Your submission was sent successfully! Close

You have successfully unsubscribed! Close

Thank you for signing up for our newsletter!
In these regular emails you will find the latest updates about Ubuntu and upcoming events where you can meet our team.Close

Enterprise private cloud

Total automation. Flexible technology choices.
Enterprise features. Stability and security above all.

Everything you need for your mission–critical workloads.

Build your own production-grade cloud

Watch the webinar: "Open infrastructure for data science" ›

General purpose architecture

For organisations of any type looking for a high performance enterprise private cloud at a low cost.

  • The best price–performance guaranteed
  • Fully automated day–2 operations
  • Hyper–converged and other architectures
  • Interoperability across various platforms
  • Integrated observability stack

See the benefits of Charmed OpenStack ›

Public cloud architecture

For service providers looking for a cost–effective alternative to hyperscalers.

  • Fully automated scale out capabilities
  • IaC and CI/CD integration
  • Extendable containers layer with K8s
  • Various storage platforms supported
  • Ubuntu and third–party workloads

Read how Nayatel built public cloud with Charmed OpenStack ›

Telco NFV architecture

For telcos looking for a network function virtualisation infrastructure.

  • 100 Gbps networking with hardware offload
  • Numerous performance extensions included
  • All major SDN controllers supported
  • Data encryption and data protection
  • Tenant–level IPv6 support

How service providers leverage OpenStack ›

Finance and government architecture

For institutions that value stability and security above all else.

  • Predictable release cadence and upgrade path
  • Up to 10 years of security updates
  • US and EU compliance programmes
  • OS, OpenStack and K8s hardening
  • Containerised control plane

Why financial institutions use Ubuntu ›

Need an architecture tailored to your needs?

Talk to us about your plans

Better enterprise private cloud platform

No two cloud platforms are the same. Have a look at the table below to see how Canonical’s Charmed OpenStack differs from other enterprise private cloud platforms.

Charmed OpenStack VMware AWS Outposts Azure Stack
No mandatory subscription
Full automation
Region freedom
Technology freedom
Professional services
Enterprise support
Self–managed option
Fully–managed option
Stable release cadence
Open source
Check other OpenStack distributions ›

Move to Charmed OpenStack

Wherever you are coming from, Canonical offers hussle–free migration

The main driver for most OpenStack deployments is the cost benefit of using a leaner and more open IaaS. Canonical's Charmed OpenStack also offers a range of cloud services, and compatibility with hybrid cloud and multi-cloud operations tools. While legacy virtualisation continues to be important, the future is cloud APIs and container-based operations, which Charmed OpenStack delivers.

Canonical provides workload analysis and migration services from VMware and other cloud types. Not every workload is best run on a cloud, but 80-90% of enterprise workloads can move to OpenStack easily, and are much more cost-efficient to run there.

Move from VMware to OpenStack ›

Total automation

The biggest challenge of OpenStack is everyday operations automation, year after year, while OpenStack continues to evolve rapidly. Canonical solves this problem with total automation that decouples architectural choices from the operations codebase that supports upgrades, scaling, integration and bare metal provisioning.

OpenStack Charms package cloud operations

In traditional automation environments based on Ansible, Chef or Puppet, architectural choices are hard-coded in the configuration management system. That means the operations code cannot be shared across various architectures, organisations and market sectors. Every cloud becomes its own unique snowflake. This approach does not scale and results in significant costs.

With OpenStack Charms:

  • OpenStack operations code is packaged for reusability across different architectures, organisations and industries
  • Upgrades are guaranteed between consecutive OpenStack versions
  • Daily operations, such as cluster scale out or database backups, are fully automated
  • OpenStack complexity gets abstracted in favour of model-driven operations
  • Security and quality improves for every deployment at the same time

You don't need to hire a large operations team that follows every aspect of OpenStack development. Just focus on high-level operations and your business workloads.

Watch our OpenStack upgrade webinar

MAAS turns a data centre into a bare metal cloud

Running private clouds requires setting up and maintaining the underlying physical infrastructure. Before organisations can even start deploying the cloud, all machines have to be racked, powered, cabled and the base OS has to be installed on them. Performing all of these tasks manually becomes expensive as the number of machines grows.

