I've said it before and I'll say it again; most IT hardware consists of electromechanical components that will most assuredly fail, so how can you minimize downtime when something in your server fails? I mentioned in a previous, three-part post, the importance of redundant components, but just as important is repairing/replacing what has failed.
Most new servers from the larger vendors such as Dell, IBM, HP or Sun, will include a limited manufacturers warranty that covers repair for 1-3 years and response during business hours. While this "free" service is good, the response times or level of effort by the vendor or their contracted support arm may leave something to be desired.
Most of the vendors mentioned above will also offer support contracts with better response times, extended coverage hours and even guaranteed parts availability. These support contracts can be expensive, but could be worthwhile for your most critical servers. Purchasing a support contract packaged with the server can soften the financial blow, especially if your organization capitalizes their assets as the contract could possibly be charged as capital rather than expense.
Lastly, many vendors offer post-warranty support contracts and this might be preferable to maintaining a parts depot of your own, purchasing replacement components from 3rd party vendors or cannibalizing other equipment to keep things running. The only drawback with these types of support contracts is that they usually become more expensive as the equipment gets older and at some point you have to consider whether buying a replacement server is more economical than paying through the nose to keep those legacy servers running.
Comment Preview