Would you like to be redirected to Logicalis United States?Close
Data centres are secure buildings that house networked servers, storage and backup facilities.
A typical data centre design build service will provide space for hardware in a controlled environment, for instance using power and environmental cooling and air conditioning to enable the equipment to perform at its optimum level with maximum system availability.
Data centre services provide various levels of resilience in the form of backup power supplies and additional communications connections that may not be used until a problem occurs with the primary system - this is known as redundancy.
The main purpose of our data centre design build services are to run core business applications and store operational data as well as providing Disaster Recover (DR) facilities. Typical applications will be enterprise software systems such as Enterprise Resource Planning (ERP) and Customer Relationship Management (CRM) services.
Common components include firewalls, VPN gateways, routers and switches, database servers, file servers, application servers, web servers and middleware - all contained on physical hardware or on consolidated and virtualised platforms.
The Telecommunication Infrastructure Standard 942 provides guidance on standardisation of data centre design and classifies data centres into four tiers with level 4 being the most fault tolerant and guaranteeing 99.995% uptime, compared with Tier 1 which will guarantee 99.671% uptime. This standardisation is important for customers to understand and measure service providers against.
Most data centre deployments are carried out for the following reasons:
Power usage efficiency (PUE)
When designing new services or indeed a new data centre, it is important to consider this factor. An internationally recognised metric used to measure the efficiency of data centres in terms of energy, it measures the efficiency of power being provided to the whole facility, compared to power used by the ICT equipment, and provides a ratio of efficiency. For example a facility’s ICT equipment may draw 800Kw, and the cooling and environmental systems may draw a further 800Kw, this would be a PUE ratio of 2.0, where it is taking twice as much energy as is required by the ICT equipment to run the overall facility. PUE ratios of close to 1 are the most desirable, as they are the most energy efficient.
To consolidate, virtualise or not?
Consider whether all infrastructure is going to be hardware-based, or virtualised. An all hardware model will use more energy, while full or part consolidation and virtualisation will reduce energy consumption.
How is the data centre to be monitored and who will be doing this? What is the escalation process and where do lines of responsibility start and end? Is energy efficiency being monitored on an ongoing basis?
Service Level Agreement (SLA)
The increasing role of the CTO in negotiating contracts is crucial, in particular the SLAs that govern data centre and provider performance.