With MAAS:

  • A data centre turns into a bare metal cloud
  • All machines in the data centre are automatically discovered and analysed for their base resources
  • Organisations can automatically provision physical machines for the purpose of running their cloud services
  • Hardware inventory and IP Address Management (IPAM) solution become an integrated part of the cloud platform
  • OpenStack cluster can be scaled out in minutes when the demand for workloads increases

You no longer need to focus on physical infrastructure operations. Just set up your data centre and let MAAS discover your machines and provision them for you on-demand.

Read our MAAS ebook

Managed OpenStack

OpenStack is a complex system consisting of various interconnected components. Therefore, its operations are challenging even in a fully automated environment. We recognise that and offer fully managed services for organisations who do not have sufficient OpenStack knowledge, but want to go to market fast. With Canonical's Managed OpenStack our customers get 24x7 cloud monitoring, daily maintenance, regular software updates, OpenStack upgrades and more. Managed OpenStack is also a more economical option up to 300 nodes.

Learn more about Managed OpenStack ›

Hyper-Converged or other architectures

Designing the cloud for the best price-performance starts with making an architecture choice. Although the Hyper-Converged architecture is the most often chosen by organisations, Canonical supports other architectures as well to meet customers' needs.

  • Hyper-Converged architecture puts guest VMs, storage and OpenStack control services on every node in the cloud, using modern kernel quality of service (QoS) capabilities to defend the performance of critical services while reducing the number of overhead nodes required to service the cloud. Such an approach allows standardisation of hardware configuration and enables maximum resource utilisation, leading to the best price-performance.
  • Converged architecture puts guest VMs and storage on every node in the cloud, but isolates OpenStack control services. Since three units of each OpenStack control service are needed for HA, the converged architecture requires three additional nodes. In order to isolate control services even more, each of them is running inside of a dedicated container which increases the overall security level.
  • Fully Disaggregated architecture allows for even more freedom in terms of services placement. Users can run dedicated compute and storage nodes or even configure various classes nodes. Designing custom architecture is often a requirement of local compliance regulations. It is also very common in telco NFV environments.


The biggest challenge when building a private cloud is to maximise performance of workloads, utilisation of data centre resources and cost efficiency at the same time. This is challenging as multiple factors shape this multidimensional curve, from cloud architecture to hardware choices.

Canonical partners with leading silicon and hardware vendors to ensure the latest improvements in the compute space are tested and validated on Ubuntu. This enables us to design a cloud architecture that answers the performance needs. We also assist with workload analysis to make sure that the cloud we build has an optimal capacity. In order to guarantee the maximum cost efficiency, we run a bidding process with leading hardware vendors on behalf of our customers.

With Canonical's Charmed OpenStack you get:

  • Access to the latest technologies, including GPUs, FPGA and other accelerators
  • Architecture and hardware guidance
  • Workload analysis and cloud capacity assessment
  • Total cost of ownership (TCO) per VM estimates
  • Performance benchmarks

Virtual machines

OpenStack leverages the leading virtualisation stack based on QEMU/KVM/libvirt to provide the best user experience for virtual machines management.

Open virtualization stack ›


Kubernetes provides a cloud-native containers layer on top of OpenStack with automated scheduling, rolling updates and autoscaling support.

Enterprise multi-cloud Kubernetes ›


Network is an essential component in every data centre. A poorly designed network can significantly affect the overall performance of the cloud and become a bottleneck in the traffic flow. On the other hand, network improvements should always be driven by economics so as not to affect the overall budget.

Canonical's partnership with hyperscale public clouds ensures that we have deep insight into the dynamics of cloud network performance and security best practices for large-scale multi-tenanted operations. Moreover, our experience with telcos for NFV and edge clouds ensures that our solution works well in complex environments where latency and security are critical.

With Canonical's Charmed OpenStack you get:

  • Up to 100 Gbps tenant-level performance
  • Underlay network architecture and security design
  • Hardware offloading and other acceleration technologies
  • SR-IOV, DPDK and other performance extensions
  • Access to a wide range of SDN platforms

Open source SDN

Open Virtual Network (OVN) and Open vSwitch (OVS) are a foundation of an open source SDN which is gaining popularity, especially in general-purpose OpenStack deployments.

How to migrate from OVS to OVN

Proprietary SDN

Many organisations are still tied to their existing vendors. We recognise that and partner with leading network equipment providers (NEPs) to ensure their SDN platforms are tested and validated with Charmed OpenStack.

