How to deal with the inevitable…
Data centers are a brutal environment for mechanical and electrical equipment. The core data infrastructure of switches, servers, cabinets, and communications wiring requires a supporting cast of mechanical, electrical, fire alarm, and emergency back-up technology to keep it all running. As with everything in nature and technology, this supporting equipment has a service life…which also means that it will eventually die.
Given the existential nature of their task, planning for the eventual retirement or end of life (EOL) of a data center’s mechanical and electrical equipment is a required exercise for the data center operator. It may be a distant thought in the heady days of data center planning or acquisition, but the time comes remarkably quickly in a 24/7/365 data center. Unlike most commercial buildings, critical electrical and cooling equipment needs to be refreshed before the invariable high maintenance end of the service term or – worst of all – catastrophic failure.
From an economic and even emotional perspective, replacing equipment that is operating as designed seems frivolous or somehow cruel. Nevertheless, waiting too long, particularly for items that fail suddenly, will lead to far greater expense. Most data center operators understand this, but initiating an EOL planning process can be daunting, particularly for organizations with multiple locations or addresses.
The question for an owner, particularly those with sites of varying designs and providence – is how and when to replace equipment. More specifically, how does an owner determine what equipment is at or near its EOL and needs replacement? How can it be tracked across multiple sites or even multiple continents? What implications does this have in relation to re-investment, divestment, or re-building of a location or multiple locations?
And, perhaps most importantly, how does an operator prioritize limited capital funds in a dynamic and changing environment? This economic challenge leads to more nuanced and complex questions, such as:
- Which items have the highest priority in terms of remaining service life?
- Conversely, what equipment can wait until next year, or maybe even five years from now?
- How bad is the pain if the equipment fails? Is it an inconvenience, a major expense, or an outage?
- Can I repair or refresh some items in my inventory to extend life without a full replacement?
- What economic gains do I receive from new equipment in terms of efficiency, maintainability, or reliability that might help prioritize some investments over others?
- And – which items are so costly or risky that they need to be upgraded in an urgent fashion, regardless of other investment opportunities?
From a management or operational perspective, the more pragmatic and pressing questions are – when do I need to start planning for EOL, and how?