Data Centre Management March 2016

EDITION: SPRING 2016

SE

E

A 50 OV IN LIV 0 I ER SI E D ND 50 DE A U 0 F TA ST GL DAOR CE RY OB TA YO NT LEA AL S C U RE TU EN R C ANDIN UP RN TR OM D G S PLI TO E W PR SO PE ERS PA OR EHE MU AKE , GE LD NS CH RS 27 20 IVE MO 16 G R !

UI E DE TO

In association with

POWERED BY

ENERGY EFFICIENCY IN OUR MODERN WORLD

Save the planet while also minimising costs and risks

THE FUTURE IS IN YOUR HANDS Global protection against the threat of climate change

HOW TO EMBRACE A HOLISTIC APPROACH

Adopting a new route towards sustainability

PROBLEM OR SOLUTION?

Are data centres really to blame for climate change?

000_DCMCOVER_Spring2016.indd 1

ORGANISED BY

08/03/2016 17:12

002_DCM_SPRING016.indd 1

01/03/2016 22:45

Welcome

To your freshly updated Data Centre Management You may have noticed that things seem a little different around here... we felt it was time to give Data Centre Management a bit of a redesign and we’ve been working very hard to bring you this new-look version of the title.

We’ve also teamed up with the lovely people at Data Centre Alliance who will be sourcing four expert contributors to write on a dedicated theme for each issue; kicking things off in this Spring edition with some excellent articles on energy efficiency.

Look out for the next issue, which focus on physical security and data centre access control. If you would like to either nominate yourself or someone else to be an expert contributor, please get in touch with the DCA team.

Contact info

Meet our contributors 3

Peter Dyment

Alex Dorn

Peter is technical manager at Camfil Ltd. His principal areas of activity are: waste, printing, heavy industry, public and commercial buildings, particulate and molecular/gas phase air filtration, upgrading existing air systems and standalone air purification systems.

Alex is the lead engineer at Futire Facilities. He manages the engineering team, which works closely with the software developers, making extensive use of the software to provide pioneering data centre performance assessment and design consultancy services.

John Booth

Beth Whitehead

In 2009, John set up Carbon3 IT, a sustainable IT consultancy, providing information, products and services to clients looking to reduce energy costs and to minimise their IT impact on the environment. This is the perfect blend of his experience and something he is passionate about.

Beth is involved in energy and risk reduction workshops and assessments with Operation Intellignce. She has a multidisciplinary knowledge of buildings, and an in depth understanding of energy use in data centres. She has completed a PhD on the sustainability of data centres.

Editor

Production

Exhibition sales

Heather Grimes E.heather.grimes@closerstillmedia.com T. 07950 295944

Helen Blunt E. helenblunt@sky.com

Rabinder Aulakh E. rabinder.aulakh@closerstillmedia.com T. 0207 348 5770

Advertising

Head of portfolio

DCA partnership

Joanne Knowles E. j.knowles@closerstillmedia.com T. 0207 348 4902

Thomas Standley E. thomas.standley@closerstillmedia.com T. 0207 348 5273

Steve Hone & Kelly Edmond E. info@datacentrealliance.org T. 0845 8734587

Suit 17, Exhibition House, Addison Bridge Place, London W14 8XP • T: +44 (0) 20 7348 5250 • ISSN: 1753-9897 • Printed By: Stephen & George Copyright Closerstill Ltd 2016. All rights reserved. No part of this publication may be produced ina any material form (including photocopying it or storing it in a ny medium by electronic means and whether or not transiently or incidentally to some other use of this publication) without the written permission of the copyright owner except in accordance with the provisions of the Copyright, Designs and Patents Act 1998 or under terms of a licennse issued by the Copyright Licensing Agency Ltd. Applications for the copyright owner’s written permission to reproduce any part of this publication should be addressed to info@turretgroup.com

DATA CENTRE MANAGEMENT

003_Eds Letter_Spring2016.indd 3

SPRING 2016

04/03/2016 12:04

004_DCM_SPRING016.indd 4

01/03/2016 22:48

Contents

COVER STORY The schemes and guidelines you can adopt to efficiently protect your customers against threat

64. Energy efficiency How a perfect partnership can improve resiliency

66. Critical equipment Improving flexibility while also minimising loss

5 44. Critical equipment How a remotely monitored solution can tackle energy expenditure issues

48. Energy efficiency The four puzzle pieces that data centres need to join up

67. Security A white paper on IT security tackles physical hazards head on

69. Security Easy and secure access when you need it most

Our regulars 6-8

12-23

68

NEWS All the latest industry news that you need to know about

HOT TOPIC Informative articles from the DCA’s four expert contributors

COMPANY SPOTLIGHT Coreix tell us about the launch of their new Here East development

10

58

70

INSIDE SCOOP Steve Norris talks frankly about data centres and climate change

5 WAYS TO...

Five top tips on how to successfully deploy the edge network

QUIZ TIME A just-for-fun quiz to enjoy with a cup of tea or coffee

DATA CENTRE MANAGEMENT

005_Contents_Spring2016.indd 5

SPRING 2016

08/03/2016 17:14

NEWS

““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ We’ve teamed up with ““““““““““““““““““““““““““““““““““““““““ News the DCA! ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ & Views ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ So what’s been occuring in ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ our world? ““““““““““““““““““““““““““““““““““““““““ “““““““““““““ “““““““““““ “““““““““““““ “““““““““““ “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ Survey reveals “““““““““““ ““““““““““““““ customer pain “““““““““““ ““““““““““““““ points “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ “““““““““““ ““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ 20 ““““““““““““““““““““““““““““““““““““““““ n Stand J o s u e e S ““““““““““““““““““““““““““““““““““““““““ LD 2016 R O ““““““““““““““““““““““““““““““““““““““““ W E R ““““““““““““““““““““““““““““““““““““““““ TA CENT he Excel London A D ““““““““““““““““““““““““““““““““““““““““ pril at t ““““““““““““““““““““““““““““““““““““““““ 12 -13 A ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ IT““““““““““““““““““““““““““““““““““““““““ infrastructure ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ from smallest ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ to““““““““““““““““““““““““““““““““““““““““ largest. ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ The Data Centre Alliance (DCA) is delighted to be supporting Data Centre Management with the creation of a regular Expert Panel section within the new look magazine.

Each edition will drill down and focus on a different critical area of importance and interest when it comes to either building, upgrading or procuring outsourcing data centre services.

The topics have been aligned to the new EN50600 EU Data Centre Standar,d which has four primary pillars: resilience, energy efficiency, security and overall operational professionalism. This first edition is centred around ‘energy efficiency’ and includes advice and guidance on those considering both the value of investing in a more energy efficient data centre infrastructure and those who are evaluating services that are offering energy efficient solutions.

Recent research conducted by Coreix has revealed that 18.6% of customers who use data centre services will be looking to migrate their services to another provider in the coming months.

6

In Q3 of last year, Coreix conducted a survey with 110 senior executives who use data centre services in the UK, Ireland and the Netherlands. It found that there were a surprising number of pain points that they are being faced with including lack of transparency around extra charges and poor levels of support.

The survey found that 18.5% of respondents were extremely unhappy at the lack of transparency about extra charges from their colocation provider. It highlighted that customers are also frustrated that skilled personnel are often not permanently on site, they are faced with poor response times, and that additional services such as

anti DDoS solutions are often not offered.

Commenting on the findings, Paul Davies, technical director at Coreix, said: ‘The colocation sector would do well to remember that customers are increasingly astute, and service providers have to pay attention to these pain points if they wish to meet the demands of their customers. ‘

This research is highlighted in the ‘How To Guide: Choosing a Right Colocation Provider’, which can be downloaded from the Coreix website, www.coreix.net.

th

ENCLOSURES SPRING 2016

006-008_DCMNEWS_Spring2016.indd 6

Steve Hone CEO and cofounder of the DCA says: ‘I’m pleased to confirm that all members of the DCA are able to apply or can nominate someone else within their organisation to be on one of the four expert panelists.’ He went on to add: ‘Today the DCA has more than 200 organisational partners and over 800 individual registered members, which when combined, creates an incredible untapped wealth of knowledge and first hand real life experience.’ The DCM expert panel is designed to capture this experience and deliver it in a way that is both informative and of true value to customers seeking advice and guidance in their quest to find the right solution to meet their business’s needs.

th

POWER DISTRIBUTION

CLIMATE CONTROL DATA CENTRE MANAGEMENT

04/03/2016 12:06

NEWS

““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ Google envisages new ““““““““““““““““““““““““““““““““““““““““ hard disk format for ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ data centres ““““““““““““““““““““““““““““““““““““““““ “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ ““““““““““““ “““““““““““““““““““““““““““ David Purvis to power ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ operations at RMD “““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ A streamlined effort ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ Reducing the width of the platter naturally increases IOPS via reduced seek time at a cost of storage capacity, and the paper argues that the correct solution could be a combination both of the ‘stacked higher’ model proposed, and a new, smaller HDD with improved seek as a rapid-response buffer between the network and the lessaccessed storage in a data centre.

Changing the circumference of the platter is impractical, the paper argues, since greater width will increase available storage but lower Input/Output Operations Per Second (IOPS) due to the longer journey the read head must make to the next sector. Brewer’s group proposes increasing the height of the standard HDD, currently established at an average of one inch for 3.5” disks and 15mm for 2.5” drives, in order to store more platters per HDD – economical from the point of view of packaging, optimal use of printed circuit boards and the drive’s motor actuator.

ROL

IT INFRASTRUCTURE DATA CENTRE MANAGEMENT

006-008_DCMNEWS_Spring2016.indd 7

David started his career as a marine engineer with British Shipbuilders, before moving into the commercial side of engineering. His career has spanned sales and customer support roles for a number of multinational industrial businesses including Applied Chemicals, ITW ROCOL, and Bel-Ray.

Speaking of his new role, David said: I am looking forward to working with this already very successful team and helping in the guidance of taking the company onwards into new markets across the UK. Power and cooling is a sustainable and growing market, with that firmly in mind I intend to help push RMD forward in new directions in the market as well as expand alongside the opportunities presented by the continuing growths in technology.’

Google has called for manufacturers to consider developing new hard drives, intended primarily for data centre use, which abandon the traditional 3.5” dimension format in favour of taller designs.

In his research paper Eric Brewer – a professor at UC Berkeley and VP of Infrastructure at Google – argues for a revised form factor, contending that the individual hard disk is always part of a collective in data centre infrastructure, and that a larger and more capacious unit would reduce points of failure and bring in new opportunities to optimise HDD design and abandon the precepts and limitations that led to the current standard.

We have a strong base and David’s expertise will help us to forge new partnerships and develop the business further.’

Alloa-based RMD has strengthened its management team with the recruitment of David Purvis, who joins from infrared sensor specialist Pyreos where he was senior director of global sales.