Deploying SDN on OpenStack ›


Storage is one of the most tricky components when architecting a cloud. Storage costs per TB vary by several orders of magnitude depending on the storage types and technologies used. Thus, making careful selections when designing storage architecture is essential from a price-performance point of view.

Canonical's approach towards storage architecture optimisation leverages multiple storage tiers. While big, cheap, low performance disks are used as ultimate storage devices, several smaller, more expensive, high performance devices serve as a cache in front. This approach ensures the required performance of the entire cloud, while not affecting hardware costs significantly.

With Canonical's Charmed OpenStack you get:

  • Up to 192TB raw storage per node
  • Data encryption in transit and at rest
  • Data protection and data durability
  • A mix of block and object storage based on personal requirements
  • Multi-tier cloud-native storage based on Ceph or other technologies


Ceph is an open source distributed software-defined storage platform, providing a flexible storage option for OpenStack, Kubernetes and standalone storage clusters.

Ceph storage on Ubuntu ›

Other platforms

While Ceph provides the best price-performance rate, it is not always an option. Therefore, Canonical's Charmed OpenStack provides support for various other storage platforms including Swift, iSCSI and more.

Explore other storage options

OpenStack consulting

Designing an enterprise private cloud is a non-trivial task. Making wrong architectural choices results in an increased TCO and a lot of rework on the next states of the cloud journey. We understand that and offer consulting services to help our customers choose the right mix of features to find the best cloud architecture for their needs. Moreover, our experienced team of cloud experts also deploys the cloud according to the requirements. All of that at a fixed price.

Learn more about our OpenStack consulting ›

Security and compliance

Building an enterprise private cloud entails additional security and compliance challenges. While in the public cloud, these activities are handled by the cloud service provider out of the box, when deploying an on-premises infrastructure, all must be undertaken by internal teams.

Canonical's Charmed OpenStack includes up to ten years of security updates, cloud hardening options according to common security benchmarks and compliance programmes to meet local US and EU regulations for enterprise customers.

What's included:

  • CIS and DISA benchmarks
  • CSEC and GDPR certification
  • HIPAA for health care institutions
Read a whitepaper about private cloud security ›

NVIDIA vGPU technology

GPU devices serve as a foundation for modern compute-intensive workloads, including data analytics, data visualisation and high performance computing (HPC). However, dedicating the entire device to a single workload may lead to sub-optimal resource consumption.

NVIDIA vGPU software democratises those contstraints, making highly performant GPU resources available on demand for all OpenStack users.

What's included:

  • Native host and guest drivers for NVIDIA vGPU Software
  • Automated deployment and operations of Charmed OpenStack with NVIDIA vGPU technology
  • Joint commercial support by Canonical and NVIDIA
Build open infrastructure for data science ›

OVN hardware offloading

In high-speed network topologies, such as those using 100 Gbps connections, network operations become a dominant factor. This creates high pressure on hypervisors' central processing units (CPUs).

With OVN hardware offloading all those computationally-expensive operations can be moved to SmartNICs and data processing units (DPUs). This frees up CPU cycles for the workloads and accelerates data plane network traffic.

What's included:

  • Fast path and slow path hardware offloading
  • Automated deployment and operations of Charmed OpenStack with OVN offloading capabilities to SmartNICs and DPUs
  • Joint commercial support by Canonical and NVIDIA
Read NVIDIA's reference implementation ›

Observability stack

Operations in highly coherent large-scale distributed clusters require a new level of observability. Cloud operations teams have to be able to check cluster health, monitor cloud capacity and investigate individual server logs from a centralised operations dashboard.

Canonical delivers an integrated observability stack for logging, monitoring and alerting based on established open source technologies with every deployed infrastructure component, including Charmed OpenStack.

What's included:

  • Filebeat for log forwarding
  • Graylog for log aggregation
  • Elasticsearch for log analysis
  • Prometheus for cloud monitoring
  • Telegraf for metrics collection
  • Grafana for metrics visualisation
  • Nagios for alerting

OpenStack support

Since OpenStack is open source, enterprise customers rely on the vendor to provide ongoing support for their infrastructure. Canonical provides full-stack commercial support under the Ubuntu Pro + support subscription which includes phone and ticket support, production-grade SLAs, ten years of security updates and security and compliance programmes. The subscription is charged per node and covers all layers of the open infrastructure stack: from bare metal to microservices.

Get support