The process of purchasing a server is relatively straightforward, but working out the details of a hardware service...
contract tends to require significantly more effort.
The need for a support contract is often overlooked because many in IT assume the hardware warranty protects the company if any problems occur. Although a warranty offers some assurances, it is often inadequate on its own.
For example, suppose a server's system board fails, but it is covered under warranty. Each vendor has its own way of handling this type of issue. Typically, the administrator would need to ship the system board to the vendor before it sends a replacement. In contrast, a support contract can provide same-day service for the replacement and professional installation by a certified technician.
Map out the company's needs
Prior to negotiating a hardware service contract, consider what matters most to the organization. Why obtain a support agreement in the first place? Does the organization require immediate access to hardware parts during a critical outage? Does the IT staff lack the technical skills to handle hardware-level repairs? Make sure that any service-level agreements the organization must adhere to are part of the equation.
Keep these factors in mind during discussions with a support vendor, and make sure you address three key areas in a hardware service contract.
Pin down terms to avoid a lengthy outage
First, negotiate the response time. When a critical issue hits the data center, there should be no doubt about the availability of the support vendor.
Most rapid response support contracts are expensive because they might require the provider to hire extra staff members. One way to reduce this cost is to negotiate a two-tier response time. For example, the contract might require the provider to respond within 48 hours for noncritical outages, but also to have a tech on site within an hour for any outages the organization deems critical.
Second, lock down the availability of replacement hardware. It's pointless to have a contract that requires the provider to respond to a critical outage within an hour if it takes the needed parts a week to arrive.
At one time, organizations relied almost exclusively on physical servers, and the server's operating system was tied to its specific hardware configuration. Backups could not restore to dissimilar hardware. To account for this, most service agreements required providers to have exact duplicates of the organization's hardware so it could swap out an entire server if necessary.
Server virtualization makes this less of an issue, but the provider's inventory remains an important consideration. During an outage, an organization needs to get back online as quickly as possible. As such, a good contract for hardware service should require the support vendor to maintain an inventory of spare parts that match your server hardware. It is also a good idea to make sure the agreement provides loaner servers if the service vendor does not have the required parts immediately available.
The hardware support agreement should address the quantity of repair parts the vendor needs to keep in stock. Multiple servers can break at the same time. The support contract should eliminate any chance a cascading failure would leave the company vulnerable.
Third, consider adding warranty handling to the hardware service contract. This is less critical than the other items, but it is worth considering. Because some of the hardware is covered under warranty, ideally, the support provider should handle the warranty claims.
If a system board fails, then the service vendor should replace that system board with a spare, file a warranty claim and ship the failed part to the manufacturer. This offers the dual benefit of a quick system recovery and frees the IT department from dealing with warranties.