How to buy cloud hosting

October 27, 2014 by iWeb Technologies in: Hosting Essentials

As our lives become increasingly intertwined with the internet and mobile technology, cloud services are becoming an important component of more and more commercial ventures. For business leaders, there is more of a need than ever before to understand and evaluate the suitability of the different cloud hosting options for a given project, service or application.

This article draws on new standardized definitions for cloud terminology published by the International Standards Organization (ISO). With the help of these definitions, and the business benefits explained in this article, you’ll be able to approach the decision of selecting a cloud hosting model with confidence.

Cloud for Business

Cloud Service Categories: IaaS, PaaS & SaaS

Cloud hosting often is synonymous with cloud Infrastructure as a Service (IaaS), sometimes referred to simply as ‘Cloud Infrastructure’. However, it can also be extended to include cloud Platform as a Service (PaaS), or simply ‘Cloud Platforms’. When considering cloud hosting, it’s therefore important to understand the differences between IaaS and PaaS, two of the three cloud services categories.

The ISO identify and define three cloud service categories :

  • Infrastructure as a Service (IaaS):  A cloud service category … in which the cloud service customer can provision and use processing, storage or networking resources.

  • Platform as a Service (PaaS): A cloud service category … in which the cloud service customer can deploy, manage and run customer-created or customer-acquired applications using one or more programming languages and one or more execution environments supported by the cloud service provider.

  • Software as a Service (SaaS): A cloud service category … in which the cloud service customer can use the cloud service provider’s applications.

IaaS is close to a more ‘traditional’ hosting model. The customer has access to the underlying computing and storage resources, operating system, web server, application server, database server and other configurations. This level of access provides all the control you could possibly need, but can also be more than you are able manage.

PaaS (for example Google App Engine, Amazon Elastic BeanStalk, Salesforce Heroku) is designed to make life easier for developers. As an extension of IaaS, PaaS removes server configuration  from cloud hosting, allowing the customer to simply get on with coding, often within a framework designed to further speed up this task. PaaS is a little like a self-service version of a ‘fully-managed hosting’ service, where all configuration and update management is done for you, plus a development framework added on top.

The tradeoff for the convenience of PaaS is a lower level of control, limited development language options, limited software options and, most importantly, vendor lock in. In PaaS your applications are designed to run on a specific platform. So once you’re on a platform you have to stay there, employing developers who are familiar with the platform and the programming languages it uses.

Completing the three cloud service categories, SaaS is the provision of a finished application over the internet. This service is hosted and developed on cloud services, and offers a product with features, functions and benefits designed for a specific market.

PaaS is great for those who have the skills to develop an app, but not the time, skills or desire to handpick, install, configure and install OS and server applications. IaaS may be preferred by those with system administrators in their IT departments, or developers who are comfortable configuring servers, as is often the case with website development. PaaS complements the scalable, elastic and convenient nature of cloud hosting. IaaS offers far more control, programming options and mobility.

From here on in, we’ll focus on IaaS as cloud hosting. For some more information on PaaS, see this article on why PaaS can be good for developers.

Cloud Computing: Cloud vs Dedicated

The ISO definition for Cloud computing is “a paradigm for enabling network access to a scalable and elastic pool of shareable physical or virtual resources with self-service provisioning and administration on-demand.” Let’s break that definition down further, and consider the benefits.

The key attribute for cloud computing, the one that differentiates it from dedicated hosting, is the way that resources are allocated and provided to the customer.

In dedicated computing, physical servers are configured and deployed one at a time, and owned by the customer, who will usually pay a fixed monthly price until they upgrade, close or migrate servers.

With cloud computing, all customers share a pool of underlying physical resources, managed by the cloud service provider. The customer can deploy Virtual Machines (VM) – the cloud hosting equivalent of a physical computer – that use an allocation of the resources in the pool. The computing resources of these VM are very easy for the customer to reconfigure (“scaling up and down”), and new VM can be added instantly and easily (“scaling out”), all of which is far easier than physically deploying dedicated servers and upgrading pieces of hardware. Along with the flexibility of scaling your infrastructure, cloud computing also offers the ability to purchase on a pay-per-use basis: by the hour, for the amount of resources you use that hour.

If the benefits of cloud hosting are flexibility to scale up and out instantly (even automatically) according to the resources needed, and to pay only for what is used, what are the relative merits of dedicated hosting?

Some applications do not need the flexibility offered by cloud computing. And many will benefit from the inherent stability of a dedicated server. A website that sees consistent traffic levels does not need to scale up and down quickly or erratically. A large and busy database may achieve more consistent read and write speeds on a dedicated server.

Benchmarking by Internap has shown that CPU intensive operations like file compression, audio encoding and video encoding performed 2-4x faster on dedicated servers than on cloud servers with similar specifications. In other tests, dedicated hardware outperformed similar cloud servers in disk performance stability 4 times out of 5, outperformed disk speed by 500%, improved RAM performance by 150% and improved database performance by 65%.

