Need to get 6 more months out of a server?

Need to eek out just a little more time out of an old server?  Don’t have the time or resources to reconfigure the rest of your server roles?

At ITP we use a neat trick.  Vmware Converter.  What is Vmware Converter?  (http://www.vmware.com/products/converter/)It is a free technology which captures a physical server and backs it up to a couple of files which can be moved to any server or computer and with the use of a free or inexpensive “player” you can get that old server configuration running on reliable hardware.  This is a great way to get a few more months out of hardware it is also an excellent disaster recovery option for small to large enterprises.  Even if you have never used virtualization this tool can be a great introduction to the technology and a real business saving tool in the case of hardware on the edge.

When is it the right time to replace a server?

We at the office have had a rash of rushed server replacements in the last few months.  We ourselves are a business who struggle with when it is time to make capital investments like servers and desktops.  I just wanted to take the time to pass along some facts about hardware.  I am an engineer here at ITP not the sales guy sorry Joe.

When is the right time to replace a server?

The answer which no one likes to hear is:  it depends.  Like your car you can drive it until it strands you and then get it fixed and get back out on the highway again.  Others chose to replace their car or have schedule maintenance on it in order to avoid this down time in their life.  The important thing is to realize the difference between a tire blow out which can just be replaced and fixed and a general wearing out.  The other car analogy I will throw at you about servers is that your equipment has miles on it.  In the industry it is called a MTBF or Mean Time Before Failure or failure rates.  It is possible depending on your environment that you have put 100,000 miles on your server in 2 years.  Think of servers more like you would your copiers after so many copies it just seems to start jamming more, right?  Microsoft.com tells us the research they have done says this is around the 5th year:

The percentage of servers experiencing some form of component failure jumps from 10 percent in year four to 50 percent in year five, according to James Browning, a research vice president at IT analyst firm Gartner Inc.

Microsoft and we at ITP urge you to use a combination of factors the most important of which is the performance of the system and it’s age in combination.  I say use both because if you have a 7 year old server even if you feel it’s performing well the risk of complete failure is so high that its not worth the risk.  The other factor I consider is manufacture, not that it be Dell or HP or IBM but if it was a “white box” or built server or PC.  Part of the process of becoming a Dell or HP server system is intensive research and burn in process which ensures longer more consistently running hardware.  If the server we are reviewing was built we tend to assign another year to it’s life at least.  ITP has never built PC’s or Servers because we think it is not good business and we would question any company out their still building PC’s.

Risk is really what its all about right?  If you replace your servers every 3-4 years on schedule and keep your servers at a cool consistent temperature you will have less issues.  Meaning less IT costs, less intermittent and major downtime.  I often wonder all the clients who call us when 1 desktop is down because of a virus or MS Office needs to be reinstalled or a corrupt OST file if they realize the effect on their business for email to be down for an entire day or to have no access to your calendar or contacts.  You say that will never happen but risk management methodology says we must think about it.  For almost all our clients a day of downtime would cost the firm in lost time and money more than the cost of a new server.

Budget for it, the core of the network should not be taken lightly.  Make sure you have a solid disaster recovery plan meaning backup, policy and plan to get the business back up and running.  Hold off on the new printer or the new laptops but make sure even in this economy you don’t take risks on the machines that carry the core of the business.