- IT Support
- About Us
Servers represent a sizeable investment for many businesses so it’s no surprise that server life expectancy is an area of concern for many organisations. Given the sums of money involved, the last thing you want to do is find yourself shopping for a replacement six months down the line. There’s truth in the old adage ‘you get what you pay for’ and the key to server longevity is to allocate sufficient budget to ensure you can procure a server with the necessary feature set to support your business needs both now and in the future.
Most entry level servers offer little room for future expandability. In order to hit a low price point, manufacturers have to forego the extra processors, drive bays, PCI and memory slots typically included in mid-point and high end servers. Whilst this might be fine in the short term, if your business is growing rapidly, you’ll soon find you reach a point where you can’t upgrade your server to keep pace with your company’s expansion. Technology marches on at a relentless pace too and it may be that the new application or software upgrade you’re planning to install demands more RAM, processing power or disk space than you can physically install. Once your server becomes obsolete, your only option is to upgrade to a server offering more power and capacity. Resist the appeal of a low ticket price and be sure to factor you mid to long term requirements when choosing a server.
Quality and redundancy
It’s not just server capacity you need to be mindful of, the quality of the installed components and the amount of built in redundancy can both have a bearing on server life. Branded servers like HP ProLiant and Dell PowerEdge dominate the market for a reason. There are cheaper brands out there but the premium price buys you superior quality and reliability, qualities you wouldn’t want to skimp on, especially if your server is mission critical.
Redundancy is your insurance policy, keeping your server operational and your business productive when a hardware component fails. If you can ill afford any server downtime, a redundant server solution should be high on your agenda. Lavish more money on your server and you’ll usually get better components and more redundancy thrown in which means less downtime, more productivity and therefore a longer useful lifespan.
Horses for courses
A server’s life expectancy is directly related to the nature of the workload it is tasked with. Much like a car, a server will incur wear and tear, becoming less efficient and costlier to run before eventually breaking down, more so the older it gets and the more intensively you use it. Most servers are designed with a 3-5 year lifespan in mind. Task your server with something that isn’t particularly taxing (e.g. use it as a DNS server) and you’ll probably get many more years of good service out of it, particularly if the nature of the work doesn’t change and/or the number of supported users is small. Charge your server with something a bit more mission critical however and you’ll realistically be looking to replace it as it gets to four or five years old.
When is it time to replace your server?
Stick or twist? You need to balance the various costs (power, admin, cooling etc.) involved with keeping the server versus ditching it in favour of a newer machine. Weigh up the potential productivity gains too; how much more efficiently and effectively can a new server deliver the required workloads? Breaking this down a bit further, here are some key areas to look at:
Warranty cover – Your server will come with a time limited warranty provided by the manufacturer. Many companies choose to replace their server once this warranty expires as it’s a costly exercise to extend it and it’s often not worth the risk of incurring any downtime due to a potential server failure.
Availability of parts – The older your server gets, the harder it will become to procure spare parts and therefore the more expensive such components will be. If you can’t replace a failed drive or power supply, your server loses productivity. The risk alone may warrant switching the server to a less critical application and replacing it with a more current alternative.
Software support – It’s not just about the hardware, servers are employed to run software after all. Is the server capable of being upgraded to a new operating system? Can you install the latest new application your business needs to run? Are software vendors still supporting the applications you have installed with ongoing updates and patches?
Cost-performance ratio – The power savings alone might provide sufficient justification for retiring an ageing server, not to mention the performance gains from switching to newer technologies. If it’s taking too long to get things done, too much time is being spent on server maintenance or the server is burning too much electricity perhaps it’s served its useful life?
Server administration – Would you need to rebuild the server in the event of a catastrophic failure? If so, do you have staff on hand familiar with its setup, unique workarounds, security settings, login scripts and so on? Over time, your staff, standards and systems change and an older server might prove increasingly troublesome to administer due to its lack of conformity with the surrounding infrastructure. Once a server becomes more trouble than it’s worth to maintain, it’s time to decommission or replace it.
To coin another adage, nothing lasts forever and eventually there’ll come a time when every server has served its useful life. Hopefully this article can help you identify when you’ve reached that point allowing you to take proactive rather than reactive action.