CDCP Data Center Tiers and Availability Standards — Questions and Answers
Question 1: According to the Uptime Institute Tier Standard, which tier requires fault tolerance with multiple active power and cooling distribution paths, such that any single failure causes no impact to IT load?
- Tier I
- Tier II
- Tier III
- Tier IV (Correct answer)
Correct answer: Tier IV
Tier IV is the highest classification and requires fault tolerance — all components are fully redundant and multiple active distribution paths exist simultaneously. A single failure or planned maintenance event does not interrupt IT operations. Tier III only requires concurrent maintainability (N+1), not fault tolerance (2N).
Question 2: A Tier III data center is described as 'concurrently maintainable.' What does this mean in practice?
- All redundant systems automatically switch over without human intervention
- Any planned maintenance can be performed on any component without shutting down IT equipment (Correct answer)
- The facility can operate on generator power alone for up to 96 hours
- Two separate utility feeds are active simultaneously at all times
Correct answer: Any planned maintenance can be performed on any component without shutting down IT equipment
Concurrent maintainability means that maintenance, repair, or replacement of any single capacity component or distribution element can be performed without shutting down the IT load. This requires at least N+1 redundancy with an alternate distribution path. Tier IV extends this by also requiring fault tolerance.
Question 3: The Uptime Institute Tier I standard specifies an annual downtime allowance of approximately how many hours?
- 0.4 hours (about 26 minutes)
- 1.6 hours (about 96 minutes)
- 22.0 hours
- 28.8 hours (Correct answer)
Correct answer: 28.8 hours
Tier I has an availability of 99.671%, which corresponds to approximately 28.8 hours of allowable downtime per year. Tier II allows ~22 hours, Tier III ~1.6 hours (99.982%), and Tier IV only ~0.4 hours (99.995%). Higher tiers have both greater redundancy and higher availability targets.
Question 4: In data center redundancy notation, what does '2N' mean?
- Two independent utility power feeds from separate substations
- Full duplication of every component — twice the capacity needed to run the load (Correct answer)
- The N+1 system plus one additional cold standby unit
- Two data centers operating in active-active mode
Correct answer: Full duplication of every component — twice the capacity needed to run the load
2N means every active component (N) has a complete duplicate — the system has twice the capacity required to support the load. This enables true fault tolerance: if one entire path or system fails, the duplicate carries the full load immediately. Tier IV requires 2N (or 2N+1) redundancy for power and cooling.
Question 5: TIA-942 is the standard governing data center design. TIA-942 Rating 3 is most closely equivalent to which Uptime Institute Tier?
- Tier I
- Tier II
- Tier III (Correct answer)
- Tier IV
Correct answer: Tier III
TIA-942 uses Ratings 1–4 that broadly align with Uptime Institute Tiers I–IV. Rating 3 / Tier III both require redundant components and concurrent maintainability. Note that TIA-942 and the Uptime Institute are separate certification bodies with distinct audit processes, so 'equivalent' is approximate rather than officially interchangeable.
Question 6: A data center with N+1 redundancy for its UPS systems has a load requiring 4 UPS modules. How many UPS modules are installed?
- 4
- 5 (Correct answer)
- 6
- 8
Correct answer: 5
N+1 redundancy means the number of components needed to support the load (N=4) plus one additional spare (1), totaling 5 UPS modules. If any single module fails, the remaining four still carry the full load. This is the minimum redundancy level associated with Tier III concurrent maintainability.
CDCP Data Center Tiers and Availability Standards
According to the Uptime Institute Tier Standard, which tier requires fault tolerance with multiple active power and cooling distribution paths, such that any single failure causes no impact to IT load?