With a background in engineering and more than 20 years of sales and customer support experience,, David will be responsible for planning and controlling change at RMD as well as working closely with external marketing, design and PR consultants.

Gavin Maxwell, RMD managing director, said: ‘David is a high calibre hire and I am delighted to welcome him to the RMD team.

ProLabs has announced the opening of its new integrated headquarters and test and measurement facility in South Cerney, Gloucestershire. The facility allows the company to combine its operations and management with its vigorous product testing labs, streamlining company efforts and allowing its growth to continue.

7

Over the last two years, ProLabs has gone from having just one distributor in Europe to distributing internationally – the opening of its new expanded headquarters mirrors this development. The move results from ProLabs’ considerable expansion over the last 18 months.

SOFTWARE & SERVICES www.rittal.co.uk

SPRING 2016

04/03/2016 12:06

NEWS

““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ Cardiff Snapchat data breach “““““““““““““““““““““““““““ University: “““““““““““““““““““““““““““ increasing energy “““““““““““““““““““““““““““ efficiency “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ ““““““““““ “““““““““““““““““““““““““““ “““““““““““““““““““““““““““ ““ “““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ ““““““““““““““““““““““““““““““““““““““““ Comtec Power and Schneider Electric have delivered a DCIM software upgrade for Cardiff University’s high performance computing (HPC) data centre.

The upgrade, completed in the first quarter of 2015 in conjunction with a new cooling solution, utilises Schneider Electric’s StruxureWare for Data Centers: Data Center Operations Energy Efficiency module to provide complete visibility of all power, cooling and IT assets, lowering the existing PUE level from 1.8 to 1.2.

Snapchat suffered a huge data breach recently after an employee was caught out by a phishing email scam that impersonated co-founder and CEO Evan Spiegel requesting payroll information.

While the video messaging app’s servers were unaffected and user data remained completely safe, both former and current employees were informed that some of their sensitive information had been leaked.

This in turn has halved the Universities electricity bill, saving them approximately £125,000 in electricity per year and should provide the university’s key stakeholders with a return on investment (ROI) in less than five years.

8

‘The University has a longstanding relationship with Comtec,’ said Hugh Beedie, chief technology officer for ARCCA and the general IT Services department at Cardiff University. ‘Having completed our original data centre design and build, plus the previous upgrade requirements, they were able to deploy the Energy Efficiency DCIM module quickly and without causing disruption to the data centre or ARCCA. This latest project has provided immediate energy efficiency improvements, and has repaid our trust in both them and Schneider Electric.’

Snapchat immediately reported the incident to the FBI and has offered affected staff two years of free identity theft insurance and monitoring.

comms room, server room and data centre solutions to SME’s, operating within the private and public sectors throughout the UK, Europe, and North America.

The company delivers efficient, highly available and scalable solutions, to provide resilient power, cooling, management and security for business-critical environments.

For more information please visit www.comtecpower.com.

Comtec continues to deliver high-value, resilient critical infrastructure solutions for customers ranging from high performance Formula 1 teams to global rolling stock organisations, from global exploration, independent oil and gas corporations to leading asset management businesses. The company’s experience across the data centre spectrum has continued to ensure efficient, critical uptime and peace of mind for every customer.

Established for more than two decades, Comtec Power has specialised in the design and provision of efficient

DATA CENTRE MANAGEMENT

006-008_DCMNEWS_Spring2016.indd 8

SPRING 2016

04/03/2016 12:07

009_DCM_SPRING016.indd 9

01/03/2016 22:51

INSIDE SCOOP

SENDING THE RIGHT MESSAGE

10

E G N A H C M

UM

PT

LI

NS

C

CO

AT E

ER

EN

Y

G

IO

N

When data centres come under fire for their impact on climate change, Steve Norris has a very simple message for us to repeat...

I

t seems that Microsoft are so concerned about the impact of data centre energy consumption on climate change that they are apparently going to bury their servers under the sea... Sounds intriguing although we don’t yet know what the effect will be on marine life.

Whether they get round to doing it or not, it only goes to emphasise how concerned policy makers are that, like it or not – and we can conclude we like it! – the world is increasingly driven by data in all its forms.

However, there is an ever increasing cost in terms of energy consumption and therefore on climate change. After the winter that never quite was, but interspersed by floods and huge storms, you’d be bound to sympathise.

LOUD AND CLEAR In the UK, the government is still not sure how it should treat data centres. As we are unquestionably world leaders in creative industries and financial services, both of which rely hugely on data, they are keen to maintain our lead. But I detect a tendency to see data centres as a problem rather than a solution. I’ve said this before but it bears repeating. Data centres do the job of managing energy consumption far better than the average office data room or the cellar of the Town Hall. We need to go on repeating that very simple message until government actions speak for themselves and we are no longer seen as the bad boys but rather as the infrastructure that carries this country’s future on its shoulders.

“Data centres do the job of managing energy consumption far better than the average office data room or the cellar of the Town Hall’

Steve Norris was a Conservative member of parliament for 14 years and minister for transport for five years. He had earlier served as parliamentary private secretary at the Departments of Environment, Trade and Industry and the Home Office. He retired from the Commons in 1997 and was twice his party’s candidate for mayor of London. He is chairman of Virtus Data Centres, operators of four state-of-theart colocation facilities in London. He is also chairman of BNP Paribas Real Estate and of Soho Estates. As president of the Data Centre Alliance, the data centre industry trade body, Steve plays a pivotal role in highlighting just how immensely important the data centre sector is to the ongoing sustainability of the digital age we now live in.

DATA CENTRE MANAGEMENT

010_InsideScoop_Spring2016.indd 10

SPRING 2016

03/03/2016 13:19

011_DCM_SPRING016.indd 11

01/03/2016 22:53

HOT TOPIC in association with

BENEFITS OF CLEAN AIR Peter Dyment, technical manager, discusses the importance of clear indoor air for energy efficient and reliable performance in data centres

12 We are all routinely using web-based services, either directly and/or through our personal and business activities. Data centre demand is increasing each year and is projected to increase for the foreseeable future.

Data centre buildings have different needs to other large buildings that are used principally by people. We need to focus on the needs of the data handling and storage equipment and the best means of ensuring reliable performance with

minimum running costs. Clean air at the correct temperature and humidity improves reliability and reduces running costs.

Energy consumption Power Usage Effectiveness (PUE) is the measure widely used to benchmark energy efficiency in data centre sites. The ratio of IT energy use is compared to other services using energy on the site. Historically a PUE of 2 was common but as system design has improved this has moved towards a PUE of 1.2.

A large part of the energy consumption is taken up with air management in these facilities. Moving the air, cleaning the air and cooling the air. A reduction in non-IT energy use from 100% of IT down to 20% of IT is principally the benefit of more efficient air plant.

Data centre air filter wall containing about 500 full size filters. Energy efficient F7 class filters can save up to ÂŁ300 per filter per year, in this instance it could be ÂŁ150,000 at 10p per kWh. Air filters with micro-fibre glass medias should be preferred as they have a high and consistent life performance.

DATA CENTRE MANAGEMENT

012-013_DCA Camfil_Spring2016.indd 12

The hidden cost of airborne contaminants Data centre servers generate significant amounts of heat and so require a constant draught of cool air. This warmed air has to be re-cooled for recirculation or extracted and replaced by external cooler air in a process known as free cooling. This, if available, offers the opportunity to minimise air plant running costs because large amounts of energy would be otherwise consumed by the chiller compressors needed to cool recirculated air. However, it is only possible to use

SPRING 2016

04/03/2016 10:48

Effective air filtration of supply air into the data centre is necessary to clean the air but careful selection is required to ensure as much of the harmful particle and gas contaminants are removed as possible. Current guidance advises F7 class particle filters to standard EN779:2012 are required (F7 is equivalent to MERV13). The dust particle concentration should not exceed Class 8 of ISO 14644-1 (1999). To minimise the air system fan motor energy consumption it is recommended that Eurovent A+ Energy rated filters, as defined by 4/21 2014, are used.

For harmful gaseous contaminants such as NOx, SOx, and H2S molecular gas, filters using activated carbon are recommended as they will have the required high level of filtration efficiency.

Apart from main air supply system filtration, it

Copper corrosion rate, angstroms/month

1200 1000 800 600 400 200 0 10-Aug-10

18-Nov-10

26-Feb-11

6-Jun-11

11-Sept-11

Silver corrosion rate, angstroms/month

2000 1800 1600 1400 1200 1000

is possible to use standalone recirculation air purifier units that can capture contaminants in local areas that are not adequately serviced by a main air system. These units can be fitted with the appropriate combination of particle and molecular gas filtration. Freestanding recirculation units are low energy as they have no ducting attached.

Health and safety High levels of humidity over 60% should be avoided where possible because deposited dust from air and corrosive gases will react with metal surfaces to a greater extent causing increased problems. In general, the benefit of clean air is underestimated in maintaining the working efficiency of data centre servers, air moving plant efficiency, and the working efficiency of people working inside these centres.

CONTACT CAMFIL LTD: 01706 238000 uk.filersales@camfil.com • www.camfil.com

Air monitoring It is possible to use a particle counter to measure airborne particle concentrations and copper and silver coupons to measure possible corrosive gases present in harmful concentrations. For a data centre, reliability of operation is critical and results from this study shows that for high levels of corrosion rate increased hardware failures result.

800 600 400 200 0 10-Aug-10

18-Nov-10

26-Feb-11

DATA CENTRE MANAGEMENT

012-013_DCA Camfil_Spring2016.indd 13

6-Jun-11

11-Sept-11

Recommended corrosion rate limits of 300 angstroms per month for copper and 200 angstroms

per month for silver coupons clearly fit the results of this study.

The strong link between hardware failure and high corrosion rates can be seen.

1400 1200

Corrosion rate, angstroms /month

free cooling with outside air if it is clean and free from airborne contaminants.

1000

Silver

800 600 400 200

Copper

0

29-Sep-10 18-Nov-10 7-Jan-11 26-Feb-11 17-Apr-116 -Jun-11

13

26-Jul-11

Date Figure 6

The study shows the clear benefit of using gas phase filtration to reduce data centre corrosion rates, which leads less

hardware damage and improved reliability. Data centre outages incur prohibitive costs and can be significantly reduced with careful air management.

SPRING 2016

04/03/2016 10:48

014_DCM_SPRING016.indd 14

01/03/2016 22:56

hot topic in association with

Providing global protection

In light of COP21, John Booth, managing director, looks at the schemes and guidelines that data centres can adopt in order to efficiently protect customers against threat

The historic events at COP21 in Paris culminated in a new agreement reached between the world’s governments to replace the Kyoto Protocol. Plans were put in place for 195 nations to work together on the mitigation, adaption and to support those at risk from climate change. This requires the data centre industry across the globe to provide the means for those nations to collect, collate, process and store vast quantities of data used to monitor, adapt, plan and support our villages, towns, cities and coastlines that may come under sustained threat.

