Infrastructure as a Service (IaaS) or cloud hosting is one of the three primary types of Cloud computing, along with PaaS and SaaS. Many small web hosting companies now offer Virtual Private Servers (VPS) as Cloud servers and masquerading as “Cloud hosting,” we thought it would be helpful to clear the air:
So what is a Virtual Private Server?
A VPS is a traditional single server operating in a multi-tenant hypervisor-controlled environment. Customers will typically order the VPS with fixed attributes (CPU, RAM, Storage) in a package and pay a fixed price. Other customers do the same on this dedicated server, and the provider must make decisions about density – most VPS servers are over-subscribed. This is why they are so inexpensive and prone to so much variability in performance. VPS servers are also not fully-networkable – they don’t offer multiple ports and the ability to connect them to private networks. Most traditional VPS offerings are billed by the month and have pre-sized configurations.
What is Cloud Computing?
Pure Cloud computing IaaS offers resources from a group or pool of technology resources – CPU cores, RAM, Storage and network ports. IaaS users are free to create servers of any size (if their provider does not limit them to pre-set server or instance sizes). IaaS users are then able to connect storage devices, firewalls, load balancers and networks (public and private) together to create a true virtual data center. Cloud computing providers then add additional features to differentiate themselves, like InfiniBand networking technology that connects servers to storage and servers to servers at 80Gb/s or “Live Vertical Scaling” which is where IaaS users can add more RAM or more CPU cores to a server without restarting a server. IaaS provides far more flexibility, power and isolation that any VPS platform.
Taking it a step further, Cloud computing IaaS should be thought of as a utility, just like the local electric company, where cloud service providers offer the resources needed to configure virtual servers or instances that do not have fixed computing attributes. The computing configuration of the cloud server can grow or shrink dynamically depending on the resources requested by the specific application or user. Billing for the resources are based on actual consumption over a specific time period for CPU cores, memory, cloud storage, and bandwidth, rather than a preset package that many first generation cloud providers and VPS providers offer. Read more about Cloud Computing in the Cloud Terminology.
|Instant provisioning with graphical data center designer||Yes||Varies||Varies|
|Fully-flexible server sizing - Choose CPU, memory and disk independently||Yes||No||No|
|Guaranteed, dedicated CPU Cores and RAM - No resource sharing!||Yes||No||No|
|On-demand live vertical scaling||Yes||No||No|
|Flexible private network with up-to 8 ports/server||Yes||No||No|
|High-speed (80Gbit/s), high-performance, state-of-the-art InfiniBand mesh network||Yes||No||No|
|Run any x86 operating system||Yes||No||No|
|Clean, simple API||Yes||No||No|
|Simple, easy to understand, pay-as-you-go utility pricing||Yes||Varies||Varies|
|Friendly, senior-level phone and e-mail support as standard||Yes||Varies||Varies|
When choosing between VPS servers and Cloud Servers, it’s most important to select the server that best meets your specific needs and requirements. There is no doubt that more and more Cloud servers are replacing VPS servers. The lack of VPS scalability, combined with its overall lack of flexibility and the latest pricing of Cloud computing services simply makes Cloud a better choice.