If you have a website with relatively consistent and high levels of traffic and transaction, resource-intensive processing applications or databases that processes a lot of transactions and store a lot of data, dedicated servers may provide better price-performance value than cloud computing. For more information on buying dedicated hosting, read our concise guide on how to buy dedicated hosting.

The price-performance ratio of cloud and dedicated hosting will depend on how well suited your application is to the benefits offered by cloud hosting – scalability, elasticity and pay-per-use – and those of dedicated hosting – high performance stability and fixed monthly costs. The more computing and storage resources needed, the more expensive the cloud pricing model usually is relative to the dedicated hosting fixed-cost model.

Cloud Deployment Models: Public, Private & Hybrid

The cloud hosting described up until have been examples of public cloud hosting. The ISO definition of public cloud is:

Public cloud: “Cloud deployment model where cloud services are potentially available to any cloud service customer and resources are controlled by the cloud service provider. A public cloud may be owned, managed, and operated by a business, academic, or government organization, or some combination of them. It exists on the premises of the cloud service provider. Actual availability for specific cloud service customers may be subject to jurisdictional regulations. Public clouds have very broad boundaries, where cloud service customer access to public cloud services has few, if any, restrictions.”

Put simply, a public cloud is cloud IaaS managed by a cloud service provider (examples: iWeb, Internap, Rackspace, Amazon AWS, Google Cloud Platform) and available to pretty much anyone, usually with the features described above, and sometimes with a PaaS option.

A private cloud uses the same cloud computing techniques and offers the same features as the public cloud, except that the underlying pool of computing and storage resources is controlled and exclusively available to the customer. This infrastructure may be hosted by the customer (on-premises), or hosted by a hosting provider (off-premises). The latter is sometimes referred to as a ‘hosted private cloud’.

The ISO definition of private cloud is:

Private cloud: “Cloud deployment model where cloud services are used exclusively by a single cloud service customer and resources are controlled by that cloud service customer. A private cloud may be owned, managed, and operated by the organization itself or a third party and may exist on premises or off premises. The cloud service customer may also authorize access to other parties for its benefit. Private clouds seek to set a narrowly controlled boundary around the private cloud based on limiting the customers to a single organization.”

Private cloud provides additional control, ‘privacy’ and an alternative cost model. With private cloud, the customer pays a fixed price for the underlying infrastructure plus any software licensing. For customers who value the flexible and responsive features of cloud hosting and the fixed-cost model of dedicated hosting, this is an interesting choice. It also offers the opportunity for large organizations to allocate costs within their organizations by charging It resources to internal customers on a pay-per-use model.

Another big benefit of private cloud is the ability to apply the additional control and benefits of cloud computing to existing data center operations. As the manager of the private cloud, a customer has access to a global view of their infrastructure. This allows the customer to monitor their infrastructure, apply templating, and automate operations – things like automatically scaling VMs in response to an applications needs.

Public clouds offer a relatively simple way to access the scalability and pay-per-use  benefits of cloud computing. Private clouds offer a compromise between public cloud scalability and private infrastructure control and fixed costs. Private cloud is a more involved choice, and often a ‘bigger solution’ but offers many additional operational benefits when compared to more traditional dedicated hosting.

Multi-Tenancy: Risks & Perceptions

When considering public cloud, many potential customers are turned off by the negative perception of multi-tenancy. The single tenancy of a private cloud, by contrast, is one of it’s differentiating benefits. Here’s the ISO definition of multi-tenancy:

Multi-tenancy: “Allocation of physical or virtual resources such that multiple tenants and their computations and data are isolated from and inaccessible to one another.”

We have already seen benchmarking studies that show a lower level of performance and performance stability in cloud hosting when compared to the equivalent specifications of dedicated server. In addition to these performance concerns, there is also a lingering concern of the security of data in a shared environment, despite the isolation if tenants from one another in public clouds.

In a 2013 survey of our hosting customers, iWeb identified the following list concerns amongst existing and potential cloud hosting users, the four most important of which all relate to the real or perceived effect of multi-tenancy and sharing resources.

concerns about cloud-based hosting 2013

One of the main areas of focus for cloud hosting providers is to overcome the performance and security concerns that are evident in today’s market. Attempts to combat security and stability concerns have taken the form of SSD-Only public clouds like that offered by iWeb, and “bare-metal cloud servers”, first released by Internap and now available from a handful of providers.

Bare-metal cloud servers are dedicated servers that are deployed and billed as a cloud server would be, giving some of the flexibility of cloud, with the performance and privacy of a dedicated server, and some of the ease of scaling.

Need more information?

For more information and assistance choosing a cloud hosting option, speak to a sales advisor or solutions architect by contacting iWeb. You can also find out more about some of our specific cloud options and some of the resources mentioned in this article by following the links below.

Comments

No comments yet.