All IT administrators have to eventually deal with aging equipment: A storage system purchased several years ago is now short of capacity, a key network segment doesn't have enough bandwidth for a growing user base, or a server can't host enough new virtual machines to suit today's needs.
In these situations, administrators can either upgrade existing equipment or purchase new systems. Here are some things to consider when choosing between upgrades and purchases.
Upgrades bring limited benefits
In many cases, an upgrade can resolve a system issue.
Adding storage capacity is often a matter of adding more disks to the cabinet or systematically replacing smaller disks with larger disk models. Virtualization makes it easy to migrate
Similarly, some server lifecycles can be extended by adding CPUs to unfilled sockets or by adding memory to available slots. Again, virtualization can migrate all of the workloads off of the server so that applications remain available while the work is performed.
Upgrades are often faster and easier to deploy than new systems, however there are many compelling arguments for purchases.
For starters, an administrator may have no other choice. The system may not be upgradeable or -- based on testing and discussions with the manufacturer -- an upgrade won't yield the desired capacity and performance improvements.
IT products have a finite lifecycle, and a technology refresh may be necessary when a manufacturer discontinues parts and support for aging systems. For example, a vendor chooses not to renew a service contract after the current contract expires. It could be possible to engage third-party support using parts secured from the secondary market, but this is a deeply flawed and problematic approach that can result in extended downtime and unplanned expenditures.
New technology brings new capabilities
Regardless if purchases are done out of necessity or out of choice, there are several benefits to adding new technologies to the data center.
A new system could provide additional resilience through redundant components such as dual power supplies or network adapters. Improved availability and the underlying reliability of a new system translate into greater resilience (less downtime) for the corresponding workloads.
In addition, a technology refresh brings new features and capabilities that can enhance a business.
For example, a new server slated for service as a virtual host will likely include the latest processors with virtualization-assist features (Intel VT or AMD-V) designed to boost server performance with virtual workloads. As result, the server can host many more virtual machines than the previous server, and this benefits the business.
Similarly, an aging storage system with a 2 GB Fibre Channel (FC) interface may be replaced with a 4 -- or even an 8 -- GB FC interface, which provides greater storage capacity and performance. This is vital for storing virtual machine files and snapshots.
Ultimately, an administrator's biggest challenge is selecting products that not only fit a business's current needs, but can scale to fit the company's future needs.
Consider the cost and business impact
Upgrade or technology refresh initiatives much include a cost analysis and a careful evaluation of the business impact.
Although it may seem that tight IT budgets disqualify new system purchases, upgrades of aging systems can also be expensive, especially if proprietary parts are needed (or even discontinued) or the system is not covered by a service contract. The greater upfront investment of a technology refresh is frequently offset by creative financing offers, a period of free support, staff training and greater system capabilities that require fewer future purchases or upgrades. As a result, an organization may find that it's more cost effective to implement a technology refresh.
A cost sometimes overlooked in regards to a technology refresh is the changes a new system imposes on workflow, policies, procedures and production. Traditionally, installing a new system was highly disruptive -- and painful. A system would need to be backed up, taken offline and removed. Then, the new system would be installed, tested, restored from backups, and finally configured for service -- assuming there were no other errors or failures to contend with.
Virtualization has eased much of this pain. It allows a new system to be installed and configured in the production environment alongside working systems. Once the new system is ready, workloads can be migrated non-disruptively to the new system and tested. The old systems can then be withdrawn or reallocated as time permits.
|Stephen J. Bigelow, senior features writer, has more than 15 years of technical writing experience in the PC/technology industry. He holds a BSEE, CompTIA A+, Network+, Security+ and Server+ certifications and has written hundreds of articles and more than 15 feature books on computer troubleshooting, including Bigelow's PC Hardware Desk Reference and Bigelow's PC Hardware Annoyances. Contact him at email@example.com.|