I call CloudSigma a pure Infrastructure-as-a-Service (IaaS) company but what exactly do I mean by that? Well, I feel our platform is much closer to the true understanding of what IaaS should actually offer its customers in relation to control, flexibility and the direct relationship between usage and cost.
I’m aiming to set out in this post how we feel we are, in many ways, fundamentally different in the way we approach IaaS and how this can benefit the end-users of our platform.
Over the past year or so there has been much debate about what constitutes a cloud service, what is the appropriate definition to use etc. etc. This debate has in my opinion overshadowed the more important questions so I don’t propose to go over old ground. Suffice to say that there are three basic layers to the cloud computing world, Infrastructure-as-a-service (IaaS), Platform-as-a-Service (PaaS) and Software-as-a-Service (PaaS). You can read more in the cloud computing section of our site.
One of the main reasons I co-founded CloudSigma was because I couldn’t actually find an IaaS provider on the market. Yes at the time Amazon, GoGrid and others had started to roll out their services but when I went to use them, instead of just infrastructure being provided, apparently they wanted to control much of how I could use my software. Now I didn’t find anything about making restrictions on software usage in any of the definitions or discussions about cloud computing and IaaS. That is why we developed the Cloudsigma platform, to delivery true IaaS without those restrictions.
Pure IaaS should give the user full control of the software layer, allowing them to run any operating system and software that they desire. Networking and other services such as load balancing and scaling should also be user controlled. For example the idea that one form of load balancing is applicable to all users of a public cloud just isn’t correct. Users should be able to tailor and tune in a solution that directly addresses their needs in the same way that they can when using dedicated hardware. No-one should change the way they work to fit some arbitrary architecture decisions of a cloud provider. IaaS should deliver the flexibility, scalability and transparency of the cloud without introducing new restrictions and ways of working.
As cloud computing moves out of the hype phase there is a lot of debate regarding issues and concerns regarding using the cloud. Many of these are actually not problems inherent with cloud computing but with the implementations of it delivered to date. This is particularly the case with regards to IaaS. The cloud can and should delivery the same levels of control, data portability and security that traditional dedicated solutions do.
So what should an IaaS cloud look like? In our opinion it should incorporate at a minimum the following concepts:
- Use of open standards
- Participation in interoperability schemes
- A clear exit path for users to export all their data in a convenient and inexpensive way
- Maximum control for the user to configure their computing resources as they see fit
- Transparent billing and pricing based on actual usage
- Open networking
- No restrictions on operating systems and software (beyond base hardware architecture requirements)
- Clarity regarding how data is stored (storage solution used, redundancy etc.)
- Where data is stored including any backups and other copies
Incorporation of the above principals would go a long way to addressing the key concerns raised by prospective and current IaaS users.
What would be the real tangible benefits of embracing these principals and characteristics? A good example of the real affect is how our Software-as-a-Service (SaaS) client using our cloud can benefit. Firstly, they install custom server images using the operating system of their choice. Often these are customised distributions or at least ones optimised for their particular usage. Secondly, they are able to purchase resources in the quantities they actually require, without getting bundled resources that they don’t need as they scale up. Thirdly, they are able to optimise their load balancing and networking to work in the optimal way for their particular application. In many cases different SaaS providers are implementing very different ways of scaling and load balancing as a reflection of their differing products. Finally, if they wish to migrate cloud, they can do so quickly, easily and inexpensively by simply FTPing out their drive images securely.
The end result is that customers are buying only what they need and, because their software layer and related services are completely customised to their particular usage, they end up having to purchase a lot less computing resources to get the same performance from their cloud solution.
That’s the real value that IaaS can delivery users; reduced waste, increased efficiency and costs that scale directly in relation to their computing needs. We believe our platform goes a long way towards pushing IaaS towards those realities and we are constantly finding new ways to make the user experience better. If you have suggestions for our platform please contact us, we welcome an open and frank discussion with our user base.
- Manage Docker resources with Cgroups - May 12, 2015
- Docker, Cgroups & More from ApacheCon 2015 - April 30, 2015
- How to setup & optimise MongoDB on public cloud servers - March 24, 2015
- Presentation deck from CloudExpo Europe - March 17, 2015
- CoreOS is now available on CloudSigma! - March 10, 2015