Servers are the “foundation” of the tech industry. Without servers, there’d be no web server, internet or any central area where files are stored for sharing. Servers make the web spin and store immense amounts of data. They’re pretty ubiquitous to the average web surfer or office worker.
There are many different kinds of servers and this blog isn’t really written to cover the wide variety. There are small servers dedicated to one function, appliance servers performing several functions and there are huge massive servers scaled to many sizes depending on how big the database or file storage needs are. There are “server farms” coupled together in many fashions to increase the computing power and for those applications that scale way past one single massive server. Suffice it to say, there are an infinite amount and variety of servers doing all kinds of things.
Regardless of size and scale, there are several problems that plague any server environment. These are two problems that need a solution every time a new server comes online somewhere in some data center. While there are a plethora of issues that can cause issues with any server, the two major ones are:
- Data backup: backup is the Achilles heel of any server environment due to the potential of failure. Backup means many different things from data being stored to operating system and applications running.
- Disaster Recovery: this is the process of performing a recovery after a failure. While backup is important, if that problem is solved and the data is comfortably stored somewhere waiting for recovery, the next biggest issue is, where and how to restore.
These two areas seem to be one in the same. However, they’re usually addressed separately and in fact, they should be considered in tandem when setting up a server(s). It is paramount that each area be addressed together. Unfortunately, many systems engineers only focus on one specific area, backup only. Recovery is usually an after thought. The reasoning and oversight is vast.
So, what does this have to do with server virtualization? Well, one of the biggest obstacles to these two major problems are “hardware”. Simply put, the easiest approach is to have a duplicate server for each primary server. In fact, in some mission critical needs, this is an option many install. However, the downside is cost. Cost is always a factor with any capitol intensive project.
Why is hardware a factor? Well, most if not all server operating systems are a one-to-one installation on top of any hardware platform. Or traditionally, that’s how it’s been configured. You buy said server and install an operating system. Then, you install the application stuff on top of that. Connect it all and you have an environment that is complex. But, hardware and the system are intricately tied together.
This is where virtualization comes into play. In it’s simplest form, virtualization uncouples the environment and hardware by making them mutually exclusive. In short, building a virtual server removes the dependency on how the hardware is configured. There are other benefits to virtual servers, like scalability and flexibility. However, the quickest win with virtual servers is minimizing the dependency on specific hardware configurations.
There are issues surrounding setting up a virtual server and depending on which solution you go with, the flexibility will vary. Each solution has pros and cons. However, in theory, virtualizing a server really makes for a flexible environment.
In short, it also can solve the two problems mentioned above. With virtual server technology, many of them offer incremental backup options. So, instead of backing up each piece of the puzzle separately, one simply has to backup the virtual server and the data. For recovery, many virtual server technologies allow the environment to be recovered onto any size system and during the recovery process, it doesn’t matter what hardware is in place. In some cases, a large complex system can be recovered to a smaller, less powerful hardware system and run with penalties such as speed and performance, but the system could be run and information accessed, which is the ultimate goal. Accessing the data and information after a failure.
While there’s no way to discuss server virtualization in one basic blog, my intent was to simply put out an very cursory overview on why virtual servers are a great technology and are the NEXT big thing as server environments change and grow. The early adopters, web hosting companies, have already proven that it’s a viable scalable solution. Enterprise and corporate environments are catching up and it’s becoming a standard.