We must do this at the same time as adapting and changing our own systems, using less resources and energy so that we do not become part of the problem. I’m pleased to say that in the data centre industry, our suppliers and operators have already stepped up to the mark but more is needed. The data centre industry has a wide array of white papers, standards, guidelines, certification schemes, etc, to help provide the most efficient digital services to the customers, here a few of the more important...

Institute. The EUCOC best practices will be known as a Technical Report under EN50600 and will be contained in part 2-3, the report will be titled EU50600 2-3 TR99-1.

In effect, there will be two versions of the same best practices, albeit in different formats and titles. The participation scheme, which allows organisations to carry the EU Logo on their marketing materials is not affected by the addition of the EUCOC best practices into the EN50600, however many organisations have adopted the EUCOC but have not become participants in the scheme and it is recommended that such organisations participate in the future.

15

EUCOC Since 2008, the European Union Joint Research Centre (EU-JRC)’s EU Code of Conduct for Data Centres (Energy Efficiency) (EUCOC) created in partnership with DEFRA and the BCS, Chartered Institute for IT, has provided a series of best practices covering management, IT equipment, cooling, power, design, and the monitoring of data centres. These are developed by the industry for the industry, are vendor neutral, and are reviewed every year.

The EUCOC best practices for 2016 have also been adopted by CEN/CENELEC/ETSI, the European Standards Authority, reformatted and incorporated into EN 50600. This is an EU regional standard regarding the design, construction, implementation and management of data centres and is the Euro equivalent to TIA942 or BICSI 002, which are US standards developed by their equivalent to CEN/CENELEC/ ETSI the ANSI, or American National Standards

Data Centre Management

015-016_DCA Carbon 3IT_Spring2016.indd 15

Spring 2016

03/03/2016 13:29

hot topic in association with

ISOs Our international cousins have various guidelines and best practices, the most important are the International Standards Organisation’s (ISO) these are as follows: ISO31032, ISO 30133, ISO30134-1, ISO30134-2, and InISO 30134-3.

These Standards are due to, if they haven’t already, be published very shortly.

EURECA Project The Horizon2020 funded EURECA Project (www. eureca-project.eu) – a consortium comprising the University of East London, Data Centre Alliance, Green IT Amsterdam, Carbon3IT Ltd, Certios, Maki Consulting, Telecity and CBRE Norland – has reviewed all the data centre standards globally, and has constructed a definitive list of current best practice covering all related aspects, including energy efficiency.

This project is designed to give public-sectorfunded organisations advice and guidance to make better data centre/server room/equipment room procurement decisions and allows five days of free consultancy for qualifying organisations.

16

One of EURECA’s activities is to recommend potential policy development to the EU. While nothing has been finalised at present, one idea is to have a mandatory list of ISO’s or other guidelines that must be in place for public sector data centres; this will include cloud or outsourced digital service contracts in third party data centres. Thus energy efficiency will be a decisionmaking factor in public sector procurement contracts and it will be incumbent on those third party data centre organisations to be compliant with the policy.

The EURECA tool is being designed to allow public sector procurement and technical staff to provide data that applies to their facility with a view to providing technical and potentially financial output so to what they should do next, be it a refurb, outsource or new build. If your organisation is publically funded and is seeking assistance with a data centre, this might be your EURECA moment!

DCMM The Green Grid, which comprises many members that originate within the EU, is reviewing its Data Centre Maturity Model (DCMM), which was first created in 2010 and was designed to provide a thought-provoking roadmap for the data centre five years away. Unhappily, many facilities are still five years away, although some would say 10 or even 20 years away for some older facilities.

However, the DCMM is a very useful tool, as it allows organisations to benchmark where they are now and to see where they could be in 5/10/20 years time.

Other Standards that are of interest and useful in our quest to be energy efficient, include: ISO9001:2015 and ISO14001:2015. Some technical standards that we think should be included in all data centre accreditation plans are ISO22301:2012 and ISO27001:2013.

Finally, but not the least, is the ISO 50001:2013, which defines how you can adopt good management practice perhaps using the EUCOC and DCMM in a defined structure to continually improve your use of energy across the organisation.

You can of course get certified to the ISOs listed above and these are well worth having on your marketing and sales literature as they define professional status and approach that is internationally recognised.

‘It makes sense to ensure that your organisation’s ICT systems are backed

by IT companies that are seen to manage threat’

Formal energy efficiency accreditations The last area is formal energy efficiency accreditations, the first being the DCA’s certification system. This is aligned with the EN50600 and covers resilience or class level, security aspects, energy efficiency aspects and operational management aspects. There are currently two data centres certified in the UK, with more in the pipeline.

Another accreditation is the Certified Energy Efficient Data Centre Award (CEEDA), this is based upon the EUCOC, DCMM and uses the Green Grid metrics and is ratified by the BCS, Chartered Institute for IT. Obtaining any externally recognised accreditation, certification or award is always good, but when the world is looking for ways to mitigate and adapt to the very real threat of climatic change, it makes sense to ensure that your organisation’s ICT systems are backed by IT companies that are seen to manage the threat.

CONTACT Carbon3IT: 0019 2684 3835 info@carbon3it.com • www.carbon3it.com

Data Centre Management

015-016_DCA Carbon 3IT_Spring2016.indd 16

Spring 2016

03/03/2016 13:29

017_DCM_SPRING016.indd 17

07/03/2016 13:03

HOT TOPIC in association with

Your challenge, if you choose to accept it...

18

...is to strive to offer efficiency without increased risk, say Alex Dorn and Mark Seymour People like efficiency; it remains a modern word.

It is associated with positive things, like high productivity and saving the planet. Striving for efficiency means putting the least in to get the most out. Who doesn’t want that? It’s no different in the data centre industry. Improving efficiency leads to cost savings, better management practices, and comes with a fashionable green badge. In any case, whether financially justifiable or not, legislation will make it a necessity.

Energy efficiency for a data centre means using as little power as possible to provide the IT services.

Historically, focusing on reducing infrastructure energy – primarily cooling – has created the industry measure Power Usage Effectiveness, or PUE. PUE tells us how much power the data centre needs compared with the requirements of the IT alone. So to improve efficiency we must target both IT and infrastructure. The processes are simple: electricity is supplied through a power delivery system, computers turn the electricity into heat, the cooling system removes the heat from the computers. Yet despite this simplicity, efficiency in the data centre is synonymous with risk. When making changes, a lack of insight can result in poor implementation, leading to reduced or non-existent operational margins. Alternatively, careful design can result in both increased savings and margins. For example, it’s possible to save significant energy by increasing cooling

DATA CENTRE MANAGEMENT

018 020_DCA Future Facilities_Spring2016.indd 18

temperatures, while at the same time making your servers cooler by taking measures to prevent recirculation. You can be a hero: offering efficiency without increased risk.

The design challenge The principle of energy efficiency can be built in at the design stage, and many features can be added in an upgrade, but they must be implemented with diligence:

‘Improving efficiency leads to cost savings, better management practices, and comes with a fashionable green badge’

• Power - Power from renewable sources - Modern PDUs and UPS systems minimise conversion losses and offer high efficiency for lower utilisations - Smart PDUs and/or SNMP monitoring show you where energy is being used, which is invaluable when optimising cooling - Choosing DC power can reduce conversion losses - In the future, fuel cells may provide a less environmentally damaging power source.

• Cooling - Free cooling (but often consumes water): direct, indirect, air based, or water based. Sometimes combined with traditional cooling to raise efficiency - Liquid cooling (e.g. rear door heat exchangers or

SPRING 2016

04/03/2016 12:08

019_DCM_SPRING016.indd 19

01/03/2016 22:59

HOT TOPIC in association with

in the IT itself): High capital cost, but an efficient transport mechanism suited to high density loads. However, typically a lot of heat is still removed by air, and these devices introduce cool air into the ambient; this may detrimentally affect data centre efficiency if mixed with other cooling.

• Air management - Containment segregates hot and cold air, but leakage can occur. To avoid this requires careful management and sophisticated controls to balance cooling air supply and IT demand, and to avoid excessive fan energy - Pressure or temperature controlled variable speed fans help match air supply to demand.

Beware: the flow profile may change with fan speed and may not suit your existing IT layout.

• Heat recovery - Unless you are next to a swimming pool or district heating system, this is confined to the future.

20

Whatever you choose, you must be confident your design will operate as intended. If it doesn’t, your efficiency aspirations won’t be met and you risk impaired performance. Engineering simulation and modelling tools to assist with design can be used to test your proposals before breaking ground.

The operational challenge Data centre environments contain many technologies. When looking at options, it’s important to remember you only need to optimise the three basic processes. Use your power responsibly! Keep phases balanced to minimise waste. For backup, consider high efficiency mode on your UPS if it suits your business needs.

The most obvious target is to reduce the compute energy required to provide your IT services. This is not news; using efficient IT helps minimise the energy requirement for a given workload. Keeping IT energy usage as low as possible through technology refreshes is a smart policy, because all other energy requirements in the energy chain are proportional to your IT power. IT equipment efficiency can still be improved by the IT vendors, especially consumption at low utilisation. More product efficiency information should be made available to allow people to make informed decisions when selecting hardware. Requesting this information now will make your life easier in future.

to maximise utilisation and minimise the number of powered up devices.

Standardising IT deployments, can result in greater efficiency. However, redundancy must be built in somewhere.

A data centre with a PUE of 1.1 is impressive, but what if a second one is needed to provide redundancy? While it is a valuable metric, PUE is limited in its scope. Future effort must be extended to include evaluation of the enterprise, continuing reduction of IT power requirements, reuse of waste heat, and more environmentally friendly power production.

Some things to consider • Design options mentioned previously • Operation under cooling maintenance or partial failure • Floor void obstructions • Revising control strategy • Internal rack airflow management • Cable and cable penetration management • IT deployment.

Whichever cooling system you use, the vast majority of designs will still rely on air in some part. It’s up to the data centre owner to ensure effective operation; this means delivering sufficient cool air to every server to remove the heat generated. Doing this efficiently is difficult, because you need to minimise or eliminate the mixing of hot and cold air, which you cannot see. No mixing means no recirculation and no bypass, so your server inlet temperatures are the same as your cooling supply. As a consequence, you don’t have to reduce supply temperatures to compensate and this is where the efficiency gains benefit you most. Higher air (and water) temperatures mean chillers operate more efficiently, or free cooling can be used for longer periods of time. For these reasons EN 50600 specifically recommends undertaking an analysis to find an efficient approach to environmental control that suits the business needs. This can be done using engineering simulation to visualise the otherwise invisible airflow.

