Making the most of server maintenance contracts

When trying to obtain the most value from your servers, watching terms and doing homework on server maintenance contracts can save at the negotiating table.

With IT budgets under increasing scrutiny from CFOs and CEOs, IT managers are finding less capital available to replace aging servers or maintain current ones. At the same time, vendors searching for ways to improve sales of new hardware products will often increase maintenance and support fees on aging servers. This presents a challenge for data center administrators; find a way to support your aging servers while getting the best value for your dollar on all of your service contracts?

No two server maintenance contracts are the same
First, it’s important to understand that no two service contracts are identical. There are certainly similarities, but every server vendor has different terms, conditions, service levels and price points. The service level (or response time) is probably the most recognized issue, and faster response times can cost dramatically more than slower response times. The trick is to negotiate a contract that provides the service you need at a price that makes sense. This is often overlooked or misunderstood considering the vast diversity of contract terms. For example, some vendors have collapsed their offerings into a flat service level or price, while other vendors have expanded their offerings to give customers a variety of cost/service options for less critical systems.

While not every vendor offers multiple levels of response, every vendor offers the option of combining your hardware and software support. Often, the price for combining these is less than buying them separately. But lookout for “split” response times. Traditionally when dealing with vendors, service contracts cover not only the hardware but the operating system software as well. But just because you're getting hardware and software support doesn't mean they're necessarily the same response levels. Sometimes, buying 24/7 support with a four-hour response for your hardware does not translate to a four-hour response for software problems.

Looking for ways to save on server maintenance contracts
Like many things in IT, buying service contracts is fraught with pitfalls and caveats that become increasingly important to be aware of in a budget-constrained world. There are numerous considerations that can dramatically affect the contract terms and pricing, so let’s cover some of the more important issues.

Avoid the tendency to overbuy. In the past, overbuying support was often a way of life, a way of covering those unforeseen "just in case” scenarios that help busy administrators sleep better at night. Rather than deal with the confusion and page after page of contract terms, most businesses would select the highest service level they needed and then put all systems under that one contract. It made things much simpler and enabled organizations to make decisions rapidly. These days are largely over, especially for small and mid-sized shops operating on a shoestring. Today, the goal is to select the correct level of service for each machine (or group of machines) wherever you can. For example, it’s probably not a good idea to buy a one-hour 24/7/365 response time for a development server that’s down three days a week anyway.

Be skeptical of operating system or business application support. When it comes to operating system or bundled application support, beware of any company trying to sell you third-party software support. The fact is that only the vendor producing the operating system (such as Microsoft) or other business application can offer you support. Some partners may be able to offer you discounted rates on operating system support from a vendor, but only the vendor can offer you things like patches, updates and installation media. It might make sense to forego software support for stable and mature servers that you don’t expect to update.

There are other ways to save money on your service contracts when it comes time to negotiate. Always try to have as many systems on the same expiration date as possible, as most vendors will incorporate the number of systems into any discounting plan. Ask for quotes for hardware and software separately as well as together, and get a quote from reputable third-party service providers even if you won't necessarily use them. Look back at your previous service contracts and have a clear and concise list of any problems or concerns you have with your current contracts. If you have unresolved issues, bring details of those issues and ask how they intend to resolve them. And don't simply accept a blanket policy that may be more or less than what you need because that's what the vendor has on its price list; most will work with you to some extent.

The move to third-party server maintenance contracts
Another concern is with systems that simply can't afford to be replaced during this refresh cycle, but which support is running out on and will become unavailable before they can be replaced. Many businesses are putting off replacing systems which are still meeting performance requirements but are at or near the end of support availability from the original manufacturer. Third-party support may be the only option for these systems, as vendors tend to make support on older systems prohibitively expensive to encourage replacement.

Vendors universally employ extremely strong language to caution against third-party support, claiming that service and parts are likely to be inferior or the staff will be unqualified. However, when evaluating companies that provide third-party hardware support contracts, it’s easy to question the background of the field engineers -- many third-party field engineers working on systems were formerly employed by the actual vendors. In addition, third-party providers understand that the onus is on them to prove they're up to the challenge, and they actively seek the best technicians they can find. The better a technician is, the less time a system or part is unavailable and the quicker problems are resolved more, making it possible to answer more service calls. The greatest concern from system administrators, however, was the backing of the vendor. In spite of experience and education, there are some problems that simply can't be diagnosed or solved by anyone but the original manufacturer. It is possible for third-party providers to have agreements in place that let them call the vendor on these issues.

Another major difference with third-party hardware support is the availability of parts. While vendors build enough parts for new systems plus a supply for spares and replacements, a third-party vendor has other options. Many third-party vendors are able to use additional sources for parts to repair these systems, including direct from vendor channels, from other suppliers, and in some cases from their own warehouse of parts or off-lease systems. For some older systems, this can give third-party service contracts a significant advantage over paying for extended life support from a vendor who may not always guarantee part availability. The questions of engineering expertise and parts availability are crucial in third-party service discussions.

The option of server self-maintenance
What about maintaining servers yourself? In most cases, this is the least attractive and least reasonable option. Servers that are past their support life are generally much less modular than current systems, and have an added degree of challenge in the age and scarcity of parts. Opting to buy a second working server as a source for spare parts is usually questionable from a financial standpoint, but the more pressing problem becomes who and how those parts will be installed. When it comes to old enterprise hardware, you can’t just run down the street to Radio Shack for a new mezzanine cable. Most system administrators would be very uncomfortable with the added responsibility in production environments, and would only do it as a last resort on very old hardware.

Knowledge and involvement can leverage service contracts
But perhaps the most important negotiating tactic you'll have is to be prepared and present a unified front as a business when dealing with your maintenance contract vendor or third-party providers. Discuss issues and concerns with your system, storage administrators and application administrators, as well as developers, and determine what service levels are really needed. Set realistic and achievable goals for your business and for your vendor and match hardware and software support levels everywhere you can. Seek opinions from every stakeholder, start discussions early, and get everyone on board with the right service-level expectations before you enter negotiations for your next round of service contracts. Because the best way to save money is to know exactly what you need and what you're buying.

Dig Deeper on Data center careers, staffing and certifications

Cloud Computing
Sustainability and ESG