There are, however, very important differences in what colocation, managed hosting and the public cloud each provides, which are often a source of confusion.
Colocation: a simple, cost-effective option. Colocation vendors essentially rent space in the form of racks and/or cabinets that are physically secure and provide what could loosely be termed the utilities necessary to sustain servers. These include power, cooling, cabling and, typically, a dedicated Internet connection with a pre-determined bandwidth allotment – often referred to as “ping, port, and pipe.” IT departments are responsible for buying their own servers, installing them, maintaining them and ultimately disposing of them at the end of their life cycle.
The advantage of colocation is that companies don’t have to deal with the complexities of housing, powering and cooling their servers. The primary drawback is that they are still responsible if anything goes wrong. This means that if something goes wrong, e.g., a server crash, someone has to drive over to the colocation site and fix the problem. (Some colocation vendors will supply so-called “remote hands” at an extra charge to perform routine functions such as re-booting when necessary.)
Managed hosting: for greater focus on business problems. Managed hosting is a much more comprehensive offering. To begin with, managed hosting vendors provide all the physical hardware, which means that a significant portion of a company’s total IT spend can be converted from CAPEX to OPEX – a significant benefit and often the primary driver for companies that choose managed hosting. Companies not only avoid initial capital expenditures. They also eliminate the expenses of hardware replacement over time and the headaches associated with accurate capacity planning. Growing companies, in particular, avoid the hidden cost of having to buy CPU and storage capacity that may sit unused for months until the company grows into it.
A second economic benefit of managed hosting is that, in addition to power, cooling and a bandwidth allocation, companies receive a range of important services. These include monitoring of server performance, management of the operating system(s), OS updates, patching and back-up. (Some managed hosting vendors move higher up the stack to offer application support as well.)
As any IT manager knows, routine tasks like OS maintenance are costly, because they consume a huge number of work hours. Each task that is off-loaded frees up some of those hours so they can be devoted to tasks that will more directly impact the business. For example, IT can train and develop experts on specific applications for the various departments (marketing, manufacturing, finance, etc.). This enables IT to help maximize the business value of these applications, rather than simply keep them up and running.
With managed hosting, if a system goes down, the IT department can focus on recovery and minimizing user downtime, instead of having to worry about infrastructure downtime.
The major drawback for most companies considering managed hosting is lack of control. For some, control is a non-issue. But for others, the fact that their applications are running on virtual machines that may hop from one physical server to another on a moment’s notice and potentially compete for resources poses problems. For example, regulatory compliance in some industries demands that companies know where their data physically resides.
Cloud Computing: Taking Managed Hosting to the Next Level
While there are many issues related to cloud computing – enough to fill a book if not several, three deserve special focus for those weighing the cloud vs. managed hosting.
Firstly, on the plus side, cloud computing isn’t only a technical approach to providing IT resources. It’s also a business model. In this regard, it can be seen as managed hosting taken to the next level. Cloud customers enjoy all the CAPEX vs. OPEX benefits plus the financial benefits of usage-based pricing.
With cloud computing, companies can very quickly scale up capacity to meet peak needs such as end-of-month financial processing, and then scale back down after the need is over – literally in a matter of seconds. They can therefore avoid paying for unused capacity. This same quick scale-up/scale-down, often combined with self-service provisioning, is the most cost-effective approach to IT development projects.
Secondly, on the technical side, the cloud can sometimes offer substantial performance improvements.
Third, and on the downside, migrating to the cloud is not as straight-forward as the hype would indicate. Most applications require at least some coding modifications to run in a cloud. Some cannot take full advantage of the cloud’s benefits. Some cannot be migrated at all. In short, the cloud is most definitely a look-before-you-leap proposition, and small or mid-sized companies should make sure their cloud vendor can provide adequate technical support for a smooth migration.
The New Hybrid Reality
For many companies, the choice between these three options will not be either/or. Today, most companies have a hybrid environment where some computing services remain on-site, some reside outside the company, and some (in the case of cloud-bursting) actually fluctuate between inside and outside depending on demand. It is extremely important that any vendor chosen for any of these options have a deep understand the complexities of hybrid IT, and be able to support migrations when needed.
Increasingly, computing power is more and more often becoming a service companies buy rather than one they provide for themselves, and this trend will surely gain even more momentum over time. There are many options now available, and they are all worth exploring.
For more information, read the whitepaper
Related Business Solution: Colocation Services