Expectations of data centers are focused on the five ‘pillars’ of availability, cost-effectiveness, flexibility, manageability, and security. These pillars let us make a data center checklist of best practices for the DC infrastructure.

Subscribe to the  Data Center Sales & Marketing  (DCSMI) Update Newsletter

Availability

Data centers have to stay up and running. Outages can be expensive in both financial and reputational terms. Data centers need to be organized to prevent such problems or at least to detect them at the earliest possible moment, including:

  • Resilient data center design with fire barriers and robust building architecture
  • Multiple connections to power providers, preferably entering the data center at different points
  • Uninterruptible power supply with battery backup and generators in case of power cuts
  • Redundant servers and storage with failover provisions at hardware and software levels
  • Multiple high-speed network links entering and exiting at different points
  • Sufficient air conditioning for all of the equipment being operated in the data center
  • Smoke, fire, humidity, and flood detection, including underneath data center raised floors

Cost-Effectiveness

Data center and colocation service providers need to be profitable. Customers expect cost savings. Data center operations, therefore, need to be high-quality yet also be efficient and avoid unnecessarily high levels of expenditure. What’s expected?

  • Tiered data storage according to needs for everyday, occasional, or archival use
  • Virtualization to maximize physical server productivity via virtual machines (VMs)
  • Automation of systems administration routines
  • Power and cooling analysis to prevent excessively hot or cold spots from forming
  • Ergonomic shipping and installation facilities, including weather-proof receiving docks
  • Proximity to access routes and fuel storage, while avoiding hazards like airports and oil refineries

Flexibility

Even if data center buildings seldom scale physically, their internal organization must facilitate scaling in power and storage capacity with new systems and handling peak user demands. This includes:

  • The ability of the data center to accommodate new technology with different power and cooling needs
  • Ease of switching to new operational procedures or ways to meet new safety requirements
  • Modularity in floor layout, electrical and mechanical design to adapt to market requirements
  • Management of requests for very popular information via overflow servers

Manageability

‘Lights-out’ operations may be possible, but at the very least, data center teams and customers must be kept appraised of performance and possible problems via:

  • Service level agreements with customers, including service response time and escalation paths
  • Overall and specific (per customer for multi-customer DCs) monitoring of service levels
  • Speedy registration and resolution of support requests, trouble tickets, and alarms
  • Coordinated computing equipment refresh process with a roadmap for customers
  • Standards compliance and certification now and into the future (e.g. PCI, TIA-942, SAS 70)

Security

Both in terms of staff and site safety and customer data confidentiality, security looms large in the data center checklist for infrastructure best practices, including:

  • Full compliance with safety regulations, including fire exits
  • Physical security with protection of power and networking links, and cable vaults
  • Video surveillance and motion detectors, badges, ‘mantrap’ entrances, data center guards
  • Data encryption, SSL certificates, firewalls, and also virtual firewalls (for VMs)
  • Locked cages with ceilings for customer systems, and locked cabinets as an option
  • Intrusion detection and prevention systems, behavioral analysis, and alerts to staff
  • Protected and tested data backup and disaster recovery procedures
  • Reliable and complete data destruction procedures (old hard drives, contract terminations)

This data center checklist is naturally a general one. Do you have an additional best practice for your own data center? Tell us about it in the Comments section below.

Other Related Resources

If you're in the data center, mission-critical, or cloud services industries, or you sell to the data center industry, don't miss our weekly update newsletter -- Data Center Sales & Marketing Institute (DCSMI) Update Newsletter. Get notified about new reports, events, podcasts, and blog posts.

Subscribe to the Data Center Sales and Marketing Newsletter (DCSMI)

Comments (4)

You may also like

Data Center Cabinet Costs with Shrinking Hardware Size
Data Center Cabinet Costs with Shrinking Hardware Size
28 December, 2015

Data centers have been growing rapidly, and this phenomenon is expected to double with the evolution of the Internet of ...

8 Data Center Asset Management Problems You Can't Ignore
8 Data Center Asset Management Problems You Can't Ignore
7 July, 2015

A data center, its servers, storage and networking, and the peripheral equipment needed for operation all add to a manag...

Comparing Software Defined Data Centers vs. Traditional Data Centers
Comparing Software Defined Data Centers vs. Traditional Data Centers
9 February, 2016

Data center technology professionals are faced with constantly escalated demand for computing capacity. The constraints ...