers to use objects like Persistent Volumes, Ingress Rules and Services to abstract applications from infrastructure.
The combination of containers and Kubernetes can make applications deployable to multiple clouds and portable among on-premise and Cloud environments. But as with their development tooling, Cloud providers are offering Cloud-specific Kubernetes implementations, threatening to bring a new wave of Cloud vendor lock-in.
But Development and Operations teams can choose to use a more open version of Kubernetes. With the right development practices, the portability challenge becomes an Operations challenge of being able to deploy, run, and manage containerized applications across multiple environments. And to do so in a way that provides consistent security, built-in monitoring, back-up and restore capabilities, and that makes it easy for Operations teams to deploy, run, and manage clusters.
Kublr’s kubernetes management platform enables containerized applications and services to be deployed on-premise or on multiple cloud providers. Kublr even makes it easy to set-up the infrastructure that your container clusters are being deployed on top of by leveraging cloud specific and VM Specific provisioning APIs. And if you prefer you can deploy to bare metal servers.
Giving your operations team the ability to Deploy Where You Want enables you to leverage the right cloud economics strategy, whether it’s spot, reserved or based on some sustained usage formula. It can enable you to ensure that your cloud workloads can meet regulatory requirements on where data resides, whether that be on-premise, in a particular geographic region, or in an approved data center. And it provides you with the flexibility to move workloads to meet future, unanticipated business requirements.