Optimising data centre and IT strategy ‘in the round’

Q Please can you provide some background on the company – when formed and with what objectives?

A I formed the company in 2010 with the backing of investors that I had worked with in the past, as well as new investors such as Carbon Trust Investments. The aim was to take advantage of a mature software product that had been developed for managing High Performance Computing (HPC) systems, and to deploy this into the broader data centre market to help drive energy and operational efficiencies.

Q Who are the key personnel involved in the company?

A I’m responsible for the technical management of the company and the overall vision. Terry Hale is the recently-appointed CEO; Terry brings a lot of experience in sales and operational management, having previously run a number of VC-backed businesses across several sectors. Our executive chairman is David Morriss, who was a senior general manager at IBM with responsibility for UK and EMEA until the late 90s; he has since held board positions with a number of SMEs. The executive management is completed by Martin Hall, our CFO, but we have a fantastic team of developers, engineers and sales professionals who are our biggest asset.

Q What are the key milestones in the Concurrent Thinking story to date?

A We were founded in 2010 and acquired our first customers in 2012. We have been fortunate enough to benefit from some game-changing events over the last twelve months: firstly, we secured a large grant from the Department of Energy and Climate Change’s “Energy Entrepreneurs’ Fund”, then recently closed an investment from both new and existing shareholders. These have helped us achieve critical mass in terms of technical development capability, engineering support and sales momentum.

Q What are the USPs that differentiate the company in what is quite a crowded DCIM marketplace?

A We designed our Concurrent COMMAND DCIM application with the delivery of business value and technical flexibility in mind. The vision was to provide a single DCIM solution that could monitor and manage virtually all IT and facilities devices within the data centre using industry standard protocols. The sheer breadth of the capabilities that we can offer, along with our vendor-neutral stance, marks us out from most of the DCIM products out there. We also think of our DCIM solution as a toolkit, so that the user is not constrained by what we think he might use our software for, but can instead configure the product to do what he wants it to do - to the extent that he can even write his own scripts that tightly integrate with our system.

Finally, and bearing in mind there is 3-4 orders of magnitude more data available from IT systems, operating systems and virtual machines compared with data available from facilities equipment, our system was designed to scale to the requirements of large data centres from its inception. When you start to think about the problems of scale, you also think about intuitive GUI design, because a bad GUI becomes an unusable GUI at scale, and a good GUI at scale is generally a great GUI for the average user.

Q Part of the problem with the DCIM market is that there is no agreed definition on what it is – how does Concurrent Thinking define DCIM?

A For me, DCIM is about bringing all or most of the important systems in the data centre together and providing management information that allows the customer to continuously identify and then make operational improvements. So a true DCIM suite monitors and/or manages the data centre building, environment, power distribution and IT. It should also understand the interrelation between these systems if it’s going to provide valuable business information.

Q So, the ideal DCIM offers complete end to end management (and analysis) of the complete data centre infrastructure, including IT?

A Yes. If you exclude IT from the equation you exclude the opportunity to make a good proportion of the potential savings available to you, as many of these potential savings revolve around IT utilisation. If you talk candidly to a typical data centre manager, he often has a keen suspicion that a reasonable amount of the IT equipment within his data centre is not doing that much at all, or should probably be replaced by newer equipment. He just doesn’t have data to support this. The problem is that without data, he can’t easily make the case for change, so he is forced to make decisions in isolation.

Ultimately, the data centre’s sole purpose is to provide a suitable environment for the IT equipment it houses, and optimising facilities management without taking into account potential changes in IT is not, in my view, DCIM. Ideally, organisations should look at their data centre and IT strategy in the round. What systems might be virtualised and how should they size the requirement for each new virtual server based on historical data? For those systems that can’t be virtualised, could they be replaced by newer more power-effective servers? What services might be retired altogether? How can the PUE be improved based on these requirements? What are the potential costs and benefits of these changes, and how do they relate to both CAPEX and OPEX?

These are the kind of questions a DCIM solution can help answer. And because the world isn’t perfect, these changes can’t generally all be made at once, but a good DCIM product helps the customer make iterative changes, each of which provides incremental benefit.

Q Presumably, some kind of modularity within this end-to-end approach is desirable, to allow customers to build a DCIM solution over time, rather than go from nothing to everything overnight?!

A Absolutely. Consequently we make our software available as individual modules delivered as part of a single application framework, so that the immediate needs of each data centre can be met whilst providing the opportunity for additional modules to be added in the future when budgets and resources allow.

Q How useful is a DCIM offering that merely monitors what is going on in the data centre, in isolation from the true purpose of data centre to underpin an organisation’s business applications?

A Data centres exist to deliver IT services and therefore a DCIM offering that ignores the efficiency of its business applications really is missing a trick.

Our DCIM solution can track and monitor power down to the server level, and it can also monitor application performance using protocols such as SNMP, WMI, or via the in-built scripting interface. The system can then easily be configured to give real-time end-to-end performance metrics – think cpu utilisation /watt or even emails / watt, SQL queries / watt or web refreshes / watt; similarly for cost-based metrics. If you’re monitoring things like PUE, you can also add in the cost of the cooling overhead. Things get even more interesting when you combine this information with information from the in-built asset management database, as you can start adding in CAPEX depreciation etc. What you then have is end-to-end information that helps you make real business decisions.

