OpenStack started as a joint venture between NASA and Rackspace. In 2012 the OpenStack Foundation was set up to create a private cloud experience available to everyone as an open source project. Globally there are approximately 3000 developers working on OpenStack code at any time, and the foundation boasts 17,000 members in 140 countries. Many of the developers are ‘gifted’ to the foundation from companies such as RedHat and IBM. Developers come together twice per year in a week-long open forum where they can share ideas, plan future functionality and agree to the way forwards. These ideas are then developed with the aim of including them in the next major software release.
The OpenStack software is actually a collection of many software packages each providing services that work together to create pools of infrastructure resources; compute, network and storage. The pools are made transparently available to end-users as cloud resources. The administrator can then configure individual tenants (or projects) to have their own resource quota sliced from the overall availability within the cloud environment. These quotas are typically shared among multi-tenanted projects accessed through the Horizon web portal, or via the OpenStack APIs.
[easy-tweet tweet=”#OpenStack is a massively scalable #cloud solution. ” hashtags=”Cloud, Data”]
OpenStack is a massively scalable cloud solution. Its greatest advantage is that it is licence-free, giving it the edge over the likes of VMWare. Many businesses around the World have now implemented OpenStack. They range in scale from just one physical node up to the hundreds of thousands of cores, such as that in CERN, where 190,000 cores keep the data analysis going. It powers financial and e-commerce businesses, such as PayPal and eBay. In 2016, Intel announced it was moving virtual machines away from VMWare and onto OpenStack. A commercial drive has also seen OpenStack evolve significantly and what was originally developed to provide private cloud environments has now been deployed successfully by public cloud providers.
Opportunity for new business
The current issue anyone looking to adopt OpenStack faces is that it can be tricky to implement. With so many software packages, developed by different teams, needing to interact with one another, it has been a challenge for the foundation to keep accurate installation guides meaning the implementer is required to be highly skilled and prepared to spend months getting the installation and configuration right. In fact, Suse quoted that “Half of all enterprises that tried to implement an OpenStack cloud have failed.” and once it is running correctly, it needs maintaining (sometimes even fire-fighting). This has provided a growth industry around creating commercially supported solutions. The products range from professional services, managed service support contracts, pay-per-use public solutions and physical OpenStack appliances capable of being integrated into IT estates quickly and easily.
OpenStack is complex and therefore can be extremely difficult to implement correctly. It’s imperative that the core is implemented correctly and therefore the design and delivery of the software should follow standard IT architectural design methodologies. You should seek comfort in the fact that there are thousands of developers supplied by over 500 supporting companies actively working on the maintenance and roadmap and that there are hundreds of OpenStack clouds in production operation, including many in the Fortune 500.