We've been working in the field of Green IT for over 10 years now, and its been an interesting journey.
Some readers will know that I had my "Road to Damascus" in 2003, when I visited the Eden Project in Cornwall, we'd spent the day walking through the tropical and semi arid desert biomes and the external areas and on our depature, we had to exit via the shop. The shop was full of all the sorts of things you'd expect a living laboratory site to have, books about gardening, tools, sustainably sourced knick knacks etc etc etc. However, there was one single item hidden high on a shelfing unit that did make me "double take" it was a recycled circuit board, something that had failed final inspection, and instead of throwing it in a landfill, or indeed sending it with a load more electronic waste and dump it in Africa or China, someone had decided that it would make an excellent plate or mouse mat.
That moment, was my conversion, from a systems engineer working for a major UK MSP, I spent the next 6 years making my plans, I would seek to obtain a degree from the Open University, that covered the environmental issues and perhaps add development and climate change aspects to the problem, I graduated in 2010 with a BSc (Hons) Tech with environment and development (Open) degree, and started my own "Sustainable IT consultancy, Carbon3IT Ltd in August 2009 and we officially opened in Jan 2010.
The last 11 years or so have been the most satisfying of my career to date, I've travelled to places (paid by other people) that I'd never thought I ever get to, Sydney, Hong Kong, Sao Paolo, Austin, most of Europe just for starters, I walked around some interesting data centres, some great, some in need of a lot of energy efficiency attention.
Anyway, I digress, the purpose of this post is to muse on the data centre of the future.
Now, many commentators of all things data centre will go on record and state that the data centre of the future looks very much like the data centre of today and to be honest there is a lot of credence that should be paid to those that have this point of view, it is a very practical approach, and given that data centres are designed to last between 15-20 years, a valid one.
However, this is very 20th Century thinking, and we have to consider the wider aspects of energy security, climate change and to determine exactly what we are doing with our IT systems and question some of the traditional facts.
So, where can we find a "blueprint" that can make us really think about our ICT systems and how we deliver them?
The EU Code of Conduct for Data Centres (Energy Efficiency) is the best thing I have come across that can assist us in our quest.
Put simply, it forms the basis of a radical strategic and cultural change programme that will yield significant benefits across the entire corporate landscape, reducing cost, reducing energy consumption and thus carbon emissions, and combined with the 5 major ISO Management Systems, making the organisation, leaner, fitter and with the necessary control systems to deal with literally anything that can be thrown at it.
During the EURECA project, we identified the Magnificent 7, a toolbox of data centre standards, guidelines and ISO management systems that all prudent DC operators should follow that can help them acheive an ICT ECO-SYSTEM that actively reduces energy and carbon, meets business goals and most impending or current regulation and provides a roadmap to the future.
So, what are the Magnificent 7?
ISO9001 Quality Management Systems
ISO 14001 Environmental Management Systems
ISO22301 Buiness Continuity Management Systems
ISO 27001 Information Security Management Systems
ISO50001 Energy Management Systems
EN Code of Conduct for Data Centres (Energy Efficiency)
EN 50600 Series of Data Centre Design, Build and Operate Standards
The ISO Management Standards speak for themselves, the correct application of the requirements of each standard (and obviously, the integration of common areas) build a foundation of excellence.
However, and I believe that this is a major problem, too many organisations pay lip service to the "spirit" of the standard and fail to really understand what the system is for, they often see it as a necessary evil, that they only undertake "certification" to meet other procurement or tender requirements.
But, if they really understand what the Standard is designed to achieve and implemented the requirements correctly, then the organisation will be able to meet the energy and carbon legislation, reduce cost, and meet business goals.
The EU Code of Conduct for Data Centres (Energy Efficiency), provides for a radical strategic and cultural change, briefly, as the actual implementation actions that can be taken, are our USP and we charge for it! are as follows:
Management, Section 3 of the EUCOC covers Data Centre Utilisation, Management and Planning, in this section we see that the very 1st best practice is the "enabler" of all the others, it is quite simply "group involvement" in essence, gather a group of people that touch the data centre, you could call them stakeholders and have regular meetings so that everybody knows what everyone else is doing , this helps prevent those Friday night emergencies, when someone in the business has ordered loads of kit and whats it up and running by Monday morning, but didn't tell you. So, in order to fulfill this request you have to run and around and organise, new power, new connectivity and the installation of this kit.
