OpenStack
Architecture & Consulting Services
Expert OpenStack consulting and architecture services for companies building or scaling private or public clouds.
From initial idea to production, we help organizations design and deploy reliable and scaleable OpenStack environments
Cloudification is a Long Standing OpenStack Contributor
OpenStack Projects at a Glance
Professional Services for OpenStack by Cloudification
OpenStack is complex and initial architecture should not be underestimated as design decisions are hard to revise when the platform is in production. Our OpenStack architecture consulting covers topics such as:
- Multi-region Architecture
- Availability Zones and Nova Cells
- Highly-available Deployments
- Converged vs Hyper-Converged Deployment
- Stretched vs single Keystone per region
- Maintained and recommended OpenStack components
- OpenStack Regions vs Availability Zones vs Nova Cells
- Multi-tenancy in OpenStack: Keystone Domains, Projects and Sub-Projects
- Scaling and best Practices for OpenStack in Production
- Security, RBAC, Policies and Audit
Networking is perhaps the hardest part of any cloud platform and OpenStack is no exception. Cloudification team is ready to help with:
- Neutron design that scales
- Neutron ML2/OVS vs OVN
- Neutron Extensions and RBAC
- Neutron QoS and traffic shaping
- Neutron BGP dynamic routing
- Octavia HA LBaaS setup
- Windows guest VM network performance troubleshooting
- Correct MTU settings
- Native OpenStack VLAN management
- VXLAN vs Geneve tunneling
- NIC Offloading and Multiqueueing
With over 30 supported storage drivers in OpenStack it is easy to get lost. About of 50% of all OpenStack installations are using Ceph, yet Ceph performance might not fit all workloads. Cloudification will help you picking the best solution:
- Ceph deployment with Rook
- Ceph RGW Scalability and S3 API
- Ceph Replication vs Erasure Coding
- Ceph full OSD encryption
- Hardware specs for Ceph storage
- Metro Replication setups
- 3rd party backup tools: Storware, Trilio, Coriolis
- NetApp, Pure, Dell storage integration
- Cinder Backups vs Snapshots
- Cinder QoS
OpenStack upgrades have a reputation to be hard, but with the SLURP process there is only one upgrade per year. At Cloudification we manage over 20 OpenStack clouds and we have a well established process with GitOps automation to make upgrades smooth and fast. Other topics we guide on:
- DC capacity planning
- Hardware life-cycle, testing and monitoring
- Hardware tiering and DC operations
- Change management for Cloud providers
- Metering and Billing in OpenStack
- Over-provisioning in OpenStack
- Custom patches for OpenStack
- OpenStack release upgrade consulting
- OpenStack upstream development
Migrating from VMware, Nutanix, Proxmox or public clouds to private OpenStack cloud is a strategic decision that requires careful planning and execution. Talk with our OpenStack experts to cover advanced topics such as:
- VMware vs OpenStack feature parity
- Cloud workload analysis
- Cloud Native Architectures
- Kubernetes as a Service for OpenStack
- Tools for migration and DR
- os-migrate/virt v2v/migratekit/coriolis
- Performance benchmarks
- QoS configuration
- SLO, SLA and Error budgets
- Securing workloads in cloud
OpenStack Components We Work With
Häufig gestellte Fragen (FAQ)
Which hardware can be used for OpenStack?
Pretty much any modern x86 server hardware can be used regardless of the vendor.
Different hardware generations, CPUs and vendors can be mixed in the same cluster, although this imposes an additional operational overhead.
We recommend a minimum of 6 servers for production deployments, especially if using Ceph for storage.
Dedicated storage nodes and network are a must for public clouds and recommended for larger private clouds (30+ nodes).
For more information on picking hardware for OpenStack cloud check our blog post here.
Does OpenStack provide own hypervisor?
No, most of OpenStack deployments rely on KVM (Kernel-based Virtual Machine) which means a Linux base OS is required.
OpenStack supports other hypervisors such as XEN, however, KVM is the most used and best option.
For a complete list of features and hypervisors supported refer to Hypervisor Support Matrix und OpenStack Nova documentation.
How often should OpenStack be upgraded?
There are two releases each year and every second release is so-called SLURP.
SLURP stands for Skip Level Upgrade Release Process, an OpenStack release cadence model allowing to upgrade annually rather than every six months.
We recommend to upgrade at least once a year. Or twice a year in case you want to get the latest features.
Which base OS is recommended for OpenStack?
The requirements for the base OS distribution to be installed are rather relaxed, especially when OpenStack is deployed in containers. It should be Linux with modern kernel version supporting KVM, Libvirt and OVS/OVN.
Acording to recent OpenStack User Surveys about 50% of OpenStack installations are based on Ubuntu Server LTS. And that would be our default choice for c12n.cloud deployments.
In case you’d like to use a distribution other than Ubuntu or Rocky we strongly recommend verifying support of your deployment tools and additional drivers you might need (e.g. for GPUs or other devices).
How many OpenStack services (components) are out there?
There are 35 official, supported OpenStack services as of 2026.1 release.
Unfortunately, some services are not in a great shape or don’t receive many updates and therefore are not recommended for production use.
About 20 services of OpenStack can be recommended for production deployments with the rest requiring careful testing and evaluation.
All the core services such as Keystone, Nova, Neutron, Cinder, Glance, Placement are well supported and actively developed with hundreds of active contributions between the release cycles.
Where is the source code of OpenStack hosted?
The official home of OpenStack source is opendev.org and repositories can be found at https://opendev.org/openstack/
The same repositories are automatically mirrored to GitHub at http://github.com/openstack/
What is the official bug tracker of OpenStack?
Most OpenStack projects rely on Launchpad for bug and specification tracking.
Is there billing component available for OpenStack?
Ja! ✅
OpenStack has standard components for enabling metering and billing of cloud use.
There are also commercial tools available that are often combined with an alternative dashboard and payment provider integrations.
Check out our blog post for detailed overview of options.
Does Cloudification contribute to OpenStack upstream?
Ja! ✅
Over the past few years we have successfully contributed 150+ changes to OpenStack Nova, Cinder, Manila, Neutron, Designate, Ironic, Octavia and Barbican projects.
Our team includes Active Technical Contributors (OpenStack ATC), Core Reviewers and Summit Technical Committee volunteers.
What is OpenStack PTG (Project Teams Gathering)?
PTG stands for Project Teams Gathering, which are collaborative events organized by the OpenInfra Foundation. These online events bring together developers, operators, and community members to plan upcoming releases, fix bugs, and discuss project development from a technical standpoint.
Ist OpenStack noch beliebt?
Ja! ✅
Laut OpenInfra-Nutzerumfrage werden weltweit über 55.000.000+ cores operated with OpenStack worldwide. There have been contributions from 10.000+ developers and hundreds of companies. Many Fortune 500 companies run clouds based on OpenStack.
Stuck in the nets of Neutron and OVN?
Read our Blog and attend OpenStack Workshops:
Evaluating OpenStack?
