From Emerald Green to Murky Brown. Is This Your Data Centre Journey?29/04/11
You have consolidated and automated and virtualised and de-duplicated and contained and coupled.
You have obtained an above average PUE, DCiE. WUE, ERE, CUE and DCcE.
You are compliant with EPEAT and RoHS and WEEE directives.
You are rapidly moving up the Data Centre Maturity Model.
You have reduced costs and carbon emissions significantly.
The CEO, CFO, CIO and CMO all love you. Your data centre is a shiny emerald green.
Whether this reflects your reality or it is a state to which you are aspiring, the questions you need to ask yourself are:
When I reach the ideal state, how am I going to keep it that way?
How am I going to ensure that the organisation recognises the return on all that investment and hard work?
How am I going to stop my emerald green hue turning into one that is murky brown?
Why should you ask those questions?
Because – your challenge is people!
Once you open the doors to the emerald green data centre and allow people access, it can quickly become murky brown again. The organisation will have made a substantial investment in establishing a green data centre. You may not be the size of Google who invested $890 million in its data centre infrastructure in the first three months of 2011 but I would suggest you could be looking a multi-million dollar investment over a period of time. You don’t want to be held responsible for that investment going to waste.
The key is to have good processes and procedures for people to follow that will support all the technology solutions that you have adopted. The solution lies in Service Management and the best practice guidance as contained within ITIL. It has all the processes you need to protect your investment and enable your continual journey towards improved sustainability.
The Service Management lifecycle from Service Strategy through Service Design through Service Transition and Service Operation through to retirement of services contains all the processes you need to keep your green data centre clean and shiny.
The following are just some examples.
Service Strategy includes the process of Service Portfolio Management. Service Portfolio Management through the activities of “define, assess, approve and charter” should ensure that there is no duplication of services and associated applications. The introduction of new services that could have been supported by a change to existing application(s) rather than creation of new ones will only serve to unnecessarily increase the carbon footprint of the data centre. Back in 2009, Intel announced that over the preceding 2 years they had reduced their applications by 37% towards their goal of 50%. They expected that retiring applications would result in a net present value of more than USD 50 million. Consider the associated reduction in carbon emissions through the release of infrastructure that was required to support the unnecessary 50%.
When designing services within Service Design every consideration should be given to reducing the environmental impact of the service. This should include developing software in a way that can be used for long periods of time without becoming outdated; developing applications that will run in the most effective and efficient manner; developing services with the lowest possible hardware requirements. Poorly designed services will increase the carbon footprint of your data centre unnecessarily.
As well as the obvious candidate in support of sustainability – Capacity Management, Availability Management, IT Service Continuity Management and Service Level Management are key processes. These processes should ensure that the availability requirements of the business for their services are appropriate to the value and priority of that service to the business. There is an environmental (as well as financial cost) in the provision of high levels of resilience, storage and contingency. If it is not required, don’t provide it in your data centre.
Service Asset and Configuration Management (SACM) within Service Transition is a key player in keeping your data centre green.
If you haven’t already done so, a first step towards your green data centre is to do a complete inventory of servers, software and applications, including the interdependencies between them all via SACM. You need to firstly understand how each physical and virtual server is used, what software is running on it, which business applications it supports and what its actual value is to the business. Then you can work out what to remove, refresh or virtualise unused, unnecessary and inefficient assets.
Once you have done this you need to keep a track of what is in your data centre and how it is being used, so regular audits through SACM are crucial. Comatosed equipment will start to pop up as developers decommission applications without the associated removal of redundant hardware. SACM will ensure that this is detected and remedial action can be taken.
Back in 2009, Sun Microsystems undertook a clean-up of data centre facilities at four of their major campuses and pulled out over 440 pallets of equipment. There were 6,199 devices in total with 4,100 of them being servers. 64% of those servers were still powered on and consuming energy around the clock. Sun equated the environmental impact of this to puling 6000 cars off the road.
You don’t want to have to do this to your emerald green data centre a few years down the track just because of a lack of a good SACM process!
Change Management of course is key in ensuring that every change is considered for its environmental impact and that redundant equipment resulting from the change is identified (in conjunction with SACM and Release and Deployment Management) so that it can be removed in a timely manner.
Change Management should ensure that Requests for Change (RFCs) are accepted or rejected using a rationale that includes the environmental impact of that change as well as the financial, business and technology impacts.
The rationale can include environmental requirements such as:
- the use of suppliers of products and services with environmental and sustainability management systems and ISO 140001 accreditation;
- the utilisation of devices that conform to a specified energy rating;
- the use of devices that have a lifetime energy footprint within specified parameters;
- the identification and removal of redundant components – infrastructure and applications – and their reuse, recycling or environmentally responsible disposal;
- sustainable release and deployment methods including remote access and local distribution;
- the inclusion of environmental targets within SLAs, OLAs and UCs; and
- adherence to the organisation’s sustainability policy, objectives and targets.
All of this will contribute to keeping the data centre green.
In Service Operation, Event Management should be used to detect deviation from the expected environmental performance of the data centre and ensure that Incidents are created and directed to the most appropriate resource to undertaken investigation and diagnosis as soon as possible. Corrective action can then be taken to get the data centre back to its emerald green status.
Where Incidents related to deviation from expected environmental data centre performance is repeatedly occurring, Problem Management should be used to determine root cause and eliminate it.
Continual Service Improvement contains the processes that enable the data centre and facilities team to look at ways in which the sustainability of the data centre can be continual improved and become an even brighter, cleaner green. Service Measurement and Service Reporting will be imperatives in keeping a handle on current performance against targets and determining where improvements can be made.
Service Management is the data centre or facilities manager’s friend. The effective implementation of the aforementioned processes and the others within Service Management will result in the protection of the data centre from undesirable activity that starts the journey from emerald green to murky brown.
I have only mentioned a few of the Service Management processes in this article.
At Macanta we have developed the eco-ITSM service, a world first, which uncovers the sustainability aspect of every process, function and activity within service management. More information is available at www.eco-itsm.com.
 PUE = Power Usage Effectiveness, DCiE = Data Centre Infrastructure Efficiency, WUE = Water Usage Effectiveness, ERE = Energy Reuse Effectiveness, CUE = Carbon Usage Effectiveness, DCcE = Data Centre Compute Efficiency
 EPEAT = Electronic Product Environmental Assessment Tool, RoHS = Reduction in hazardous Substances, WEEE = Waste Electrical and Electronic Equipment