Communication is key In summary, data centre efficiency has improved dramatically with the introduction of free cooling and improvements in power systems, IT equipment and cooling. What’s more, when designed around each other these systems offer performance not otherwise achievable, but doing this requires all stakeholders to communicate. There is more to come: fuel cells, liquid cooling, and heat recovery.

Adopting established new approaches – and potentially techniques not yet generally employed – will continue the trend of improved efficiency.

However, all of these approaches (except perhaps if your data centre is underwater!) will still dissipate heat into the air. Failing to consider the operational challenges that come with this, ahead of time, will not deliver all the efficiency available to the business, and could even put mission critical operations at risk.

Virtualisation offers more efficiency gains with no added risk. Multiple operating systems can run on a shared hardware, which typically reduces the number of servers you need and reduces your number of single points of failure. Look out for advanced techniques for application deployment CONTACT FUTURE FACILITIES: 020 7840 9540 info@futurefacilities.com • www.futurefacilities.com

DATA CENTRE MANAGEMENT

018 020_DCA Future Facilities_Spring2016.indd 20

SPRING 2016

04/03/2016 12:09

021_DCM_SPRING016.indd 21

07/03/2016 13:05

ADVERTISMENT FEATURE

Prepare for the unexpected

Energyst CAT Rental Solutions explains why a customised, proactive approach minimises disruption while maximising productivity 22

In today’s global environment, businesses must prepare for the unexpected. Whether a manmade or natural disaster, the consequences can be devastating on a human and financial scale. The adage ‘fail to prepare, prepare to fail’ applies here and is particularly important to critical environments, whether for your Factory Acceptance Test (F.A.T) or Site Acceptance Test (S.A.T), or providing necessary contingency to support a shutdown for ongoing maintenance.

For any commercial operation, CAT Energyst aims to reduce unnecessary downtime caused by a disruption to their power supply, thereby negating the effect on their customers and employees.

Energyst provide a long-term, proactive approach to business continuity with personalised, turnkey rental solutions for power and cooling with built-in contingency. Delivered by people who genuinely care about their customers and constantly look for ways to make a difference.

‘Our engagement with customers, the will and ability to immerse ourselves in their situation, is unparalleled,’ said Gary Smith, Energyst CEO. Energyst’s relationship with CAT and other partners has resulted in a global presence across a wide range of sectors including oil and gas, utilities, data centres, infrastructure, hospitals and shipping.

Examples of use include load banks for both low voltage (LV) and high voltage (HV) requirements

that enable the client to prove the system in varying market sectors. CAT Energyst offer energy-efficient generators with ultra-low emissions that reduce NOx and CO2 emissions.

Personal approach As an example of Energyst’s personalised approach, a recent project for Finning, across the Home Counties, addressed issues with F.A.T. testing.

Three CAT C175 Engines 3.3MVA, each requiring a 10MVA test in total at 11KV benefited from the HV capability and know-how of Energyst and their brand partners CAT. After a thorough consultation process, Energyst quickly and efficiently provided a robust 10MVA 11KV Load Bank package, complete with HighVoltage Switch Gear Container and Auxiliary Generator, alongside over four kilometres of 240mm cable.

‘Delivered by people who genuinely care about their customers and constantly look for ways to make a difference’

The 10MVA package was commissioned and delivered at F.A. test and followed by the S.A.T to prove continuity of supply in the event of a mains power outage. This vigorous approach from Energyst ensures that every client has the resilience and re-assurance of the generator’s full capability.

This project was successfully delivered with the combined mobilisation efforts of Finning and Energyst working through a three-week window to commission and install the desired result.

CONTACT: Energyst CAT Rental Solutions 0845 769 7450 • info.uk@energyst.com • www.energyst.com

DATA CENTRE MANAGEMENT

022_Energyst ADV_Spring2016.indd 22

SPRING 2016

03/03/2016 13:32

HOT TOPIC in association with

On the right track

Beth Whitehead, associate sustainability engineer, looks at how data centres should take a holistic approach to reach the sustainability finish line

Operational Intelligence believes that energy efficiency is not just about reducing facility bills, but that it also plays a role in reducing environmental impact and improving sustainability.

Past The data centre industry has experienced rapid growth over the last couple of decades, and with it so too has its energy consumption. In response to rising energy costs, power usage effectiveness (PUE) was created to help owners and operators rationalise this consumption. It has, however, had both a positive and negative effect on the industry. PUE has, in very simple terms, helped make energy efficiency a high profile subject and has paved the way to a growing understanding of the industry’s impact on the environment.

However, it is used frequently as a proxy for sustainability. Sustainability and the green economy are about growing in a way that improves energy and resource efficiency, and reducing emissions and impact on the environment. Energy efficiency is only part of sustainability and therefore should not be considered in such an isolated manner. PUE has got

the industry out of the starting blocks but until a more holistic approach is employed, sustainability is a long way off.

Walk before you run The industry understands that the energy consumption in operation needs to be reduced, but there is still a long way to go to eliminate wastage.

PUE has a well-documented list of shortfalls, not least that it is frequently incorrectly calculated. Currently there is a lot of discussion about sustainable data centres, but cutting through the green wash to understand whether savings are being made can be difficult.

‘Green’ technologies are often designed into new builds and retrofits, but their operation needs to be in line with the design intent for these operational savings to be realised.

Data centres need to walk before they can run, and often little or no effort is made to manage air flows in the facility.

For example, much energy is wasted when air from CRAH (computer room air-handling) units bypasses the IT equipment, returning directly to the air handler without cooling any equipment.

This wastage of supplied air means that servers draw hot exhaust air (recirculation) to fulfil their required airflow volumes. IT inlet temperatures rise as a consequence, and CRAH setpoints are often reduced to compensate. The result is higher energy consumption and inefficiency.

Appropriate air management minimises this bypass and recirculation, enabling an environment optimised to allow for the potential raising of set points and lowering of fan speeds. Risk has also been lowered – IT equipment inlet temperatures are managed and controlled within an appropriate range with hot spots eliminated.

DATA CENTRE MANAGEMENT

023-024_Operational Intelligence Socomec_Spring2016.indd 23

Operating at higher temperatures through air management and exploiting wider operating ranges for IT equipment (defined by ASHRAE in conjunction with hardware manufacturers) increases opportunities for using free cooling technologies. In many locations, particularly in the UK, it is possible to design and operate zero-refrigeration solutions.

There are often missed opportunities for energy efficiency due to a lack of awareness of the issues and poor communication between different stakeholders. For example, coordination between IT and facilities teams to optimise air delivery and operational parameters. The EU Code of Conduct for Data Centre Energy Efficiency provides a list of best practices that can be used to help make improvements in all areas of the data centre.

23

A holistic approach A focus on operational energy efficiency is likely to continue, but for the industry to embrace sustainability it needs to understand that it is not a stand-alone matter. The industry needs to think of its consumption in terms of the energy and resources inputted throughout the entire life time of a facility – from the extraction of raw materials, through manufacturing, transportation, construction, operation and eventual end-of-life – and the resulting emissions to the environment. Using such a holistic approach will stop the use of ‘end-of-pipe’ solutions (such as PUE), which focus on one issue (energy consumption) or one life cycle stage (operation) and will ensure burden shifts don’t go unnoticed. Energy efficiency is incredibly important, but it should not be at the detriment of other areas of sustainability.

SPRING 2016

03/03/2016 13:36

HOT TOPIC in association with Publication of documents such as The Green Grid’s Data Centre Life Cycle Assessment Guidelines and inclusion of life cycle assessment in the Data Center Maturity Model show that key thinkers believe in this holistic approach, but there are very few mechanisms available to make it easily applicable. It is also time consuming to collect data

and build models that consider a whole array of life cycle stages and issues.

Research therefore needs to inform the industry to facilitate a broadening of concern. Current research, using a life cycle perspective, shows that beyond the operational efficiency of cooling, the industry needs to consider: • Energy consumption of IT and cooling • The source of this energy • The materials used to manufacture IT equipment and their refresh rates and the embodied impact of mechanical and electrical services.

There needs to be more attention paid to the 1 of PUE, not just the operating efficiency of cooling. IT equipment often operates with poor utilisation and high power consumption when idle. By considering these impacts we can begin to build and operate data centres that are more energy efficient and sustainable.

CONTACT OPERATIONAL INTELLIGENCE: • info@dc-oi.com • www.dc-oi.com

24

POWER

Fast, flexible & futureproof

Addressing evolving demands of managing data centres, critical buildings and facilities With designed-in flexibility, Socomec’s latest development ensures that every hard working electrical infrastructure can protect its organisation’s vital assets while demonstrating real adaptability; quickly and easily responding to the constantly changing challenges of power and efficiency.

multiple applications. Easy to integrate and install, while simple to manage and maintain, it provides maximum availability and power protection in a compact design; leaving space for other rack-mounted devices.

The latest modular, rack-mounted UPS development addresses short-term capacity issues while meeting longer-term growth requirements – delivering reliable, safe, high quality power to the most demanding critical applications, that can be scaled to meet demands.

Modulys RM GP has been engineered with full flexibility and fewer parts to simplify and optimise every step of the integration process, de-risking the entire project.

Rack-mounted modular UPS Socomec’s Modulys GP is a three-phase modular UPS system designed for 19” rack integration across CONTACT SOCOMEC: 01285 86 33 00 info.uk@socomec.com • www.socomec.co.uk

DATA CENTRE MANAGEMENT

023-024_Operational Intelligence Socomec_Spring2016.indd 24

As a completely modular system – designed with no single point of failure – Modulys RM GP delivers reliable power while ensuring optimum load protection, even during power upgrades or maintenance procedures.

To learn more visit Socomec on stand E65 at Data Centre World.

SPRING 2016

03/03/2016 13:37

025_DCM_SPRING016.indd 25

01/03/2016 23:01

026_DCM_SPRING016.indd 26

01/03/2016 23:02

027_DCM_SPRING016.indd 27

08/03/2016 12:02

028_DCM_SPRING016.indd 28

08/03/2016 12:04

029_DCM_SPRING016.indd 29

08/03/2016 12:05

030_DCM_SPRING016.indd 30

08/03/2016 12:06

031_DCM_SPRING016.indd 31

08/03/2016 12:07

032_DCM_SPRING016.indd 32

08/03/2016 12:07

033_DCM_SPRING016.indd 33

08/03/2016 12:08

034_DCM_SPRING016.indd 34

08/03/2016 12:08

035_DCM_SPRING016.indd 35

01/03/2016 23:03

036_DCM_SPRING016.indd 36

08/03/2016 12:09

037_DCM_SPRING016.indd 37

08/03/2016 12:10

038_DCM_SPRING016.indd 38

08/03/2016 12:10

039_DCM_SPRING016.indd 39

01/03/2016 23:06

040_DCM_SPRING016.indd 40

08/03/2016 12:11

041_DCM_SPRING016.indd 41

08/03/2016 12:12

042_DCM_SPRING016.indd 42

08/03/2016 12:12

043_DCM_SPRING016.indd 43

08/03/2016 12:13

critical equipment

Optimise your cooling systems

44

Tackle your energy expenditure issues with a 24/7, remotely monitored solution from Birdsall Energy expenditure and efficiency is a big issue today and unlikely to recede anytime soon. For data centres, by their very nature big consumers of power, the issue is even greater. Of course, there’s no shortage of product improvements with more efficient models readily available.

a year. The company goes beyond the norm of monitoring flow and return temperatures as highlighted by a BMS. Instead they suggest fully interrogating the cooling system constantly for any drop in performance or efficiency using a Swedish technology called ClimaCheck.

