Tuesday, 17 September 2019


A DATA-CENTRIC APPROACH TO MANAGING DATA CENTRES

“Data centres owned and operated by data-centre landlords, cloud services and other technology firms is expected to increase to roughly 9,100 this year, up from 7,500 last year, and are expected to reach 10,000 by 2020, IDC estimates.”  Source: The Wall Street Journal.

If all other data centres including hyperscale and enterprise were added, the total figure could be in the billions.  Businesses around the world rely upon data centres being available.  There is also more focus on the environment and climate change, so there is more focus on efficiency, and carbon-neutral designs – ergo yet more complexity to manage.

There is a reason that DCIM hasn’t been replaced with something new.  It has had a bad rep for many reasons, but it is necessary to help us manage ever more complex, hybrid environments, and so it has to evolve.  It needs to connect to facilities systems, network systems, IT systems and orchestrate changes as they are required.  No longer can the M in DCIM represent ‘monitoring’.  Perhaps the metamorphosis of DCIM more accurately should be DNIO: Data centre, Network and Infrastructure Orchestration.

DCIM is now moving into the IT stack, and integrating with systems, such as ITAM, CMDB and cloud-based systems.  It now offers the ability to analyse data across sites, and provide AI based solutions to controlling the data centre throughout the IT stack – from the BMS, through to application performance.

One of the hardest elements in a DCIM implementation has been integration, and figuring out how processes and procedures should work, and then how to automate them.  This integration piece – in the past - has either been technically challenging, or financially challenging, or seen as scope creep, or it has been something that a vendor or stakeholder has discouraged.

What is really required is an open integration suite that would allow enterprises to pull their own bespoke solutions together, without racking up expensive development bills. It seems this vision is slowly becoming a reality after some M&A activity in the DCIM space, and clients and vendors steadfastly staying the course behind the DCIM vision.

This brings with it a different way of looking at managing the data centre: it’s a data-centric view.  Instead of worrying about whether an integration is possible, it’s reasonable nowadays to assume that it is. Therefore, it is possible to design the system in the most efficient way and make use of automation where it makes sense.

Here are six encouraging areas of progression where more integration is enabling positive leaps forward:

Broader scope of infrastructure managed by DCIM:

The links to CMDB, ITAM and other systems on the IT side are bringing more data analysis opportunities, with a broader scope of data points.

Use of Artificial Intelligence:

AI is being used more readily in a number of areas within the DC.  For example, cooling optimisation, and security.  AI can learn normal network behaviour and detect cyber threats based on deviation from that behaviour. 

Open platform approach:

Instead of a silo’d approach both internally and externally, the data-centric view of the DC should take priority, which means that IT, Facilities, and vendors are all working together.

SDK / Open  API:

A number of vendors are providing SDKs or Open APIs, which are a good step forward to making integrations between systems work, and it shows that they are open to working with other companies.

CMDB and Asset Management:

There is a recent move to focus on asset management and aligning assets in ERP systems too, to provide a single source of the truth.  From a data centre perspective, having the assets managed well, is an essential building block to DCIM and data centre management.

Processes and Procedures:

Data centre operators are viewing the system as a whole and are finding areas where technology can automate processes.  For example, adds, moves and changes can be streamlined, saving around 30% of resource time by using accurate DCIM data and integrated workflows.

In a world where IT systems are becoming more distributed, and IoT is making its mark, data centres must take a data-centric approach to managing the system of systems housed under their roofs.  Silo’d thinking no longer has a place in the modern data centre:  DC and IT managers need to work together, alongside a multitude of vendors who also need to align and integrate their offerings to the clients’ needs.

This open platform approach enabling integration brings many benefits to life.  An integrated workflow capability facilitates automation, reducing resource time required for operational tasks.  With more visibility of systems, capacity management from the CRAC unit through to ports in the meet me rooms, is a reality allowing the DCIM to assist with intelligent commissioning of new assets and patching routes.  Energy optimisation now involves data from the servers themselves, allowing them to shift workloads when compute requirements are low, thus allowing a server to potentially stand down.

With this data-centric approach, the return on investment should not only be better, it should come in sooner as well.  The software-defined data centre is now in view. 


Guest Blog written by:



Assaf Skolnik, CEO, RiT Tech



Venessa Moffat, Head of Product Marketing, RiT Tech

Marketing, Strategy and Growth Hacking specialist, with 20 years’ experience in the Data Centre and tech industries. Venessa holds a BSc in Computer Science, a Post Grad Diploma in Business Administration, as well as an MBA from Essex University, where she specialised in agile IT architectures for maximum business value. She has successfully led strategy development and implementation programmes in multiple international data centre organisations. 



No comments:

Post a Comment