Technically Speaking

Organizations large and small are trying to make the most of any investment dollars they spend to improve infrastructure and reliability. It seems that more with less is the most common statement in IT shops today and with good reason. The recession is causing many employers of all shapes and sizes to consider any ways possible to save money.

Changing technology

Even the rather dismal economy doesnt slow down the changing technologies that move through corporate IT departments. From Netbooks to Windows 7 and the latest social media sites, there is and always will be some new and unbelievable technology to deal with, making IT employees giddy with opportunity. These changes can continue to be a good thing by allowing all employees to continue learning things that can and will set them apart from others.

In the recent past, companies were taken down the application server path. This led oraganizations and their employees to consider purchasing a separate server for each application needed in the organization. This is an effective way to ensure each application running within your organization has the needed firepower to operate at its best. In fact many shops still operate this way, and if the equipment exists and it works as needed, that is ok.

More servers, less money… long term

The organization I work for has a pile of servers. We have servers for each application that we have running and one for purely back office/IT operations. Keeps things nice and separated and for the most part everything is humming along nicely. We also have an edict to use or purchase less servers because we should not need that many servers to function. The number of servers we run should certainly not need to go up in the short term. If anything the way to move forward would allow for some initial investment to obtain a couple of rather high end servers with loads of processor and memory as well as a SAN to store our data on.

Implementing these hardware items and a virtualization platform like Hyper-V or VM Ware would allow us to create multiple machines on less hardware. Which in reality would provide almost an infinite number of servers while allowing the amount of physical hardware stays the same.

There will be an initial outlay for hardware and software because many organizations will not have enough horsepower in only a few of their server systems to ensure all of their applications (with the exception of a couple) can be moved to virtual servers. After this initial investment, when the existing servers begin migrating to the virtual environment and their physical counterparts are repurposed or taken off line, the organization should see a reduction in power consumption resulting in a smaller carbon footprint and lower operating costs.

The idea is that less physical servers running require less power. This will get an organization moving down the path to being green, or at least greener which is good for the environment and remains a very popular buzz word.

Being virtual saves power… but how does that help?

When you need a server for testing or a new application, you can create a clone of an existing server or build a brand new virtual machine within your environment. This will provide as many servers as necessary in very few clicks. The VM boxes will use a portion of the available hard disk space and memory within the physical server on which the machine is running which will limit the number of virtual machines that can be used and active at one time only by the physical server constraints in the environment.

For an IT manager, there will be discussion and justification needed to implement the project and initially go virtual, but over time the discussion needed to bring a server online will be more about usefulness than justification of hardware purchase. When the IT department can bring a server online in just a few clicks for testing or production use the overall cost of IT projects should go down, both within the organization and due to the purchasing of additional hardware.

Recommendation: There are certain server systems that are best left to their own servers in production environments. These include Exchange (up to version 2003) and Active Directory Domain controllers acting as global catalog servers. It seems to be beneficial to run these boxes as physical servers, but as advancements in both technologies continue to role out and as virtualization improves it stands to reason that these too will become exceptional candidates for virtualization. Besides keeping a few servers around to help hold down the rack isnt a bad idea, as long as they work.

Hopefully this post has provided some food for thought about ways to make wise investments with your IT dollars and reduce costs in the long run.