Cloud computing has revolutionized how modern companies build, deploy and scale software to keep pace with the speed of business. According to Gartner, end-user spending on public cloud platforms will grow by 20.7% in 2023 to total $591.8 billion. As enterprises accelerate their move to the cloud, open source platforms provide greater choice over proprietary vendors.
What Capabilities Are Companies Looking For?
Based on my experience advising CTOs and cloud architects at over 100 companies, key selection criteria often include:
- Flexible scaling to match workload demands
- Portability across cloud environments
- Interoperability with existing infrastructure
- Security and compliance to manage risks
- Speed and agility to accelerate releases
- Cost savings compared to proprietary licensing
- Innovation fueled by open source communities
Open source cloud platforms can deliver on all these fronts while avoiding vendor lock-in. This allows enterprises to reduce spending on data center operations and divert resources towards building value for customers instead.
What Are Open Source Cloud Platforms?
Open source cloud platforms provide pre-built infrastructure or application runtimes to deploy workloads using pooled virtualized resources. Leading options include:
- Infrastructure-as-a-Service – Resources for compute, storage and networking
- Container Orchestration – Automation for managing container deployments
- Platform-as-a-Service – Application runtimes and developer services
According to a Flexera 2021 report, 31% of enterprises leverage open source clouds while 49% use both open and proprietary platforms.
Below are 7 widely-adopted open source platforms for building cloud-native strategies:
1. Cloud Foundry: Leading Open Source PaaS
Created by VMware and now hosted by the Cloud Foundry Foundation, Cloud Foundry is the industry‘s leading open source platform-as-a-service (PaaS). With a thriving ecosystem and over 60% of Fortune 500 companies as users, it delivers a scalable runtime for deploying cloud-native applications.
Key Architecture Components
- Router – Handles inbound requests and routes them to appropriate component
- Application Containers – Encapsulates running application processes
- Service Providers – Manages bound services for access by applications
- Cloud Controller – Primary management component that monitors and coordinates processes
Backed by Pivotal, IBM, SAP and other major tech providers, Cloud Foundry supports a "build pack" model for detecting programming languages and frameworks to automatically configure containers. It allows developers to use languages and tools of their choice without proprietary lock-in. Companies can run Cloud Foundry on their existing infrastructure or use supported public cloud platforms that offer it as a managed service.
According to James Watters, SVP of Product, Business & Corporate Development at Cloud Foundry Foundation member SUSE, "Cloud Foundry provides the velocity needed for digital innovation by streamlining the application delivery experience for developers."
When To Use: Best suited for a PaaS model that avoids underlying infrastructure configuration while enabling CI/CD automation. Provides optimum developer experience and productivity.
2. OpenStack: Leading Open Source IaaS
Started in 2010, the OpenStack project provides open source software for powering public and private clouds. Used by major companies from BMW to PayPal, OpenStack runs in production environments across over 75 public cloud data centers and tens of thousands of private clouds at a scale of >10 million compute cores.
Key Architecture Components:
- Nova – Compute engine for managing virtual machines
- Neutron – Powerful networking as a service
- Cinder – Persistent block storage
- Swift – Redundant object/blob store
- Horizon – Dashboard for access and administration
With a strong ecosystem of solutions built on top of its core infrastructure components, OpenStack provides immense flexibility. It allows enterprises to optimize costs using existing hardware investments along with burstable access to public IaaS cloud providers within a hybrid model.
As Mark Collier, COO of the OpenStack Foundation, comments, "Many of the world’s largest brands use OpenStack because it provides a flexible infrastructure platform that can span multiple data centers and clouds."
When To Use: Ideal for any company wanting to offer private cloud services or with complex infrastructure needs not easily met by public cloud providers. Enables hybrid cloud models.
3. Apache CloudStack: Bring AWS-Like Capabilities In-House
Apache CloudStack is an IaaS platform used by major service providers to offer public cloud services and enterprises to run private clouds. Led by companies using CloudStack including British Telecom, Disney, Nokia and Orange, it offers an easy way to utilize existing infrastructure to deliver compute, storage, networking, security and other fundamental cloud capabilities.
Key Architecture Components:
- Cloud Infrastructure – Pool of resources like compute, storage and network
- Management Server – Allocates resources, tracks usage and enforces constraints
- Database – Storage for user accounts, resource configs and usage records
- System VMs – Specialized VMs that perform administrative operations
With over 300 contributors and many global users running production workloads at massive scale, CloudStack provides an alternative to vendors like VMware and proprietary stacks. It offers compatibility with Amazon Web Services including support for AWS APIs to help transition applications between environments.
As Giles Sirett, Director, Solutions & Technology at CloudStack collaboration partner ShapeBlue notes, "Whether deployed on-premises or by a service provider, Apache CloudStack unlocks robust, secure and highly scalable cloud environments with the benefits of open source."
When To Use: Well-suited for Telecoms, Hosting Providers and MSPs that want to mimic AWS functionality using their existing infrastructure with full customizability.
4. Kubernetes: De-Facto Container Orchestration
Initially created by Google based on their internal Borg system, Kubernetes has become the de-facto open source leader for managing Linux container deployments across clusters of hosts. Backed by the Cloud Native Computing Foundation with contributions from AWS, Microsoft, Red Hat and over 1200 other companies, Kubernetes delivers powerful automation required for running containerized applications in production.