Some are an easy win, especially if the capital cost is low but some require a significant investment to be able to enjoy the benefit of greater efficiency of the latest product.

Here’s how it works A ClimaCheck is connected and ‘Typically when fixed to the cooling unit with a modem installed to send data to optimising the cooling plant the cloud-based software. The in this manner, efficiency ClimaCheck technology is based can be improved by on the fundamental energy laws and thermo-physical data of between 10-40%’ refrigerant cycle cooling systems.

Multiple temperature sensors, pressure transducers, current transformers and voltage tappings are used to diagnose the performance in the refrigeration process. These measured results identify opportunities to optimise cooling performance and system efficiency.

In the case of data centres, cooling is one of its largest consumers of power, and while each year more efficient technology becomes available, it rarely makes commercial sense to replace existing cooling systems until their effective working life is run, for the payback just seems too long.

Birdsall’s compelling solution is to optimise your cooling technology, constantly, 24/7, 365 days

The data recorded second by second, includes

Data Centre Management

044-045_Birdsall_Spring2016.indd 44

Spring 2016

08/03/2016 11:59

CRITICAL EQUIPMENT power consumption, cooling performance, thermal load of the system, as well as chilled water temperatures and flow rates along with refrigerant conditions such as sub-cooling at the expansion valve inlet and superheat at the compressor inlet.

ClimaCheck is unique in providing this data based on the power consumption of the compressor as well as the refrigerant thermodynamic cycle as measured on site, without the need for refrigerant or water flow meters.

The early warning system can detect: • Leaks (high superheat combined with low sub cooling) • Blocked condenser coils (abnormal air on and air off condensing temperatures) • Failed condenser fans (condensing temperature high with low auxiliary power input for fans) • Inadequate water flow rate (risk of freezing evaporator with low water out temperatures) • Compressor damage (liquid return, low efficiencies, excessive discharge temperatures) • Motor burn outs (high discharge temperature and excessive power consumption, high superheats).

What is optimised?

With the data provided, Birdsall engineers can optimise the refrigerant charge, the expansion valve superheat, the control parameters of electronic expansion valves, adjust head pressure-low ambient control settings, control dead bands and much more.

Typically when optimising the cooling plant in this manner, efficiency can be improved by between 10-40%. For a data centre operating 24/7, that’s a lot of energy and cost. Equally important is the increase in reliability. By seeing any slight alteration in performance alerts Birdsall to a change that can be investigated before it becomes a fault, thus avoiding downtime and repair costs.

45

Remote monitoring Birdsall operates a 24/7 Energy Bureau. They set thresholds on the ClimaCheck so that when a limit is reached Birdsall would receive a message, enabling them to log-on remotely to investigate and if required despatch an engineer.

The financial benefits The financial savings to be achieved in power consumption alone make this an attractive proposition; saving between 10-40% of energy each year is a significant cost saving.

Birdsall provides both ownership and rental options to enable all budgets to use this technology with an Energy Bureau providing 24/7 service. The company can also provide comprehensive and semi-comprehensive maintenance package options that results in the data centre reducing its overall cooling costs by cutting energy, maintenance and running costs.

Visit Birdsall at DataCentre World 2016 on stand D42 at ExCel in April for a demonstration.

CONTACT BIRDSALL: 01442 212501 info@birdsall.co.uk • www.birdsall.co.uk

DATA CENTRE MANAGEMENT

044-045_Birdsall_Spring2016.indd 45

SPRING 2016

08/03/2016 11:59

ENERGY EFFICIENCY

Setting a Trend

Be fashion forward by building resilience into your energy management systems Visit stand M46 at Data Centre World, where Trend’s experts will be on hand to talk about how it can make today’s data centres more resilient, compliant and sustainable. Trend’s innovative building energy management systems (BEMS) based technology can help data centres operate more efficiently and maintain the high levels of uptime and energy efficiency that owners, operators and their customers expect. The recently launched IQ4E controller range will be on display during the event, with its latest additions that even further optimise panel space and enhance cost effectiveness.

Build resilience Data centres can benefit significantly from building resilience into their BEMS and minimising any risks and significant costs associated with situations such as plant failure or environmental conditions

falling outside acceptable parameters.

This leads to the development of a sustainable, long lifecycle solution that also ensures compliance with key performance indicators (KPIs) and helps co-location facilities remain competitive. Trend will demonstrate why a well-designed BEMS is crucial for any organisation that has business critical systems to protect, ‘Data centres can benefit and show how its systems can monitor significantly from building and manage data centres intelligently in resilience into their order to deliver the lowest power usage effectiveness (PUE) ratings.

BEMS and minimising

Trend enjoys a large, installed system base, with the vast majority of its highly reliable systems being supplied, engineered and commissioned by 160+ approved systems integrators, a number of which specialise in resilient environments such as data centres, healthcare facilities and CONTACT TREND: 01403 211888 telecommunications. marketing@trendcontrols.com • www.trendcontrols.com

any risks and significant costs associated with situations such as plant failure or environmental conditions falling outside acceptable parameters‘

46

046_Trend_Spring2016.indd 46

03/03/2016 13:42

047_DCM_SPRING016.indd 47

07/03/2016 13:06

ENERGY EFFICIENCY

Complete your puzzle

Every data centre has its own unique set of needs but here we break it down into four distinct pieces...

Data centres are becoming ever more strategic to the always on global enterprises they support.

Today’s leading operators view them as coordinated and optimised facilities that are built to be intelligent, highly efficient and increasingly flexible. Every data centre has its own set of needs and these can be broadly categorised into four areas: critical power, intelligent grid connections, power distribution, and data centre infrastructure management.

Critical power Critical power systems ensure that data centre customers have continuous access to their data.

Products such as uninterruptible power supplies are built to withstand the rigors of data centre operation and ensure many years of reliable, trouble-free service, such as modular UPS, genset subsupply and integration, and energy storage systems.

48

Intelligent grid connections Many operators demand their own grid connection to meet growing demand for power as well as greater levels of redundancy. ABB’s equipment, systems and expertise mean that it can reliably create new connections at up to 132 kV and its turnkey grid connection capabilities provide better management of energy demand, optimal utilisation of renewables and power assets, integration into smart grids and microgrids, and substations.

Power distribution This includes the products, systems and expertise required to distribute power reliably and safely. It also keeps operating costs down through improved electrical efficiencies. ABB supplies not just individual products but also develops and implements whole systems, from power delivery and distribution to mi-crogrids and automated monitoring and control.

This leads to data centres that more readily attain their operators’ performance goals and includes aspects such as simple maintenance, enhanced reliability, increased safety performance and reduced likelihood of human error: alternating current (AC) power systems, direct current (DC) power systems, a wide range of low and medium-voltage products, and distribution substations.

Data centre infrastructure management ABB’s Decathlon system is a tool to automate power and energy management. It allows intelligent planning of assets and capacity and incorporates alarm management, remote monitoring and other key data centre functions. Its benefits include system availability and performance, capacity planning and management, resource forecasting and energy planning, facility and IT automation, and troubleshooting and root cause analysis.

Join up the pieces of your puzzle with a visit to stand H72 at Data Centre World 2016, where ABB will be highlighting its specialist products, systems and complete experience.

CONTACT ABB Electrification Products Division UK: 0333 999 9900 • Ennquiries@gb.abb.com • www.abb.com

DATA CENTRE MANAGEMENT

048_ABB_Spring2016.indd 48

SPRING 2016

03/03/2016 17:26

SOLUTIONS FOR DATA CENTRES Cabling systems | Fibre products | Fibre containment

www.splicegroup.com SpliceGroup Europe: +44 (0)113 285 2536

049_DCM_SPRING016.indd 49

01/03/2016 23:08

CRITICAL EQUIPMENT

Near-perfect availability

Mike Elms, technical manager, looks to solve a power protection conundrum

50

Who heals the healer? In power protection terms, this conundrum is about recognising that while UPSs are essential for protecting critical loads from power problems, their success depends on their own fitness for purpose. With the continuously increasing ubiquity of online transaction processing, IT system availability is more than ever of primary concern for operators, and power availability is its largest single component.

Power availability is a measure of how much time per year an IT system has acceptable power. It is also the ratio between Mean Time Between Failures (MTBF) and Mean Time To Repair (MTTR). Its value can be improved both by maximising MTBF and by minimising MTTR times – modern UPS systems use techniques that exploit both of these.

Earlier UPSs tended to use transformer-based designs, resulting in large, floor-standing units. The advent of transformerless technology allowed manufacturers to develop smaller, more modular designs, which offered several advantages. Users could specify the UPS capacity they needed far more accurately. Modular design also supports future growth with equal efficiency, as modules can be added according to changing needs.

Parallel redundancy Another important advantage of modular design is its efficient support of redundant configurations. A UPS comprising five 10 kVA modules supporting a 50 kVA load is referred to as a capacity system. It operates without any problems, provided all five modules remain healthy and deliver full power. However, if one should fail, the other four CONTACT UNINTERRUPTIBLE POWER SUPPLIES: 0800 731 3269 • sales@upspower.co.uk

DATA CENTRE MANAGEMENT

050_UPS_Spring2016.indd 50

will have to supply 12.5 kVA each and would be overloaded. The load may be transferred to bypass, exposing it to raw mains power and all its potential aberrations. To prevent this, a redundancy system can be set up, comprising at least one more module to fully support the load. This makes the system resilient to the failure of a single module. Redundant systems offer significantly improved availability compared with capacity implementations.

We have assumed that the UPS topology is ‘decentralised’ – each module operates as a complete, independent unit with no components shared by all the modules. However, some systems are available with a more centralised architecture, in which all modules connect to the load through a shared centralised static switch (CSS).

The advantage is reduced costs, but the problem is that the CSS represents a single point of failure, which adversely affects the system’s availability.

Modular rackmounting systems The highest possible availability is achieved by using a modern, modular rackmounting system as shown in Figure 1. Firstly, these provide true Decentralised Parallel Architecture (DPA) as each module is fully self-contained. Additionally, many such systems offer a ‘hot swap’ capability, which also improves availability. If a module in a hot swap system fails, it can be removed and replaced with a working unit without having to put the system into maintenance bypass, or interrupting power to the load. Eliminating these requirements or having to effect repairs to equipment while within the UPS means that MTTR times can be reduced to minutes, compared with the repair time of hours, typically needed for monolithic systems.

