just is better.
Think of meetings, almost always better when shorter, if you could only keep Richard from opening his pie hole with one more
of his redundant
“points of discussion”.
Shorter weekends, however, are no good. You need time for Sunday Brunch and the endless bacon tray at the
As an IT professional, this distinction isn’t always clear-cut — especially when it comes to server refresh cycles. How long should you wait before your next server refresh?
The choice of a server refresh cycle has a tremendous impact on your operations and your budget.
If done too early, you are basically flushing a huge part of your IT budget down the drain. But if you wait too long, you risk downtime of your IT systems and core business functions grinding to a halt.
Previously, typical refresh cycles varied between three years for laptops and desktops and four to five years for storage, servers and network infrastructure components. Such refresh cycles were ideal in situations where information storage and data processing requirements remained roughly the same during the period. However, with the rapid advancement and evolution of new business technologies, server refresh cycles of three to five years no longer makes sense from an IT architecture and technology obsolescence standpoint.
If your organization is still using these rigid hardware refresh cycles, it’s likely that you’ve been struggling to keep up with the latest in tech trends. Business technology advancements no longer follow set cycles. It’s time you look at revamping your hardware refresh cycle so as to allow for more flexibility.
The complexity of most enterprise applications and the volume of big data synonymous with today’s business environment are fast outpacing the capabilities of server hardware — forcing enterprises to consider even shorter refresh cycles.
The development of more sophisticated mobile devices and end-user systems is precipitating the rapid displacement of traditional desktops and laptops. Although smartphones and tablets are cheaper than laptops, they have a higher obsolescence rate. The shorter product lifecycle of these devices coupled with advanced development tools and rapid development processes that enable the faster release of increasingly sophisticated applications are also forcing IT departments to consider shorter refresh cycles.
The conundrum then is — how to maximize the use of already-paid-for server hardware while keeping tabs on increasing application performance, information storage, and data processing requirements. Although enterprises look to reduce operating expenses, they also want to ensure that their continuing use of existing hardware doesn’t negatively impact the execution of core business functions and service delivery to clients.
Of course, it can be tempting to put off server replacement past their due time (for another two to three years, in most cases) due to tight IT budgets and other considerations. However, studies show that server failure rates increase by 85 percent if they are not replaced before the time recommended by manufacturers. This creates a serious risk, especially for mission-critical servers.
At the end of the day, you (and only you) get to decide the server refresh cycle that matches your IT and business needs.
However, you should note that businesses with more frequent server refresh cycles report increased efficiency and adapt more rapidly to today’s shifting technology landscape.
Let’s take a look at three reasons why you might want to consider shorter server refresh cycles.
1. Less Maintenance and Operating Cost
Although the acquisition and deployment of new servers is expensive, maintaining old ones can be even more so. An IDC (International Data Corporation) report shows that the operating costs of a server in its 4th through 6th year is 10x higher than initial acquisition costs.
By purchasing new servers, your organization can save as much as 59 percent of maintenance costs during the first three years of operation. And these savings aren’t limited to cost alone. Maintaining old servers tie up IT’s human resources and expertise. Purchasing new servers enables your IT team to devote more time and resources to innovative projects that can add more value to your business.
2. Improved Performance
Aside from increased maintenance costs, old servers also cost your business in terms of poor performance and increased downtime. The longer they stay around, the worse they become — and maintenance can only do so much. The more up to date your infrastructure is, the more capable it is of supporting the scale and complexity of a changing application landscape.
A shorter refresh cycle typically ensures improved performance — a benefit that is enjoyed not only by your organization but by end users as well. There are fewer technological disruptions and downtime, leading to increased productivity and overall efficiency.
3. Improved Support for Cloud-native and Next-gen Apps
With the hyper-adoption of private, public and hybrid cloud environments, most applications are going cloud-native. Current applications are being supplemented with (and will eventually be supplanted by) next-gen (i.e. cloud-native) applications — applications with very different infrastructure requirements.
Up-to-date infrastructure makes it easier for your IT department to support both kinds of applications. Such infrastructure enables the delivery of public cloud-like economics with private cloud benefits — making it easier to combine the synergies of on-premises infrastructure and public cloud.
Ensuring the Success of Your Server Refresh Cycle
You can no longer rely on the conventional three-to-five–year server upgrade cycles. You need to evaluate the capabilities and specs of your current system, overall IT environment, current and future business needs and the trends and breakthroughs in the technology industry to determine the ideal server refresh cycle for your organization.
