Flexible, open, fast and reliable

Architectural freedom. Fully automated operations. Telco-grade OpenStack for all.

Canonical’s OpenStack on Ubuntu gives you the flexibility to place your OpenStack services exactly where you want them, while sharing all the operational code with a large community.

Contact us for OpenStack insights

Telco NFVI architecture

With SR-IOV for line-rate VNFs and CI/CD for dynamic cloud operations, our telco reference architecture for OpenStack meets both regulatory and operational requirements for highly automated telco infrastructure.

Move beyond single-VNF clouds or NEP-specific clouds, and create a general platform for cost-effective network operations across vendors.

Finance architecture

Isolate critical services for high security OpenStack with simplified audit and compliance assessments.

Building and operating OpenStack for financial services enables us to address EU and US regulatory requirements.

Private cloud infrastructure on Ubuntu is part of the fintech revolution, from blockchain to digital transformation.

Learn more about Ubuntu and financial services ›

Media architecture

Hyper-converged storage and compute ensure maximum cost effectiveness for large-scale media processing, transcoding, digital workflows and storage.

Kubernetes on top provides a high-performance workload operations regime with direct access to underlying media content.

Read about job concurrency in Kubernetes

AI / ML and HPC architecture

GPU acceleration of AI and machine learning workloads on OpenStack requires careful configuration of the underlying hardware and host OS. Canonical’s Ubuntu is the leading platform for public cloud GPU instances and Canonical offers private cloud expertise to match.

Build a GPU cluster and share it with multiple tenants using Canonical OpenStack - then operate Kubernetes on top for HPC and high-throughput AI / ML data science.

Learn about developing AI models with Ubuntu ›

VMware migration

Migrate your existing VMs to Canonical OpenStack — quickly and cost-effectively.

Canonical provides workload analysis and migration services. 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-effective to run there.

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

Operations Automation

The biggest challenge of OpenStack is everyday operations automation, year after year, while OpenStack continues to evolve rapidly. Canonical solves this problem with model-driven operations that decouple your architectural choices from the operations codebase that supports upgrades, scaling, integration and security.

Model-driven operations deliver true open source automation

In traditional automation environments like Ansible, Chef and Puppet, your architectural choices are hard-coded in the configuration management system. That means two different companies who need two different architectures cannot really share ops code. Every environment becomes its own unique snowflake. That’s expensive.

With Canonical’s approach to OpenStack:

  • Operational code is shared across different architectures
  • Daily operations are fully automated
  • Upgrades are guaranteed for every OpenStack version
  • Security improves for every deployment at the same time
  • Third-party integrations are shared as soon as they are ready

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

Watch our OpenStack Upgrade webinar

Total automation of bare metal infrastructure

Operating private clouds means taking responsibility for physical infrastructure. Ubuntu OpenStack uses MAAS, the leading open source software-defined-data-center system and open source IPAM, to automate commissioning, allocation, provisioning and secure recycling of every 乐天堂fun88备用网址 in the data centre.

MAAS enables multi-vendor and multi-OS data centre operations:

  • IPMI Board and Chassis Management Controllers
  • HP, Dell, Supermicro, Quanta, Lenovo and others
  • Cisco UCS, Open Compute Foundation, Intel RSD
  • PXE, DNS, DHCP, NTP, IPv4 and IPv6
  • Windows, CentOS / RHEL, Ubuntu deployment
  • Hardware inventory and asset management
  • IP Address Management (IPAM) built in
  • Automated testing of NICs, disks, CPU and RAM
  • Scales efficiently with in-rack offload and proxy
  • White-box switches can provide in-rack services
  • Micro-cloud capability for VM creation
  • Edge cloud solution selected by major telcos
  • Single-command install and upgrade
  • Postgres database for reliable state
  • High availability for all services

Open source, with contributions from many companies. MAAS is used for bare metal operations in high performance computing, finance, gaming, telco edge and core, and many other sectors.

Hyper-converged, containerised or isolated services

Canonical leads the world in containerising the OpenStack control plane, enabling ultra-efficient placement of OpenStack services alongside guest compute and storage, reducing OpenStack cloud costs.

  • A containerised control plane means that OpenStack services are deployed in containers rather than dedicated machines, which reduces the number of machines required in smaller cloud environments and puts more of the machinery to work for guests and tenants.
  • Hyper-converged architectures put storage, guest VMs and OpenStack services on every node in the cluster, using modern kernel QOS (quality of service) capabilities to defend the performance of critical services while reducing the number of overhead nodes required to service the cloud.
  • Isolated service architectures use dedicated machines for specific OpenStack functions, such as messaging, databases and API services, providing simpler guarantees of service levels in cloud functions and simpler security and compliance and audits.
  • VIP-HA architectures enable almost instant failover of critical front-end API services but require those services to reside on a single physical network segment. DNS-HA methodologies use short-lived TTL DNS record updates to redirect users to backup API endpoints in the event of failure, with some delays to service recovery but more flexibility of placement and routing for large-scale environments.