Key Architecture Components:
- kubelet agent – Runs on cluster nodes to start/stop containers
- Control Plane – Manages the overall cluster and workload scheduling
- Container Network Interface (CNI) – Enables networking between pods
- Container Runtime – Underlying software like Docker that runs containers
With robust community support and an extensive ecosystem of tools like Helm and Operators, Kubernetes enables a cloud-native approach for organizations embracing microservices. It handles dynamic requests and failovers without needing manual intervention to place workloads.
As Taylor Dolezal, Technologist at CNCF member company Thanos IT notes, "Kubernetes has become so widely adopted due to its flexibility. Whether on-premise or across multiple public clouds, it provides a reliable abstraction layer to manage containers at scale."
When To Use: Any organization running containerized applications that require automated scaling, networking, security controls or data persistence.
5. Docker Swarm: Native Container Clustering
Included with Docker Engine, Docker Swarm provides simple native clustering capabilities. It pools Docker hosts and exposes them as a single virtual Docker host for deploying services across a cluster. While less common than Kubernetes, Swarm offers a fast way to create clusters without needing additional software just for users already leveraging Docker containers.
Key Architecture Components:
- Swarm manager nodes – Cluster management plane and API endpoint
- Swarm worker nodes – Host containerized application workloads
- Distributed datastore – Embedded key-value store for persisting cluster state
Swarm handles scheduling containers across nodes based on resource availability and application requirements. It utilizes the standard Docker Compose file format for declaring application services to deploy across the cluster. While missing complexity like auto-scaling policies of Kubernetes, Swarm provides a simple clustering solution accessible for Docker users.
As Gabriel Monroy, Lead Software Engineer at Deis Labs notes on Swarm, "Its tight integration with the Docker CLI makes Swarm an easy drop-in for converting Docker hosts into a multi-node cluster."
When To Use: Ideal for smaller container deployments using Docker hosts where a lightweight scheduling solution is sufficient rather than full orchestration capabilities.
6. Cloudify: Multi-Cloud Service Orchestration
Backed by over a dozen member companies and hundreds of users, Cloudify is an open orchestration platform designed by Israeli startup Cloudify Platform Ltd. It utilizes TOSCA, an OASIS standard, for orchestrating infrastructure resources, platform services, complex applications and container management software like Kubernetes across any cloud infrastructure through a unified workflow.
Key Architecture Components:
- Cloudify Manager – Orchestration logic and cluster management
- Agents – On managed resources to communicate with manager
- Plugins – Connectors to different infrastructure
- CLI/UI – Interfaces for invoking & monitoring orchestration
Service providers like Verizon use Cloudify for simplifying management across AWS, Azure and OpenStack. With robust policy-based auto-scaling and self-healing capabilities driven by metrics and events, Cloudify fits organizations using multi-cloud or hybrid cloud infrastructure.
As Zeev Bikowsky, Co-Founder & CTO of Cloudify Platform states, “Cloudify’s open orchestration significantly simplifies IT operations by delivering environment and app portability through automation.”
When To Use: Companies embracing multi-cloud or hybrid cloud for production systems where workloads span different environments requiring coordination.
7. OpenNebula: Build Hybrid Cloud Environments
Led by C12G Labs as part of the OpenNebula Consortium, OpenNebula provides an open-source management platform aimed at enterprise data centers to build private, public and hybrid clouds. Combining existing infrastructure with connectivity to external clouds, OpenNebula provides a unified provisioning layer.
Key Architecture Components:
- Front-end – Interfaces like CLI, API and web UI
- Core – Management from dynamic compute/network pools
- Drivers – Hardware, storage and platform integrations
- Scheduler – Allocates resources matching ranking criteria
OpenNebula leverages enterprise virtualization platforms like VMware vCenter along with AWS, Azure and Google Cloud integrations. This enables fully automated application scaling and migration across environments. With over 50 companies sponsoring and contributing to its advancement, OpenNebula offers an alternative open source option to VMware.
According to Jaime Melis, Director of Global Services and Support at OpenNebula Systems, "OpenNebula provides the tools needed for any enterprise looking to build a production-ready hybrid cloud that avoids vendor lock-in."
When To Use: Organizations with existing VMware infrastructure investments looking for tools to link data centers with public cloud platforms. Useful for automated application portability across environments.
Making the Right Open Source Cloud Platform Choice
With powerful options now available for open source clouds, the landscape continues to rapidly evolve with new projects and innovations delivered through collaborative development models. These community-driven platforms offer excellent alternatives to proprietary vendors and licensing fees.
However, every solution has tradeoffs. Factors like specific feature requirements, in-house skill sets, ability to handle support/maintenance and integration with existing systems should drive technology selection decisions.
By understanding strengths of different open source cloud platforms profiled here, enterprises can find the approach that best aligns to their cloud-native journey and business objectives while keeping budgets on track. The community of contributors and users surrounding these platforms also create abundance of guides and best practices to learn from others traversing similar challenges.
For companies undertaking IT modernization initiatives or needing more agile foundations for delivering customer-facing innovations, leveraging open source clouds in place of proprietary platforms offers immense strategic value. With options to run anywhere, integration support and interoperability with the wider cloud-native ecosystem, they provide the velocity and freedom required for powering next-generation solutions.
