IT and facilities personnel increasingly depend on one another to deliver high-availability services to data center clients. Is your company effectively managing the most critical parts of this key relationship? To answer this question confidently, it helps to understand which shared responsibilities between these two groups lead to the most common availability failures.
From there, we can apply proven methods to minimize availability risks. Industry data shows that among computer operations failures caused by equipment, the majority occur within the computer room rather than in the facility infrastructure building.
The reason behind this is simple: people are typically present in the computer room, not in the building housing the support equipment. The risk is often complex, partly because the computer room is managed by multiple departments, and partly because operating the various systems supporting the computer room requires a diverse range of skills.
Given these factors, the probability of confusion and mistakes is very high. Avoiding this situation requires close collaboration among all departments, with clearly defined specific processes and responsible personnel for mission-critical tasks.
Cabling and Connection Procedures
This is a top priority item. A successful IT/facilities team needs to clearly define which departments and individuals are authorized to install and remove power, network cables, and connections within the computer room. Each department needs to draw clear lines regarding their respective responsibilities and reach a consensus. They need to define which side of the server rack the facilities department should connect power cables to;
which department will be responsible for connecting the Power Distribution Unit (PDU) whip to an individual piece of computing equipment;
which department has the authority to remove network connectors, and so on. Diagrams or photographs should be accompanied by written descriptions of responsibility boundaries to further clarify each group’s duties. Each department should allow only 2~3 people to be responsible for these critical cabling tasks: one person holds primary responsibility, while the others provide support. Continuity in these roles is very important, as consistency minimizes the potential for errors.
Hardware Master Plan
Assigning one person each from the IT and facilities departments to be responsible for the preparation and management of the computer hardware master plan is the next most important objective. These two individuals need to meet weekly to decide where newly purchased computers should be installed on the floor and within racks.
A computer room floor plan needs to be prepared and updated regularly based on environmental changes. This floor plan needs to show all deployed cabinets, racks, and standalone equipment within designated zones (where up to 3/4 of the space is pre-reserved for specific categories of computer hardware). Elevation views of racks should show the hardware to be installed from bottom to top and its specific installation location. The IT representative needs to ensure the master plan accounts for all special requirements, including connectivity, communication needs with other hardware or network equipment, and distance limitations.
The facilities team member needs to ensure the master plan considers specific cooling and power requirements, for example, placing a high-heat-output server at the bottom of a rack if cooling is supplied from a raised floor. Both IT and facilities managers need to designate a backup person. No computer hardware may be installed or removed without the permission of the IT and facilities master plan owners.
Fewer People——Fewer Problems
The third risk mitigation method is to reduce the number of people with access to the computer room. If the number of people accessing the computer room is minimized, the risk of errors can be greatly reduced. This requires strong support from senior management. Only personnel who need to work inside the computer room at least one day per week should be allowed unescorted access. All other personnel should be accompanied when entering the room. Certain selected staff members, if they possess a thorough understanding of computer room work procedures, may use temporary access cards to enter. This access list should be reviewed quarterly by management to confirm individuals’ ongoing requirements for computer room access.
Ensuring Procedural Clarity
There is another way facilities and IT staff can significantly reduce risk in the computer room environment: every individual who will enter the computer room needs to read the complete set of data center computer room procedural guidelines. This document should describe all precautions for work activities and is typically 10~15 pages long. Managers from each department need to review this document together with employees and vendors authorized to work in the computer room (whether escorted or unescorted). Signed copies need to be filed and updated annually. For those individuals taking accompanied tours of the computer room (non-working visits), they need to read an abbreviated version of the document.
By specifically addressing these common disruptive factors that lead to operational failures and adopting the clearly defined procedures mentioned above, data center companies can significantly improve the continuity of their data center operations. Furthermore, continuous application operations will greatly increase the length of data center uptime.
銆怑ditor’s Picks銆?/p>