Figure 1 UPS Ltd PowerWAVE9500 rackmount modular system Reducing MTTR contributes to improving availability just as significantly as increasing MTBF. Rackmount modular systems with DPA architecture, a redundant configuration and hot swap capability can nowadays achieve up to 99.9999% availability.

This is highly valued by IT system operators seeking zero downtime.

SPRING 2016

04/03/2016 12:10

DATA CENTRE MANAGEMENT

051_DCM_SPRING2016.indd 51

SPRING 2016

03/03/2016 17:38

cloud & storage

Addressing the performance/power trade-off Helping data centres meet customer demand

52

Anybody involved in building and maintaining a modern data centre will be under pressure to deliver increased capacities and ever-higher IPOS while keeping power consumption and cost to a minimum. Managing the trade off between performance and power is only going to become more important as companies seek to drive down TCO and demonstrate their green credentials while also addressing user expectations for storage and response time.

And while the majority of digital data is still stored on traditional hard disk drives (HDDs), solid-state drives (SSDs) based on NAND flash memory technology – which allows SSDs to operate with faster data access times and lower energy consumption than HDD alternatives – are an increasingly important part of the data centre storage toolkit.

The next level Toshiba’s enterprise-grade SSDs have evolved rapidly in recent years in terms of both storage capacity and IOPS performance, as can be seen by its third generation of eSSDs, the PX04S Series.

This is Toshiba’s first family of eSSDs to offer up to 4TB of operating capacity.

In addition, because the drives are equipped with dual-port 12.0 Gbit/s SAS interfaces, they can deliver sustained IOPS read performance up to 270,000 – the highest for a 2.5-inch form factor SAS SSD. Flexibility In developing the PX04S Series, Toshiba has looked to provide customers with the optimum level of flexibility: a unique customer-tunable power and performance feature allowing optimisation for power efficiency or maximum performance. Users

Data Centre Management

052_Toshiba_Spring2016.indd 52

can choose to operate the SSD with either ‘performance’ or ‘economic’ power settings. This keeps power consumption as low as possible during normal operation but allows the write performance to be boosted by as much as 47% when it is needed. This ability to re-configure the power mode allows data centre managers to accurately match performance to demand.

DWPD and capacities up to 3.2TB. The PX04SVB value-endurance drive is available with capacities up to 3.84TB and has been optimised for data warehousing, media steaming and other read-intensive server and storage applications that require a balance of reliability, capacity and endurance.

The importance of flash Solid-state flash memory plays a fundamental role in the performance, reliability, endurance, power consumption and cost of an eSSD drive. Toshiba has complete control over both the NAND flash and the controller technology on which its eSSDs are based. This vertically integrated design and manufacturing capability ensures that not only can Toshiba designers develop industryleading SSDs, but also offer highquality customer support ranging from identifying the most appropriate solution to post-sales engineering back-up.

Flexibility is also the underlying principle for the four different PX04S drive options that Toshiba has developed. Available with capacities up to 1.6TB, the highendurance PX04SHB drive, for example, is targeted at write-intensive, virtualised data centres that require the highest levels of eSSD performance, reliability and endurance. These devices support 25 complete drive writes per day (DWPD) with a 100% random workload.

The PX04SMB Series is optimised for mixed-use, mission-critical hyperscale and virtualised CONTACT toshiba: 00800-THOTLINE (84685463) environments, spdinfo@tee.toshiba.de • www.toshiba.semicon-storage.com offers 10

Spring 2016

02/03/2016 14:15

CRITICAL EQUIPMENT

A NEW COOLING CONCEPT

Low Speed Ventilation is a new, highly eficient cooling concept for server rooms

Alfa Laval has introduced a unique and patented cooling concept, Low Speed Ventilation(LSV), which delivers air at a considerably lower speed: 1.5 to 1.8 metres per second. Reducing air speed and delivering an optimum airflow eliminates the Venturi effect, minimises pressure differences and prevents problems such as ‘hot spots’. As LSV is a normal pressure technology, data centre operators only need to control air availability, not air pressure, making it far easier to achieve optimum conditions in the server room.

As significant amounts of air are required for LSV, the coolers are large. Alfa Laval Arctigo LSV air coolers are specially designed with a large cross-sectional area. Located outside the server room, these coolers ensure the servers are supplied with sufficient temperature-controlled air at all times, while consuming far less power than conventional server cooling solutions.

Key benefits of LSV • Lowest PUE – Power Usage Effectiveness of 1.07 compared to market average for global data centres of 1.5 • Lower energy consumption – ‘Leaner cooling’, as EC fans distribute air to servers at low speed • Easier maintenance – LSV coolers are located outside the server room, making service access more convenient • Lower costs – reduces cooling-related energy bills by up to 30% depending on size, type and location of the data centre.

LSV coolers for server halls join the existing range of Alfa Laval products such as gasketed plate heat exchangers, dry coolers and adiabatic coolers, which can be combined in integrated systems to meet overall cooling needs.

CONTACT ALFA LAVAL: 01276 633833 general.uk@alfalaval.com • www.alfalaval.co.uk

53

DATA CENTRE MANAGEMENT

053_ALFA LAVA_Spring2016.indd 53

SPRING 2016

08/03/2016 17:05

054_DCM_SPRING016.indd 54

01/03/2016 23:09

CRITICAL EQUIPMENT

MAXIMUM ENERGY EFFICIENCY

Innovative cooling for a new high performance data centre

HARTL GROUP has started up a new data centre in Hofkirchen, Lower Bavaria, fully equipped with Munters’ data centre cooling solution, OASIS IEC, which met the requirements on energy efficiency, operational safety and reliability even under extreme conditions.

High power consumption and rising energy costs require data centres to achieve maximum energy efficiency. The heat from the data halls must be removed as efficiently as possible while maintaining operational safety and reliability at extreme climatic conditions.

Anton Immerz, graduate engineer and key application manager for data centres at Munters, tells us how the cooling solution for this new data centre was implemented...

What is the principle behind the cooling?

At the heart of the solution is Munters’ patented polymer heat exchanger. The result of Munters’ 15 years of experience in data centre cooling is a polymer heat exchanger whose elliptical tubes deviate due to the movement of air through the heat exchanger in the micron range. Thereby the substances resulting from the evaporation of water on the heat exchanger surfaces are removed and collected in the sump.

What’s so special?

The Munters OASIS system does not require complex and costly water softening. The scale, which at times can be deposited during the evaporation process on the surface of heat exchanger, is removed during normal operation of the plant.

The main water source on this project for the units is well water, the water in the sump is carefully monitored for temperature and conductivity levels. With indirect evaporative cooling there is zero air mixing between the outside air and the data hall process air, this keeps the data centre cooled air free from external pollutants.

And this evaporation provides cooling?

Yes exactly. The cooling process is similar to the cooling effect on wet skin. If one gets out of the water after swimming you get cold quickly.

The water that wets the surface of the skin evaporates. In this case, it is the blood that flows on the other side of the skin warms the skin, and the heat is extracted and it starts to freeze. With the Munters’ heat exchanger, the outdoor side of the heat exchanger is sprayed with water and an outdoor air fan system allows air to flow over the outside of the tubes. As the water evaporates on the outside surface of the heat exchanger the data centre air flowing on the inside of the polymer tubing (the ‘blood’ in this comparison), and thus heat is extracted.

55

CONTACT MUNTERS: 01480 410 223 datacentres@munters.com • www.munters.com

DATA CENTRE MANAGEMENT

055_Munters_Spring2016.indd 55

SPRING 2016

04/03/2016 10:47

critical equipment

Why is true modular UPS the market standard?

Shri Karve, senior technical consultant, sheds some light on merits of a true modular UPS system by comparing it with the basic modular system Over the past 10 years, many data centres have been installing true modular UPS systems that allow ease of scalability with a ‘build as you grow’ approach. Transformerless technology and ‘hot swap’ capability of are some of the key improvements that allow manufacturers to offer compact modules that form a building block of vertical stackable system that can be extended.

Future proof A number of factors must be considered when choosing a UPS system in order to protect the load from mains supply related disruptions.

These are reliability, resilience, nature of load, power quality requirements, space requirement, future scalability and total cost of ownership (TCO). Correct UPS system scalability and sizing right from the outset is important to avoid issues related to future expansions of the data centre.

56

There seems to be a lot of misconception with regard to true modular UPS system topology. It is therefore vital to differentiate between true modular solution and a scalable parallel UPS system. Choice of module sizes and associated scalability requires careful consideration in order to meet the specific needs of the data centre with regard to CAPEX and OPEX at optimum level. Small incremental steps can provide better flexibility, less costly redundancy and offer higher operational efficiency reflecting in lower running cost when compared to UPS systems made of 200 to 300 kW increments. If a UPS system is made up of larger standalone individual units then there is a risk that the system input power factor may go leading at lighter loads and much lower overall efficiency. Leading power factor can cause compatibility issues when supported by Genset.

Initial capital cost for a modular system can be more than that of

Data Centre Management

056_Gamatronic_Spring2016.indd 56

a standalone system. However, this can be offset by the reduced operating costs due to its higher efficiency, lower infrastructure costs, low cost of maintenance/spares and superior availability.

In view of the above, it is better to utilise a true modular UPS system since these can be configured, and readily reconfigured, so they are operating close to capacity. Scalability of UPS system provides flexibility without sacrificing required redundancy.

Efficiency Modular solutions provide higher operational efficiency compared to traditional standalone models because power demand and usage are closely linked and so there ‘Reduced floor space, lower is no wastage due to oversizing. running cost, improved This reflects in increased system reliability and resiliance’ efficiency and lower Total Cost of Ownership. Additional modules can be added to meet the increased load demand and still retaining the initial planned redundancy. Hot swappable type modules improve availability to six nines (offering availabilities to 99.9999 %) since maintenance and repair can be carried out with ease and can avoid any need for shutdown of critical load. Capacity of the system can be uplifted due to built-in incrementally within a true modular system to meet the demands of scalable data centres, without any disruption to service.

Benefits Let’s remind ourselves of all the benefits offered by true modular UPS systems: greener solution due to high efficiency across wide load range, flexibility and speed of future scalability within the footprint, hot swap capability improves availability due to improved Mean Time To Repair, reduced floor space, lower running cost, improved reliability and resilience.

CONTACT GAMATRONIC: 01480 479889 sales@gamatronic.net • www.gamatronic.co.uk

Spring 2016

03/03/2016 13:46

DATA CENTRE MANAGEMENT

057_DCM_SPRING016.indd 57

SPRING 2016

01/03/2016 23:27

5 WAYS TO

5 WAYS TO… SUCCESSFULLY DEPLOY THE EDGE NETWORK Edge data centres are on the rise but they require a different way of thinking.