Q Can/should a DCIM solution provide some kind of proactive management?

A Our view is that a DCIM solution should be a toolkit for the data centre manager to use. We provide the capability for a data centre manager to monitor the key performance metrics that are of interest to him, and then a powerful GUI that allows him to dig deeper and deeper into the data in order to draw conclusions and make informed decisions. If he wants to manage more proactively and even automate, then our solution provides him with tools that also make this possible. We have purposely refrained from taking a prescriptive approach and firmly believe that the end-user should drive the DCIM solution and not the other way around. After all, nobody knows his data centre and objectives better than he does.

Q Virtualisation and Cloud continue to be buzz topics in the data centre space. Does DCIM have any role to play in helping end users approach some, or all, of them?

A Our solution is already integrated with virtualisation software platforms, and this has to be a key element of DCIM. Any efficient cloud provider will need an efficient data centre, and so a DCIM. I fully expect there to be close integration between key cloud delivery platforms and DCIM solutions, and technology like this will merge as Cloud becomes more mainstream.

Q Please can you tell us a little bit about Concurrent COMMAND, your DCIM offering?

A The Concurrent COMMAND application lies at the heart of our DCIM software suite. It provides a wide range of configuration, reporting and management capabilities through a modern, intuitive and powerful GUI. The GUI is based on the latest web protocols and can be accessed from any internet-capable device.

Concurrent COMMAND supports industry standards such as Modbus, SNMP, WMI, IPMI and 1-wire protocols, as well as key vendor-specific protocols, such as Intel Node Manager.

While we deliver all of our capabilities through the same web application framework, they are licensed as individual modules so that the customer can choose to implement a single module or any group of modules based on his technical or budgetary requirements.

Q In more detail, what does each module offer?

A The modules available for customers to choose from are the following:

 Asset CONTROL, which offers visual asset management for
keeping track of IT and facilities devices within the data centre.

 Facilities CONTROL, which is used to configure and monitor
mechanical, electrical and cooling systems.

 Environmental CONTROL, which monitors environmental factors.

 Power CONTROL, which provides fine grain monitoring of power
at rack, server or device level.

 Server CONTROL, which collects a range of metrics from BMC
and SNMP-enabled devices.

 IT CONTROL, which collates information at the operating system
and application level.

Q Many data centre personnel are coming to DCIM ‘cold’. Where should they start?

A I’d suggest they start simple. To some extent it depends on the customer, but monitoring power distribution and cooling performance at the building level as well as environmental conditions at the rack level is the obvious one. It is surprising how easy it is to obtain a short term ROI by reducing PUE without making any major changes to the data centre.

Some customers are actually keen to start by getting their asset management under control, which naturally leads to some level of IT monitoring and management.

Once a customer is up and running with a simple installation, our DCIM solution can grow with him as his requirements change.

Q You recently announced new funding for Concurrent Thinking. How will this enable the company, and its solution, to move forward?

A The funding is all about growing our sales and marketing resources, and executing a commercial plan that can leverage the significant investment that we have already made in R&D.

Q What kind of traction/presence has Concurrent Thinking achieved to date, and what more is there to be done?

A Naturally, given that we are a young and upcoming company, most of our sales to date have been into the UK, our home market. We’ve seen particularly good success in the public sector, including Higher Education, where saving energy is a key driver from both a political and financial perspective.

But there’s a great deal more to be done! As I said earlier, the primary purpose of the recent new investment is to drive our sales and marketing efforts. We’re a very ambitious company and are seeking new channel partners who can help us increase our geographical reach in all vertical industry sectors.

Our product offering has the potential for universal appeal and we’re determined to exploit this potential.

Q What can we expect from the company during the rest of 2014?

A Expect to hear much more noise from us in general, including news on some new partnerships. We’ll also be announcing some interesting capabilities aimed squarely at the co-location marketplace. Watch this space!

Q Are you able to share one or two brief customer successes to show how Concurrent COMMAND is helping businesses seeking to optimise their data centre performance?

A We’ve just released a white paper on the work we have done with HPC Wales at the Dylan Thomas data centre in Swansea, where we are monitoring water cooled racks and associated auxiliary equipment, as well as proactively managing power at the server level. This is available to download from our new web site. We’re also just completing a study on data centre efficiency with the University of Surrey. This is particularly interesting as it involves monitoring virtually all aspects of their data centre, from power distribution, cooling systems and environmental conditions, all the way down to operating system utilisation. We’re also just about to start working with them on application performance metrics. Once again, a white paper on this will be available in the coming weeks.

Q Any other comments?

A Well, thank you for the opportunity to talk to DCS, and a bigger thank you for those readers who have made it this far! I hope the article has brought across what makes us different from our competition and how we might help data centre managers make a real difference to operational efficiency. We are a young, dynamic and open-minded company and would certainly welcome the chance to talk to your readers about what we can offer them.