Standardization and Modularity in Data Center Physical Infrastructure

Overview

The benefits are traced to two roots: modular architecture (scalable, changeable, portable, swappable) and greater human understanding. These map onto the three parts of infrastructure business value: availability, agility and TCO. The largest availability entitlement is human error, which causes 50-60% of data centre downtime. On reliability the paper warns that analysis based on component count alone unfairly penalises modular systems; it must consider design, function and dependencies, plus fault tolerance and reliability growth. Citing MTechnology studies, factory-repaired UPS power modules are 500-2000 times less likely to cause outages, introduce new defects or inhibit return to full operation than field-repaired modules.

On agility, standardised modular deployment can cut concept-to-commissioning time from months to weeks and reconfiguration from weeks to days, partly through factory pre-test. With typical IT refresh cycles of two years, the ability to reconfigure, upgrade or move infrastructure is a significant part of agility.

A sidebar introduces micro data centres, self-contained factory-assembled and tested enclosures from 1 to 100 kW of IT load, as the next step. One example is Deloitte's twin Amsterdam data centres, which consolidated 109 IT facilities on standardised modular infrastructure and were built in four months. The text is a vendor view and rests on cited studies as evidence.

Key points

1. Benefits come from two roots, modular architecture and human understanding; three business values: availability, agility, TCO.
2. Human error causes 50-60% of downtime, the largest single availability entitlement.
3. Reliability analysis by component count alone can be incomplete and misleading.
4. Factory-repaired UPS power modules are 500-2000 times less likely to cause outages or new defects than field-repaired ones (MTechnology).
5. Standardised modular deployment can cut concept-to-commissioning from months to weeks and reconfiguration from weeks to days.
6. Typical IT refresh cycle is two years.
7. Micro data centres are factory-assembled, tested enclosures of 1-100 kW IT load.
8. Deloitte's twin Amsterdam sites consolidated 109 IT facilities and were built in four months.

Using it in design

• Let users define a reusable power/cooling/rack module and replicate it across phases and sites, with BOM and labelling generated consistently.
• Model installed versus ultimate capacity per block so the day-1 build matches current load.
• In the reliability engine, support module-level N+1 redundancy and MTTR for swappable modules instead of penalising higher part counts.
• Offer a 1-100 kW micro data centre as a prefab template type.

Who should read it

Engineers and decision makers weighing repeatable, phased data-centre design and a standard infrastructure block strategy.

Publication details

Authors

Suzanne Niles

Publisher

Schneider Electric

Year

2015

Document number

WP 116

Document language

English

Source type

Manufacturer

Access

Free to read

Licence

Copyright retained; citation only — the publisher’s terms apply to the original document

Citation