You will probably have seen from the content of my blog posts that my view of IT over the last Ten or so years ago has been pretty blinkered. Anything within my daily role has been focused on the world of proprietary solutions and what they either do or fail to do. Over the last two or three years this has certainly changed for me, with my focus turning to what Open source projects are capable of providing to an enterprise.
When it comes to Open Source projects, Instead of learning about the intriquete technical nuts and bolts of how they work, I have focussed more on how the API, Interfaces i.e. REST, and the all round general ecosystem, and lastly what value and benefits that opensource provides to a business, and this is where my post dovetails into the relationship between this and my post title.
It appears that with IaaS solutions there is a much larger wave of projects based upon an open sourced architecture. More importantly the open source alternatives to proprietary IaaS cloud (or supposed cloud) options in the start of the Cloud computing era are extremely more ubiquitous and are becoming ever more pervasive with the more uptake of cloud. Cloud Computing is now much more different than previous computing eras when it comes down to available technological choice, which yes for the likes of you and me can become an issue when it comes to governance and control, however this provides IT with much better enablers and at a much better price point.
So is opensource going to be a better selected platform for IaaS strategy or will it still have the same uptake as experienced in Server/Client architectures? Here are some of the reasons why I think it may end up being more dominant in most larger enterprises.
Opensource and open standards have several key benefits for a cloud computing world and here are a few which I think make this statement stand true.
Functionality wise, in order to deliver reliable performant solutions to meet the usually extremely rediculous functional requirements, the technology solution must be able to scale and be readily available to devops. Proprietary solutions can scale, however when you delve into a Proprietary technology architecture it may well be operating on an underlying operating system or service which is proprietary and inherently subject to scalability limits. This underlying dependancy creates a vicious circle of the development cycles being based upon customer volumes, i.e. if companies do not invest in a proprietary product they cannot increase R&D costs.
Open source has always been everywhere and readily available, however I’ve just never known it. What I have certainly realised is that with cloud the use of open source is much much more larger. Look at solutions such as EC2, Openstack and OpenNebula and this is certain enough evidence to suggest that there is a heavy accessible available amount of community and resource to tap into. And of course each of the above aforementioned have a plethora of system integrators or Consultants that have external available skill (at a price) that can work on integration and deployment of such technologies into your existing architecture.
Well this is an obvious one, however don’t think that open source projects such as Openstack or Opennebula are all completely devoid of cost, the open source model has a indirect cost associated with the integration and design of whatever solution you want to implement. This cost however in the financial budgeting world today is much more preferred by projects or programmes that build cloud solutions. Additionally the people factor cost of open source is far more preferred I feel on budget sheets than a lengthy 3-5 Year proprietary software license cost, with this cost being susceptible to additional increases and reductions in level of service OR even death of a product of that all important investment that was set to deliver what you originally asked for.
Open source is much more agile and responsive to architecture change in both functional requirement and in response to evolving changes and trends in the IT industry. With a much larger development community and not the equivalent of how a proprietary SW lifecycle operates, with a shop of paid developers this means that in response to general changes to IT trend the opensource world can much more easily adapt.
In regards to resourcing for support and maintenance of an open source model this was where the open source model in Client/Server architectures failed to succeed in uptake within larger enterprises. With most larger enterprises now outsourcing IT operation, the benefits of this include having a much larger footprint of employed outsourcers who have cheaper opensource skillbase.
By no means am I inferring that proprietary is going to falter in the next 5-10 years, organisations will always want to invest in technology which has the insurance of a solution that is “known” to function to a certain degree and level. Proprietary solutions will feed off of a IaaS ecosystem in some shape or form, and these will provide certainly what an organisation initially wants and also to be able to deploy known entities into an Architecture.
Opensource has the charateristics to be able to meet all of the demands that are imposed by the end consumer of the services whether this be a private enterprise or a public consumer based services. This provides interesting alternatives to the architect when decisions are made upon the roadmap and direction of where they will want to align there future IT strategy towards, and at the end of the day if Opensource fails to deliver there will always be the proprietary alternative to invest in to with a more direct throat to choke.