CDCP common mistakes
CDCP · Understand the traps and how to avoid them.
Treating the four CDCP domains as isolated silos and answering a cross-domain scenario question (for example a UPS room heat load question) using only the power domain knowledge, ignoring the cooling consequence.
Candidates revise domain by domain from a syllabus list and never practise questions that deliberately span two domains; the exam frames many items as operational scenarios where the correct answer depends on the interaction between power, cooling, and standards.
What to do instead
After each study block, write one sentence linking that domain to each of the other three (e.g. 'UPS losses become sensible heat load in the IT room; cooling sizing must include it'). Practise at least 20 scenario questions where you must state which two domains are involved before choosing an answer.
Confusing the roles of standards bodies: attributing the Uptime Institute Tier topology definitions to TIA-942, or attributing TIA-942 rated-4 requirements to Uptime Tier IV certification.
Both frameworks use similar words (Tier, rated, fault tolerant, concurrently maintainable) and candidates memorise the terms without anchoring each term to its publishing body and its purpose (design topology vs. constructed-facility certification).
What to do instead
Build a one-page comparison table with rows for publisher, scope, levels, and whether it certifies design, construction, or operations. For every practice question, before answering, name the body the question is implicitly invoking.
Mixing up N, N+1, 2N, and 2(N+1) when calculating capacity or interpreting a single-line diagram, and treating '2N' as meaning 'twice the load capacity in one system'.
Candidates memorise the letters but not the definition of a redundant component versus a redundant system; exam items often show a distribution path and ask how many concurrent faults the topology tolerates, which requires counting independent paths, not doubling numbers.
What to do instead
For each topology, draw the number of independent power paths from utility to the IT load and count the minimum components that must fail to drop the load. Practise explaining why 2N tolerates one full path failure while N+1 tolerates only one component failure.
Calculating UPS runtime or battery autonomy using the nameplate kW of the UPS instead of the actual connected load and the battery discharge-rate/end-voltage characteristics.
Candidates apply a simple kWh-divided-by-kW formula and forget that battery capacity is quoted at a specific discharge rate and end voltage, and that inverter efficiency and ageing margin reduce usable capacity.
What to do instead
Always compute runtime from the measured or design IT load, apply the manufacturer's discharge curve at the target end voltage, then apply the ageing and design margin separately. Work three numeric examples with different load fractions and verify the direction of each correction.
Assuming a UPS ride-through covers generator start and load transfer, or assuming the generator picks up the load instantly, when answering questions about availability during a utility outage.
Candidates treat 'UPS present' and 'generator present' as independent guarantees and never trace the sequence of events: transfer to battery, generator start and stabilise, transfer to generator, and recharge of batteries.
What to do instead
Write out the timeline of a utility failure as an ordered list with typical durations at each step, and identify which component must cover which interval. Then answer availability questions by walking that timeline explicitly.
Applying the ASHRAE recommended envelope and the allowable envelope interchangeably when judging whether a data centre space is compliant or when selecting setpoints.
Both envelopes appear on the same psychrometric chart and candidates remember 'ASHRAE says 18 to 27 degrees' without distinguishing recommended (reliability and energy guidance) from allowable (equipment tolerance limits) ranges.
What to do instead
Learn the two envelopes as separate facts with their purpose attached, and practise classifying a given temperature/humidity pair as inside recommended, inside allowable only, or outside both. Note which class (A1 to A4) the equipment is rated for.
Ignoring humidity control direction: not distinguishing the risk of too-low humidity (electrostatic discharge) from too-high humidity (condensation and corrosion) when answering environmental questions.
Candidates memorise a single humidity number and cannot map a symptom to a cause; exam items often describe a failure symptom and ask for the environmental root cause.
What to do instead
Create a two-column symptom-to-cause table for low and high humidity, including where condensation risk actually sits (cold surfaces, not room air). Test yourself by reading only the symptom column and recalling the cause and the corrective control.
Treating airflow management as a matter of adding cooling capacity: recommending more CRAC units for a hot spot instead of addressing bypass airflow, recirculation, or missing blanking panels.
Candidates equate 'hot' with 'not enough cooling' and skip the containment and airflow diagnostics that usually explain a local hot spot while return-air temperature looks normal.
What to do instead
For any hot-spot scenario, run a fixed checklist in order: blanking panels and open U spaces, cable cut-outs, containment integrity, perforated tile placement, CRAC airflow versus IT airflow, then capacity. Practise stating which measurement (delta-T, return temperature, tile flow) would confirm each hypothesis.
Answering change-management and maintenance questions by choosing the technically elegant solution rather than the procedural one (permit to work, method of procedure, rollback plan, approval authority).
Candidates with a technical background default to engineering fixes, while operations questions are testing whether the candidate follows controlled process; the distractors are often technically plausible.
What to do instead
When an item mentions maintenance, repair, or configuration change, first ask 'what procedural control applies?' and eliminate answers that skip authorisation, documentation, or rollback. Learn the standard sequence: request, risk assessment, approval, scheduling, execution under permit, verification, close-out.
Confusing reactive, preventive, and predictive maintenance, and mis-assigning a task such as thermographic scanning or battery impedance testing to the wrong category.
The categories overlap in practice and candidates classify by how often a task is done rather than by what triggers it (failure, calendar, or condition data).
What to do instead
Classify by trigger: reactive follows failure, preventive follows time or usage, predictive follows measured condition. Sort a list of 15 real tasks into the three buckets and justify each with its trigger, then revisit any task you hesitated on.