Notes from the Consultant’s Jungle

The data center consolidation and construction boom of the 21^st century is well underway.  As we work with Clients, helping with planning, advice, and project management of these changes foundational to the future performance of their enterprise, we always encounter the discussion of facility Tier Level rating.

The Tier Level rating refers to an industry standard way of describing the degree to which the facility can support constant uninterrupted operation of the data processing systems.  We know that the systems themselves can be architected with high-availability configurations.  Tier Levels though, refer to the capability of the facility itself to support the systems it contains.  Utility power, temperature increases and so on are facilities issues, and are the foundation upon which any amount of data processing fault-tolerance stands.

The American National Standards Institute (ANSI) and the Telecommunications Industry Association (TIA) are examples of organizations that formulate standards for the industry to follow.  The TIA developed a specification entitled TIA-942 Telecommunications Infrastructure Standard for Data Centers.  This is the most widely referenced standard when talking about data center facility availability.  From the title one might be inclined to think this is a specification of telecom for data centers.  It is that, but it’s much broader than that including cabling, space layout, site selection criteria, and infrastructure Tiers/availability.  This last point, represented by Annex G in the TIA-942 standard, is where all the talk about data center Tiers comes from, and Annex G was authored by  The Uptime Institute (or, ‘The Institute” for short).  Thus, when we hear people speak of, say, a “Tier-4 Data Center,” they are referring to the Tier classifications from TIA-942, or the Uptime Institute.

The Tier classification model provides an objective basis for comparing or describing the functionality, capacity, and cost of a data center’s facility architecture.  In particular, the Tier classification model is focused on the availability of the facility itself, and is driven by the infrastructure to power and cool the data processing environment.

Now, we recognize that the above brief review of Uptime Institute Tier Level genealogy is possibly well known to many readers.  What seems to be not so well known though is what the “critical to quality” features of a data center are that differentiate it from one Tier to another.  This point also, is the motivation for this article.

“Do I need a fence to have a Tier-4 data center?”

We often hear a lot of questions from Clients about how the height of the raised floor, perimeter security, number of parking spaces, and so on will impact planning for reaching the desired Tier rating of the facility.  These questions enter our Clients’ minds, based on their own reading of TIA-942 guidelines which list lots of features typically found in the various tier levels of mission critical facilities.  These points though, are not “critical to quality” in the true spirit of what the Tier Levels are intended to deliver, which is availability.

The Tier Level ratings are focused on facility availability.  The availability of the facility infrastructure is foundational to any degree of clustering, fault-tolerant systems, virtualization, or anything else one will architect into the data processing footprint to provide “always-on” service.  If the data processing equipment is architected for five-nine’s availability and the facility itself is one nine, all the ROI modeling to justify the extra nines in that data processing equipment is likely for naught.  Look at it this way:

“The only way to determine Tier Level is to objectively assess the facility topology’s ability to respond to planned and unplanned events.”

In other words, it’s the level of capacity and redundancy of MEP infrastructure components, and the topology of configuration that is important to the Tier Level rating.

The Institute guidelines cite a 30-inch raised floor as typical for Tier-3 data centers.  This is based on the assumption that the data center is using the raised floor as a plenum for cool air distribution to IT equipment, and as such needs a certain height to maintain the anticipated volume of air flow.  However, there are ways to cool IT equipment without the use of a raised floor as an air plenum.  If the facility uses such a technique and doesn’t have a raised floor at all, it can still reach Tier-3 capabilities if the capacity, redundancy, and topology of MEP systems are correct for that Tier.

Similarly, no,… one doesn’t need a fence to qualify as Tier-4.  One has to ask why there wouldn’t be good perimeter security for such an investment in facilities and systems, but technically a fence is not required.  If there is an identified risk to the facility (such as physical intrusion in the case of fences) then the risk of impact to availability should be considered in the design of the facility.  The same holds for the other ancillary characteristics that the Institute’s research has shown to typify sites of various Tier Levels.

The core factors contributing to the Tier Level rating of mission critical facilities are the following:

• Active Capacity Components to Support IT Load (N, N+1, etc.)
• Distribution Paths (single, single + alternate, dual)
• Concurrent Maintainability
• Fault Tolerance (to a single event)
• Compartmentalization
• Continuous Cooling Capability

These factors describe the level of redundancy and topology of the MEP systems, and are core to ensuring Tier Level availability ratings of the facility.  Ensuring sufficient availability in support of Business operations is the reason a certain Tier Level goal is targeted for the enterprise mission critical facilities.  The core factors impacting the facility’s availability capabilities are those listed above and are those important to reaching or assessing the corresponding Tier Level.