OpenStack Consulting

Our OpenStack consulting and design services will help you choose the right mix of these capabilities for the particular stage and scale of your private cloud journey; having built hundreds of private clouds, we’re well placed to advise on the tradeoffs and steer you to the right architecture. Perhaps most importantly, though, we can evolve OpenStack deployments over a series of iterations as your needs change, which keeps your options open.

Learn more about our consulting services ›


Every ounce of performance matters

If you’re building a private cloud you want the maximum performance for your workloads, the maximum utilisation in your data center, and the maximum economic efficiency. Canonical delivers all three.

Understanding compute price/performance in a cloud environment is non-trivial. Multiple factors shape cloud performance and efficiency - from your choice of architecture, networking, storage and ultimately the hardware you build it on. Canonical has built enough clouds to provide deep insights into your architecture for utilisation and performance.

  • How to select boards and CPUs for total throughput
  • Assess RAM, disk and network performance impacts
  • Integrate GPU, FPGA and other accelerators
  • Offload functions to specialised NIC and storage systems
  • Assess and raise the level of 乐天堂fun88备用网址 utilisation
  • Balance compute load in the cloud
  • Optimise for specific workloads
  • Assess overcommit ratios
  • Specify instance types

Economics are the ultimate measure of effectiveness in the modern compute substrate; your private cloud has to be as efficient as your public clouds or it won’t attract internal workloads.

Kubernetes support included

There’s no doubt that Kubernetes is the new standard operational layer for every multi-cloud business. Canonical partners with Google to enable K8s on demand for your devops teams - on OpenStack, on VMware, on public clouds, and on bare metal clusters with MAAS. Kubernetes support is included on a simple per-node basis in your OpenStack cloud. No additional paperwork necessary.

Deliver ‘Containers as a Service’ across the enterprise with Canonical’s Distribution of Kubernetes (CDK) which enables each project to spin up a standardised K8s of arbitrary scale, on demand, with centralised operational control.

  • Multi-cloud workload portability thanks to Ubuntu
  • Same Ubuntu OS and kernel as Google, Azure, AWS k8s offerings
  • PCI device passthrough for GPU, FPGA and SR-IOV workloads
  • AI and machine learning optimisations with nVidia and Google
  • Flavours to match public cloud k8s on Google, Azure, AWS
  • Standard upstream Kubernetes, CNCF builds available
  • Portfolio of third-party certified solutions for CI/CD
  • Day-2 operations encapsulated and included
  • Upgrades guaranteed between K8s releases
  • Security updates by Canonical just like Ubuntu
  • Full OCI compatibility with docker and containerd runtimes
  • Prometheus and ELK logging and monitoring included
  • Extensible, third-party ecosystem for storage, networking

In partnership with Google, Canonical delivers the perfect K8s for AI and machine learning with Kubeflow.


Storage performance and economics are tricky to balance in a cloud environment. Canonical will help you architect your storage across the cluster to balance price and performance, ensuring the right mix of resilience, latency, IOPS and integrity for your particular deployment.

  • 48TB raw storage per node included
  • Mix of local direct-attach and network storage
  • Distributed software-defined storage vs SAN assessment
  • Performance tuning for local and network storage
  • Understanding storage compression
  • Understanding storage resilience
  • Third-party storage solutions available for integration
  • Object stores and other higher-level storage constructs
  • Hierarchical storage configurations

Learn more about storage with Ubuntu Advantage for Infrastructure ›


An open source distributed storage platform, providing a flexible storage option for OpenStack, Kubernetes and standalone storage clusters.

Learn more Ceph ›

Other platforms

Canonical’s Charmed OpenStack provides support for various other storage platforms including Swift, SAN, Nexenta, Pure Storage and StorPool.

Learn about our consulting services ›


Aggregate network throughput and latency across the underlying layer is a major factor in cloud performance. Canonical’s work 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. Our work with telco for NFV and edge clouds ensures that we can work well in complex environments where latency and security are critical.

  • Impact of switch architecture choices on cloud performance
  • Latency measurement across distributed clusters
  • SR-IOV and function off-loading on cloud NICs
  • NFV implications of particular hardware selections
  • Underlay networking architecture design
  • Underlay network security design implications
  • IPAM and DNS for underlay networks
  • Tenant network router options and optimisation
  • VLAN, OVS and Tungsten Fabric tenant networking
  • Cisco ACI, Juniper Contrail and other proprietary SDN options
  • Company firewall operations for private clouds
  • BGP route propagation from OpenStack tenants
  • MPLS integration for telco environments

Logging and Monitoring

Operations in highly coherent large-scale distributed clusters require a new level of operational monitoring and observability. Canonical delivers a standardised set of open source log aggregation and systems monitoring dashboards with every cloud, using Prometheus, the Elasticsearch and Kibana stack (ELK) and Nagios.

These dashboards can be customised or integrated into existing monitoring systems at your business.

Find out how Canonical’s OpenStack on Ubuntu can give you the flexibility to share operational code with a large community whilst placing your services exactly where you want them.

Contact Canonical about OpenStack