It's in this section where we see the introduction of the "concept" of the management systems, note that I did not say ISO Management systems, merely management systems and this is because the EUCOC is NOT Prescriptive. Indeed, we mention 3 specific management systems, an Environmental, an Energy and an Asset management system, we point to the ISOs but do not specify that they should be an ISO certificated system.
We also introduce the concept of sustainable energy usage, alternative power generation technologies, management of air quality, conscise documentation, and a comprehensive training and development programme for staff.
In section 3.3 we bring into play, the concept of thinking about resilience levels, lean provisioning and part load concepts.
Section 4 is all about the IT with best practices on procurement, temperature and humidity ranges for new equipment (so, you can begin to operate at the higher ASHRAE levels and reap energy savings.)
Section 4.2 is IT deployment, software efficiency etc.
4.3 talks about the management of the existing estate, audit the physical and virtual estates, perhaps you've got "server creep" some virtual machines may have been fired up by a dev team and used for a project, that project finished 3 years ago, but you've still got these VMs cluttering up your stack. A comprehensive audit can identify targets for optimisation.
4.4 is all about the data, essentially, what have we got, where is it, and do we need it anymore? can we reduce our storage energy bill by the prudent management of that data?
Section 5, the largest section of the EUCOC contains the best practices appertaining to Cooling, and covers, Air Flow Management & Design, Cooling Management, Temp and Humidity Settings, Cooling Plant, CRAC/Hs, Direct Liquid Cooling and the reuse of DC waste heat.
6, is all about power, specifically UPS.
7 is other DC equipment, lighting, and having sensors to report back temp/humidity, energy etc to an advanced DCiM.
Section 8 is the building itself, like a feasibility or site location exercise, but not quite, covering some design/build principles, location stuff and water issues, but to be honest the EN 50600, 2-1 Building construction covers them in more detail.
Section 9 is the Monitoring section, and as we all know, you cannot manage what you cannot measure, so this section provides the detail and what, where and how you should be measuring all the data points within the DC.
Section 10 covers the best practices that will be implemented in following editions, so for instance, in the 2020 edition (v11) there is one best practice, 5.7.5 Capture Ready Infrastructure "Consider installing "Capture Ready Infrastructure to take advantage of, and distribute available waste heat during new build and retrofit projects. Section 5.7.x is the Reuse of Data Centre waste heat section.
Section 11 are Items under consideration, what this section is, is essentially a holding area for proposed best practices, pending additional information, sector feedback or the development of a new technology, so for instance, in the 2020 edition (v11) there are 3 best practices, the first is software efficiency development definitions and/or metrics this has been located in this section since the EUCOC began in 2009 and really is a souce of embarassment as the software community really haven't grasped the mettle and come up with any definitions or metrics.
Two new best practices were added to this section in 2019, both by myself, under the remit of the CATALYST Project www.project-catalyst.eu the first is Network Energy use, this is basically a request for organisations to calculate the "hidden" energy when they move to the Cloud or colocation service, if you take your equipment and locate it elsewhere, your users will expend more energy than if it was located on your site, it may not seem much, but there are two aspects to this, one is the energy cost itself (although this is borne by others (your carriers and the telecommunication operators, and secondly, to identify all the possible SPOF points. Although the internet is self-healing there are some SPOF points.
The second is to be aware of, and plan for energy flexibility or energy storage services that may arise from any Smart Grid implementations in your local area, as a large energy user, you will be in the Utilities crosshairs.
So, there we have it, the first musings on the Data Centre of the Future and my belief that we actually already have the foundations already laid, we now need to build on them and that will be the topic for the second part of my musings on the Data Centre of the Future, whcih will probably be published in late September/Early October.