Once you’ve decided on such a refresh cycle, it’s best to take proactive steps to ensure the success of your server refresh project. Now, let’s take a look at a few tips for avoiding disaster for your server refresh project.
1. Non-obvious Dependencies
Ever heard of an application dependency map?
When doing a server refresh, it is extremely important to understand dependencies at all levels. This could range from simple things such as multiple redundant power supplies connected to the same UPS to several applications that rely on a single database (where your staff thinks it’s only connected to one).
Taking down the UPS before moving the power supplies to another power source may cause your servers to fail. Also, removing the database along with the application will cause other dependent applications to fail. You should come up with a comprehensive application dependency map to ensure the success of your server refresh cycle.
2. Measure Twice, Order Once
You may end up exceeding your IT budget if you order new equipment (especially large ones) without first checking if it is physically possible to move the new hardware into the reserved space. You must ensure that the equipment can clear doorways, hallways, narrow openings, and tight corners and that your building’s elevators can handle its weight.
Where realistically possible, it’s a good idea to order samples ahead of time to check these factors. Also ensure that your office space has adequate power density, cooling capacity, floor strength and sufficient ceiling height for all new equipment. Overlooking these metrics will adversely impact your server refresh project. It’s no good ordering equipment that you end up returning (since it doesn’t even fit in the building) or under-utilizing due to lack of supporting facilities.
3. Avoid Being the “Guinea Pig” for New Tech
Never go all in with a new technology that has just been pushed into the market. Unsuspecting enterprises that are seduced by hardware vendors and software suppliers into purchasing the “newest architecture” often end up beta testing a product with a lot of bugs and design flaws that need straightening out. You should only pick products that fit in with your business needs, instead of migrating to new technology based on industry buzz and marketing material. If you want to try out such new architectures, it’s best to do a partial, trial migration to see how things go.
4. Don’t Assume — Do an Audit
Before undertaking a refresh cycle, you should have a clear understanding of the specs of your current hardware. Although regular IT audits have become common practice, you shouldn’t assume that your organization’s last audit is fully up to date. Take time to thoroughly catalog the memory, I/O, bus speeds, and processor power of the server hardware.
A truly comprehensive audit can help reduce budgeting requirements by helping you identify some of the old hardware that can still meet business needs. It also provides details about everything you have in your data center including the number of physical and virtual workloads and how many of them use local or shared storage.
These details will provide you with an idea of the overall costs of the refresh (in terms of new infrastructure purchases and costs) by identifying much-needed additional server capabilities.
5. Understand Your Peculiar Needs
Before undertaking a server refresh, a clear understanding of your business needs is essential. Never do a server refresh just because “it’s time.” Spell out the gains to be had by the server refresh. It could be better cost-per-performance economics, support for more advanced applications, or improved reliability and availability.
There must be a compelling IT and business reason for undertaking a major server refresh. If there’s none, and your current servers still provide dependable service, you should re-evaluate your server refresh cycle.
6. Reevaluate Your Choices
Before going through with a server refresh, it’s a good idea to re-evaluate your server environment choices. Do you need to outsource, maintain your own data center or split them between the cloud and on-premises? These choices will impact how your IT department plans and implements a refresh cycle.
Usually, it’s cheaper and easier to gain access to the newest servers through your cloud partner. However, owning your own hardware may help you amortize costs over time, especially if you decide on a longer-than-typical server refresh cycle.
7. Avoid Broken Links
Data migration is an essential part of every server refresh. However, inadequate planning and improper execution of this phase could lead to broken links — a scary byproduct that upsets users and makes bosses frustrated with IT staff.
The most effective way to ensure that all your links survive the migration process is to use LinkTek’s LinkFixer Advanced. LinkFixer Advanced facilitates stress-free data migration by automatically monitoring and fixing links in huge batches, thus protecting your file links during a server refresh.
Ensuring the success of your server refresh cycle will involve the development (and subsequent implementation) of a detailed rollout plan. This plan must include a training program for employees so that they are ready to use the new environment once the refresh project is completed. If you wait till the conclusion of your refresh cycle before equipping employees with the tools and training needed to properly utilize the new hardware, you’ve waited too long.
Implementing a server refresh cycle that takes into account your business needs, and the gains to be gotten from the very latest IT trends (while avoiding the pitfalls detailed above) will enable your business to deliver continuous and predictable products and services to consumers.
Such a server refresh cycle will provide you with an agile, elastic, and scalable IT infrastructure that will ensure a more superior service delivery model for your business.