Here are the top five factors to consider High definition video, mobile computing and cloud services have all resulted into a need for enhanced network performance in order to deliver an outstanding user experience. To improve the quality of high-bandwidth applications outside large urban areas, edge data centres are on the rise. These position the edge of the internet further away from traditional internet hubs. Large

1.

volumes of frequently-referenced applications and content are cached on servers closer to somewhat less densely networked or tier-two markets. Building these ecosystems away from traditional areas requires a different set of considerations. What do you need to take into account in order to successfully deploy the edge network?

SPACE CONSTRAINTS Edge data centres usually contain 5-20 racks and have to deal with space constraints. As a result, cabling from servers is often directly connected to a fibre platform in a central network cabinet. In confined sections of fibre rollouts, fibre must be able to accommodate the inevitable twisting, moving, adding and changing of cables. Data must move through cables at awkward angles without a loss of quality. This makes the expansion and rollout of broadband networks more efficient, faster and cost effective, thereby reducing ownership costs, cooling, downtime and carbon footprint.

CHOOSE A HIGH-DENSITY SOLUTION Since this network cabinet also needs a lot of space for switches, an ultra-high density fibre solution is required as all fibres are directly brought from server ports to this UHD platform. The latest solution can accommodate up to 50% more fibre optic connections in a traditional housing. It’s also important to realise that the use of data-hungry technology solutions might expand at amazing speeds, but the backbone can’t simply be replaced every few years. You need cables with a very high fibre count that retain the handling attributes of small cables, with as little termination-related hassle as possible.

58

3.

SUFFICIENT PORT CAPACITY High port density is key to a successful rollout, and traditional UHD solutions based on 72 ports per unit will definitely not suffice. Something far beyond 100 ports/U would do. Don’t forget, your edge network may have to support several consecutive generations of hardware and bandwidth standards. Although edge ports may be more costly than regular versions, saving money here is not a smart move. ‘Edge’ or ‘access’ switches connect directly to end-user devices; high port density and low costs per port are desirable. If there’s plenty of port capacity, that means lower cost of patching because the user can simply re-patch devices themselves.

PAY ATTENTION TO CABLE MANAGEMENT In such an environment, cable management is key. Double check measurements, make sure terminations are top quality, test wherever necessary, label and colour-code, watch out for cramped conduits and make absolutely sure no cables or bundles rest upon others. Bad cable management can result in signal interference and crosstalk, damage and failure, resulting in data transmission errors, performance issues and downtime.

5.

2.

4.

SUFFICIENT PORT CAPACITY As edge data centres are widely distributed across the country, automated asset management and tracking makes life much easier. Dynamic data centre environments require ongoing, precise and efficient asset management. You need a specialised solution for tracing and monitoring all changes to a physical network, including switches, servers and patch panels. A system that offer functions for mapping, managing, analysing and planning cabling and network cabinets can also include asset management, planned and unplanned changes and alarms. Such systems improve operational efficiency and facilitate management of passive infrastructure. Everything can be monitored and administrated from a common software tool. The entire infrastructure is represented in a consistent, up-to-date database, offering precise, real-time information on the current state and future requirements of the data centre.

CONTACT Reichle & De-Massari: 07803875339 gbr@rdm.com • www.rdm.com

DATA CENTRE MANAGEMENT

058_RM 5 Ways to_Spring2016.indd 58

SPRING 2016

03/03/2016 13:46

059_DCM_SPRING016.indd 59

01/03/2016 23:14

energy efficiency

Keep your cool

Energy efficient and fully tested cooling for your data centre The Excool cooling system has the most effective, efficient and lowest energy process available to cool data centres.

In the colder winter months, the system relies on sensible heat exchange typically above 70% overall efficiency, however as the ambient temperature increases to typically 20°C, water is introduced into the process increasing overall efficiency in excess of 90%.

The first stage of water introduction follows an adiabatic path along the wet bulb line on a psychometric chart using a closely controlled atomising technique, enabling relative humidity approaching 100% to be achieved.

60

The second stage uses an effective evaporative technique. The heat exchanger plates are ‘charged’ with water on the ambient side to achieve heat exchange in an effective 2mm spacing between the heat exchange surfaces.

The water is held onto the heat exchange surface during this charging process and the cooling effect transfers from air to air to an effective air to water process.

To evaporate, water requires a relative

exceptionally high energy source, which in this application is provided by the air returning from the data centre.

A relatively small amount of water evaporation provides an exceptional cooling effect all harnessed within the Excool heat exchanger. This is manufactured from a composite material, which makes it resistant to corrosion, more robust and leak tight, and able to be ‘charged’ with water.

Using evaporative and adiabatic cooled outdoor air as a cooling medium without introducing this into the data hall eliminates the dependency on mechanical cooling in thousands of global locations.

Savings In more extreme environments, there is a significant reduction in mechanical cooling relative to traditional systems.The reduction in compressor dependency leads to further capital expenditure savings in the electrical and BMS infrastructure costs.

Major savings in the costs of generators, transformers, switchgear, diesel storage, BMS and other items can be realised.

Alternatively the ‘stranded’ power

CONTACT EXCOOL: 01527 492750 info@excool.co.uk • www.excool.com

Data Centre Management

060_Excool_Spring2016.indd 60

normally reserved for the cooling system can be released for additional IT. In most cases this can add 25% to the IT power available.

Fully tested Cooling represents almost half of the energy consumed in legacy data centres and the need to improve this situation is a pressing one.

Excool’s 400m² test facility allows clients, consultants and data centre operators to witness performance tests of the Excool units in real life conditions with airflow parameters of upto 100,000m³/h.

The latest technology, such as fully variable heat loads, temperature and humidity sensing equipment, large capacity data logger/taker, energy and water usage monitoring systems are utilised.

This allows Excool units to be thoroughly tested in real life conditions across the world such as external ambient conditions ranging from 0°C upto +50°C dry bulb and +30°C wet bulb; Internal heat loads of upto 600kw simulating a working data centre.

After seven years of testing, Excool can confidently quote low pPUE’s of 1.025 in many locations across the world. The latest installations in the UK, Europe and South America are testament to this.

Spring 2016

02/03/2016 14:16

DATA CENTRE MANAGEMENT

061_DCM_SPRING016.indd 61

SPRING 2016

01/03/2016 23:29

energy efficiency

Boost your bottom line

Why the perceived risks of grid demand balancing schemes may be significantly lower than many believe...

The opportunity to enjoy revenue from Dynamic Rotary UPS equipment by participating in grid demand balancing schemes is a well-established concept.

However, data centre take up has been slow; primarily due to misconceptions that third party control is required bringing additional risk to the critical operation. Alex Greenhalgh of EURO-DIESEL explains why the perceived risks of these schemes are significantly lower than thought.

The real time frequency of the National Grid is a relatively accurate measure of the balance between power generated and consumed. It is also a good indication of the health of the grid or whether grid instability is imminent.

62

Grid stability The National Grid has an obligation to supply power within a frequency tolerance of Âą1% of the nominal frequency - 50Hz. Stability of the distribution system is managed by constantly controlling the balance between supply and demand. Predicted demand levels allow the grid operators to forecast and plan supply requirements.

However, constant variation in demand necessitates real time control and allowances made for any difference between planned and actual.

Traditionally, the grid was supplied by thermal plant power stations with high inertia machines synchronously locked via the distribution grid. Today, a higher portion of the supply mix is derived from asynchronous renewable sources such as wind and solar. The reduction of high inertia feeds has led to a supply that is less able to tolerate demand change without

an impact on stability.

Traditionally, balancing the demand could only be achieved by funding power stations or private operators to adjust their generation output. There are now opportunities to participate in demand reduction programs through frequency response contracts which significantly reward owners of critical power systems with spinning reserve technology, such as Rotary UPS systems.

Another way The EURO-DIESEL Dynamic UPS system, sits between the incoming mains supply and the sites’ critical load. It comprises a synchronous machine, choke, flywheel storage device and diesel engine. The system constantly measures the quality of the incoming mains supply. When the supply is healthy, it takes power from the

grid in an energy efficient manner to both condition (filter) power to its critical load and to maintain its high inertia flywheel energy store. In this state, the diesel engine is not participating and remains stationary in a standby condition. When the supply frequency falls below a predetermined threshold, the Rotary UPS is automatically released from the grid by opening the mains incoming breaker. Critical load is supported throughout by taking energy from the kinetic source whilst the engine is started and engaged.

When applied to National Grid demand reduction schemes - the Rotary UPS is already automatically reducing demand as a function of its operation. No further control or third party intervention is required. The Dynamic UPS owner remains in control of their plant at all times and risk to the operation remains unchanged. An interface that transmits energy information from the incoming billing meter is all that is required for the scheme operators to ensure that participants have actually reduced demand.

By selecting Dynamic UPS as the technology of choice the Data Centre owner can realise a revenue stream from the Frequency Response Program that will contribute to the bottom line profitability of the business.

CONTACT EURO-DIESEL: 01285 640879 info@euro-diesel.co.uk • www.euro-diesel.com

Data Centre Management

062_Eurodiesel_Spring2016.indd 62

Spring 2016

03/03/2016 13:48

DATA CENTRE MANAGEMENT

063_DCM_SPRING016.indd 63

SPRING 2016

01/03/2016 23:29

ENERGY EFFICIENCY

A perfect partnership

Working together to improve resiliency

DSM has built up an established reputation as a provider of business continuity and data centre services; maintaining a reliable power supply to customer equipment is the lifeblood of its business. With reliability at its core, DSM maintains the highest standards of resiliency to ensure continuous business supply for its customers, which include, among others, Yorkshire Building Society, Travelex, Anglian Water and Royal & Sun Alliance.

64

The challenge DSM was looking to upgrade the UPS power infrastructure at its Sibson site to bolster the resiliency of its recovery facility and meet the demands of its forthcoming data centre expansion. The recovery facility houses nearly 700 workplace recovery positions, which are set up ready for customers who are unable to work from their own premises for any reason – such as power failure, flood or other catastrophic event, while the 100-rack data centre caters for clients requiring colocation services.

Speaking about the project, Amanda Johnson, operations director at DSM, said: ‘This was part of a major project to improve the resiliency across our entire site. We were looking for a UPS solution

THE SOLUTION:

• 2 X MULTI POWER MODULAR UPS • 126KW LOAD IN EACH • FULL SET OF NEW BATTERIES (80) • COMMISSIONING that was reliable, compact and efficient, and one that could be easily upgraded in terms of power output. We were also working to a fairly tight timescale so we needed a solution that could be implemented quickly and by a trusted supplier.’

The solution Responding to the brief, the team at Riello UPS commissioned two of its award-winning Multi Power Modular UPS’ at 126kW plus a full set of new batteries.

