Think ‘tiers’ for a standards-compliant data center
It seems as if a big chunk of future contracting work will come from designing, installing and supporting data centers. There are many reasons this is being headlined in the telecommunication trades: it’s a solution to many problems. What exactly is a data center? Today it is defined as a highly secure, fault-resistant facility housing customer equipment that connects to telecommunications networks.
What used to be called the “computer room” or “equipment room” has now earned a new status—a “leading-edge technology” status in the name of the data center. Everyone seems to win with the data center:
• Hardware manufacturers can sell new servers and routers;
• Service and Application Providers can sell the programs and Internet services a data center links;
• IT management provides these services to employees, saves the organization money on the infrastructure—after justifying any unbudgeted expenses, and consolidates services so they can be managed efficiently;
• Designers/contractors get paid for designing and actually installing the infrastructure and maybe even the hardware (for an efficient design, remember); and let’s not forget the
• Consultants, who advise the organization and may also provide validation/certification services proving the network performs as the customer expects.
Some major users have clearly said it needs to be efficient. Well, efficiency stems from an understanding of what is needed and wanted. The Telecommunications Industry Association is developing a standard that will be dedicated to this purpose. It will be the first standard for cabling of today’s demanding data center. It’s truly an efficiency document that will give you a way to classify (categorize or make efficient) the structure of the data center according to the needs of the customer.
Design is broken up into Tiers 1 through 4 and covers the architectural, telecommunications, electrical and mechanical needs of the small to the large data center. Here is a look at a few of the recommendations drafted to date. Also, for your information, each Tier should meet the recommendations of the previous level. Understanding what makes up some specific design Tiers will go a long way toward helping you and the customer save money and provide the performance expected.
Note that while the concept of Tiers is useful to showing levels of redundancy, it is possible that some systems need to be of higher Tiers than others. For example, a data center located where utility electric power is less reliable than average might be designed as a Tier 3 electrical system but only a Tier 2 mechanical system.
A Tier 1 data center is a basic data center with no redundancy with a single path for power and cooling distribution and no redundant components. It is susceptible to disruptions from both planned and unplanned activity. It has computer power distribution and cooling, UPS and generators are single module systems, critical loads that may be exposed to outages during preventive maintenance/repair work, and data center disruption could be caused by operation errors or spontaneous failures of site infrastructure components.
Some Tier 1 recommendations include:
• All yard equipment should be anchored as per appropriate code.
• All interior partitions should have a minimum one-hour fire rating (two hours preferred) and extend from the floor to the underside of the structure above.
• There should be one customer-owned maintenance hole and entrance pathway to the facility.
• Communications infrastructure distributed from the Equipment Room (ER) to the main distribution (MDA) and horizontal distribution areas (HDA) throughout the data center via a single pathway.
• Installation of lightning protection should be based on a lightning risk analysis per NFPA 780 and insurance requirements
• The HVAC (Heating, Ventilating, and Air-Conditioning) system with air conditioning units is to maintain critical space temperature and relative humidity at design conditions.
A Tier 2 data center is a data center with redundant components. It has a single path for power and cooling distribution and it has redundant components on this distribution path.
A Tier 2 data center is slightly less susceptible than a Tier 1 to disruptions from both scheduled and unscheduled activity. The UPS and engine generators design capacity is on an N+1 (this redundancy provides one additional system in addition to the minimum required to satisfy the base requirement) basis, which has a single distribution path throughout the design.
Some Tier 2 recommendations include:
• All security doors should be solid wood with metal frames and to full height (floor to ceiling).
• Intra-data center LAN (Local Area Network) and SAN (Storage Area Network) backbone cabling from switches in the horizontal to backbone switches in the main distribution area should have redundant fiber or wire pairs.
• To provide redundancy, racks and cabinets should each have two dedicated 20A 120V electrical circuits fed from two different PDUs (Power Distribution Unit) or electrical receptacles.
• If the air conditioning units are served by a water system, the components of these systems are to be sized to maintain design conditions, with one redundant unit.
• Maintenance of the critical power path and other parts of the infrastructure will require a processing shutdown.
A Tier 3 data center is a data center that is “concurrently maintainable” because the redundant components are not on a single distribution path (as in Tier 2). The data center can have infrastructure activity going on without disrupting the computer hardware operation in any way. These data centers should be manned 24 hours a day—for maintenance, planned activities, repair and replacement of components, addition or removal of capacity components, testing, etc.
The Tier 3 data centers are often designed to be upgraded to Tier 4 when the business case justifies the cost of the upgrade (additional protection).
Some Tier 3 recommendations include:
• At least two access providers should serve this data center with the provider’s cabling from their central offices or POPs (Points of Presence) separated by at least 66 feet along their route.
• Mantraps at all entrances to the computer room should control letting more than one person in by the use of only one credential.
• A signal reference grid (SRG) and lightning protection system should be provided.
• If the HVAC system’s air conditioning units are served by a waterside heat rejection system, the components of these systems are to be sized to maintain design conditions, with one electrical switchboard removed from service. The piping system or systems are dual path.
• Two independent sets of pipes are to be used for data centers using chilled water.
• Sufficient capacity and distribution should be available to simultaneously carry the load on one path while performing maintenance or testing on the other path. If there are errors during operation or any other unplanned activities, disruption will be caused.
A Tier 4 data center has multiple active power and cooling distribution paths. This design provides a higher degree of fault tolerance—because at least two paths are normally active. There are multiple power feeds to all computer and telecomm equipment and all the equipment has to have multiple power inputs. This higher tier provides the ability for activity without disruption.
As you can see, this Tier 4 provides the infrastructure capacity and capability to permit any planned activity without disruption to the critical load. It also can sustain one worst-case unplanned failure or even with no critical load impact. This requires simultaneously active distribution paths, and electrically, this means two separate UPS systems where each system has N+1 redundancy. This is the most compatible data center level with high availability information technology concepts that employ CPU (Central Processing Unit) clustering, RAID/DASD (Redundant Array of Independent Disk/Direct Access Storage Device) and redundant communications to achieve reliability, availability, and serviceability.
Some Tier 4 recommendations include:
• A separate building or enclosure should include a secured generator pad.
• Critical systems should have horizontal cabling to two horizontal distribution areas-optional redundancy.
• To accommodate two active distribution paths, electrically, two separate UPS systems are required.
• The piping system(s) are dual path so as not to cause interruption of the air conditioning systems.
After reviewing this concept, it’s easy to recognize the importance of providing a fully functioning and efficient data center, regardless of its size. The hardware installed inside still needs to be supported by the infrastructure so that it performs for the customer. The brief points listed (only a small part of what will be finally written) above will become very educational as well as a successful sales tool for any contractor or designer. Not only will they come to know how a data center needs to support itself, they will also be able to provide their customer with a valuable look at what they should have to do for what they want to do. Look for the final and ratified standard, TIA 942, (from TIA’s TR 42.1.1 subcommittee) to make sense of the data center.
To follow this standard, see www.tiaonline.org where it will be announced when it becomes available to the public. EC
MICHELSON, president of Jackson, Calif.-based Business Communication Services and publisher of the BCS Reports, is an expert in TIA/EIA performance standards. Contact her at www.bcsreports.com or firstname.lastname@example.org.