These would be situated in the powerhouse adjacent to the data centre.

WORKING TOGETHER ‘The investment underlines our commitment to ensuring power to customer equipment is 100% maintained,’ Amanda said. ‘Resiliency is what our business is all about and it’s what we build our reputation on, so choosing a reliable company like Riello is key to this.

‘We chose Riello because of its strong brand reputation in the industry, the technical ability of its engineers and the reliability and scalability of its products.

The impressive track record when it came to ongoing support and maintenance was also another deciding factor as this is an important part of power protection.’

CONTACT RIELLO UPS: 01978 729 297 sales@riello-ups.co.uk • www.riello-ups.co.uk

DATA CENTRE MANAGEMENT

064_Riello_Spring2016.indd 64

Another key consideration for DSM was a UPS that was efficient at low loads and with minimal single points of failure.

‘The UPS needed to be physically compatible with the space we had,’ Amanda said. ‘It also needed to be modular so that we could add to it or take away as required, and hot-swappable to allow for easy maintenance without downtime.’

Craig Smithson, sales engineer at Riello UPS, said: ‘We recommended the modular unit because it offers flexibility and the scalability to grow. The timescales on this particular project were particularly challenging but, because we hold stock at our UK site, we were able to quickly move from recommendation through to delivery to on site commissioning in a short space of time.’

Designed to achieve the highest levels of power and performance in a data centre environment, the Multi Power three-phase double conversion system gives complete scalability from 42kW to 1176kW. It maintains high efficiency in online mode operation of up to 96.5%, even at low loads of 20%.

SPRING 2016

03/03/2016 13:50

065_DCM_SPRING016.indd 65

07/03/2016 13:08

CRITICAL EQUIPMENT

STAY CONNECTED

A proven way of improving flexibility while minimising loss

During its 25-year history within the market, SpliceGroup has deployed its f-Tube solution into many data centres across Europe and the USA. Many of Splice’s existing clients are high-end blue chip companies and they are enjoying the full benefits the f-Tube provides.

66

So, what is f-Tube? It’s a proven way of provisioning a network for fibre connectivity utilising a tube that allows easy installations and de-installations reacting to demand quickly.

Essentially, f-Tube is a ‘pulled fibre’ solution.

By using Splice’s tube solution alongside a pre-terminated cableing approach for critical applications (such as SAN switches, storage arrays, etc), it reduces points of failure, therefore reducing channel loss, while maintaining a full structured approach to the overall cross connect environment. F-tube is modular by design, allowing the end user to be creative in their approach to connectivity.

Future proof By simply installing a tube network within your data centre not only can you meet your current demands, you will also create an environment where expansion, upgrades and replacements are effortless. The range of innovative fibre management solutions available easily facilitates

the high connectivity demands of fibre growth within your data centre.

F-tube by its very nature allows you to rapidly deploy additional connectivity without affecting the existing fibres as they are accommodated within separate tubes, which negates any risk. Not only is it easy to add new connectivity, it’s also just as easy to remove redundant fibres.

In an expanding and dynamic data centre environment, it is important to be able to create and maintain connectivity flexibility. F-tube allows you to just that, ultimately providing an any-toany cross connect. It also allows infrastructure be installed without the need to access underfloor voids or overhead containment.

The complete solution F-Tube integrates with Splice’s existing data centre solutions including Optical Distribution Frames, Pre-Terminated Cables, and Cassette Systems, all of which are supported by the company’s fully managed data centre services to provide a complete solution.

Splice Group will be at Data Centre World at Excel in London on 12-13 April 2016. Head over to stand L60 for a chat about improving your fibre connectivity for now and the future.

CONTACT SPLICE GROUP: 0113 285 2536 info@splicegroup.com • www.splicegroup.com

DATA CENTRE MANAGEMENT

066_Splice Group_Spring2016.indd 66

SPRING 2016

04/03/2016 12:17

SECURITY

Avoid an attack

Rittal’s white paper on IT security tackles the issues of physical hazards head on While the world’s reliance on digital technologies is growing, the need for heightened security around IT and data centre infrastructure has never been greater.

Rittal’s new white paper, Physical security in IT and data centre technology, tackles the issues of physical hazards head on. It identifies fire, water, dust and fumes, falling debris, EM radiation and unauthorised access as prime risks and explores what can be put in place to prevent damage when such issues arise.

The paper guides data centre planners and operators through the complexities of securing their sites. It highlights the European Standards and Certification needed to demonstrate effectiveness under test conditions and compliance not just for individual components but also, very importantly, systems.

Best practice advice The Paper takes a look at various threats and offers best practice advice. For example, it

DATA CENTRE MANAGEMENT

067_Rittal_Spring2016.indd 67

considers IP (International Protection) categories in relation to protection against dust, water and solid particles, as defined by EN 60529.

Physical threats also include deliberate human interference and any space filled with expensive hardware and important data should be secure enough to prevent malevolent or vindictive individuals gaining access.

IT security rooms are also reviewed. These help protect hardware and data from the negative impact of EM radiation, fires and water damage.

For example, Rittal’s high availability room offers the maximum physical security for data centres and IT system locations. A compact alternative are IT security safes, which are ideal for smaller data centres that only need to be connected to the appropriate supply lines on site.

CONTACT RITTAL: www.rittal.co.uk • www.friedhelm-loh-group.com Follow us on Twitter @rittal_ltd

67

SPRING 2016

03/03/2016 13:51

COMPANY Q&A

An Olympic effort

Paul Davies discusses the launch of the Here East development and what this London hub means for Coreix.

Tell us a little bit about Coreix...

Coreix Limited is a London-based managed hosting, colocation and network services provider that offers scalable and robust hosting solutions to more than 500 clients in the enterprise, financial, SMB, pharmaceutical, public sector and wholesale sectors, to name but a few.

We refreshed the Coreix brand last year and streamlined our business operations to enable our customers to have tailored solutions that match their IT strategies and budgets.

68

You are launching a new data centre in the Here East development in Stratford, what does this mean for Coreix and your customers?

Here East benefits from the rich connectivity designed to service the Olympic Games and will enable us to have access to the most advanced digital infrastructure in Europe. With many carriers in the vicinity and an extensive legacy duct network already in place, Here East is ideally positioned for connectivity. As the new site is part of the Olympic legacy and based in the building built for media, it houses the core infrastructure for BT Sport, so is a major BT Pop and on the LD4 fibre ring. Coreix provides superior ultra-fast, low latency connectivity to customers via its fibre network. This offers access to an array of national and international tier one and tier two carriers, internet service providers and internet exchange points including LINX and LONAP. The new site will look to be the most energy efficient of all data centres based in the UK and

further afield with a realistic PUE of below 1.25.

The site will use a combination of high efficiency free cooling chillers, hybrid adiabatic cooling and centralised district cooling making the most of the Olympic legacy features of this unique site.

What new developments do you have planned for the near future?

Having invested heavily in developing our service offerings over the past three years, we continue to innovate and will be launching our second generation public cloud in the coming months, which will feature full flash based storage, enterprise tools and support as standard.

We’ve also increased our portfolio of private and hybrid cloud solutions allowing us to match almost any IT strategy or budget.

What can we expect from your speaker slot at Cloud Expo?

With the launch of the Here East development taking place in the upcoming months, we will be looking to inform, educate and engage IT managers and directors about London’s newest technology hub. Those attending will be able to learn about how London’s latest data centre facility and how Coreix will enable businesses to have best in class connectivity, space, power and cooling.

‘Here East benefits from the rich connectivity designed to service the Olympic Games and will enable us to have access to the most advanced digital infrastructure in Europe’

What does the future hold for Coreix?

In the coming years, we will be looking to increase our data centre footprint in London, and the number of points of presence across EMEA and APAC – watch this space!

Paul Davies is the technical director at Coreix, responsible for overseeing the data centre and network operations teams. His expertise lies in network architecture, software and solution engineering.

CONTACT: 0800 022 6734 or 0207 183 1725 enquiries@coreix.net • www.coreix.net

DATA CENTRE MANAGEMENT

068_CoreixQA_Spring2016.indd 68

SPRING 2016

02/03/2016 14:17

SECURITY

Your flexible friend

Easier and secure access when you need it most Key management is often a concern, especially in data centres or large offices where keys are easily lost or stolen. Electronic digital keypad locks are far more convenient to manage on a large scale; if codes are forgotten or compromised, the locks can be reprogrammed in minutes.

Codelocks latest product in the KitLock range, the KL1550 Smart, provides flexible access for end users and makes it much faster and easier for operators to manage and control access to large numbers of lockers, cabinets and enclosures.

Be smart Flexible keyless access is available via code and smart card. The lock also supports NetCode, a web-based application, allowing users to generate one-time access codes from a remote location and send the codes via SMS or email direct to the person who needs it. This is useful for service engineers when visiting remote cabinet enclosures.

The integration of technology opens up new market areas for Codelocks and allows

standalone digital locks to compete with more sophisticated wireless access control systems. Using an NFC-enabled handheld device (that is purchased separately), operators can upload selective programs and features, and also download audit trail data on demand in a matter of seconds. For access to storage used regularly by a small group of people, up to 50 smart cards can be issued.

Codelocks will be exhibiting at Data Centre World, stand A70, 12-13 April 2016 ExCeL London.

CONTACT CODELOCKS: 01635 239645 • www.codelocks.co.uk/kitlock.html

69

DATA CENTRE MANAGEMENT

069_Codelocks_Spring2016.indd 69

SPRING 2016

02/03/2016 14:26

QUIZ TIME

It’s just for quiz in each issue fun! Atonew enjoy in a spare moment As part of Data Centre Management’s fresh and exciting new look, we thought we’d bring you something a little bit different alongside all of the informative features and articles.

Perhaps in the future we may be include it as a competition with real prizes to give away, but for now there’s no pressure... think of it as your warm up!

So whether it’s your lunch break or time for a quick cup of tea, take five minutes to yourself and relax with our quiz. We’ll bring a different one to enjoy in each issue and it’s all purely just for fun.

If you have any ideas for future quizzes we’d love to hear them.

Simply email Heather at heather.grimes@closerstillmedia.com and we’ll see what we can do.

Can you spot the subtle differences?

There are 10 in total.

B.

A.

70

Answers: 1. Team: 14 people 2. Ipad: SEO 3. Teamworking: 5 people 4. Network working 5. Network: 4 circles 6. Site Search 7. Lighbulb: Extra Skyscraper 8. Pie-chart: 3 segments 9. Lighbulb: Birds 10. Cloud Arrow: Points Down DATA CENTRE MANAGEMENT

070_Competition_Spring2016.indd 70

SPRING 2016

02/03/2016 14:19

071_DCM_SPRING016.indd 71

01/03/2016 23:20

072_DCM_SPRING016.indd 72

07/03/2016 12:05