Maverick or pragmatic?

RiT Tech’s Mark Acton on why deploying a DCIM to mitigate operational risk is no flight of fancy.

“Military precision” is not a misnomer.

Being in the business of bombs and bullets demands organisational excellence and exacting standards and is the reason why so many companies actively court the services of those calling time on their careers in the UK’s Armed Forces.

Soldiers, sailors and airmen are masters at mitigating risk, having been meticulously and exhaustively trained to counter myriad threats in complex areas of operation. 

This military preparedness is key to saving lives in conflict but also preserving the well-being of incredibly valuable pieces of hardware, such as the Royal Air Force’s fleet of Typhoon aircraft.

With each fighter jet costing circa £75 million, pilots do not get to fire the afterburners of the UK military’s most expensive weapon system in anger without first undergoing hundreds of hours of training in simulators and computer-driven cockpits on terra firma.

The RAF’s trust in technology to help its personnel avoid future flak is a lesson those in the data center sector would do well to note, not least given their mission’s primary objective is to ensure the reliability and availability of clients' IT services 365 days a year.

While fuelling the Cloud is not as dangerous as flying above the clouds, the financial thrust needed to do so is similarly sky-high and doing all that is possible to protect this investment – a build price in the region of a quarter of a billion pounds and annual running costs of £20 million – is far from maverick behaviour.

The right Data Center Infrastructure Management (DCIM) solution can form a key component of this defensive arsenal, providing a level of real-time situational awareness not possible using information afforded by discipline-specific tools and the human eye – no matter how experienced, skilled and alert its owner may be.

By federating all the data available from its surrounding environment, a fit-for-purpose DCIM will deliver vital intelligence on the inter-dependencies of IT assets, power and cooling that enables potential risks to operational efficiency to be identified and mitigated.

Knowledge is power but is something data centers – which are often organisationally siloed – have been historically bad at garnering. A building management system can be great at monitoring the performance of its host’s mechanical and electrical elements, just as a configuration management database might be adept at storing accurate details about hardware and software assets, but basing proactive and reactive operational decisions on isolated sources is the equivalent of flying blind.

An organisation’s servers can be the very best money can buy but if the building’s power fails those expensive blinking lights are worthless when the cause of the problem cannot be quickly addressed. A DCIM joins the dots and serves as a force multiplier.

Just as the success of a Typhoon sortie relies on the orchestration of multiple elements – the aptitude of the pilot, intelligence from mission command and fidelity of the target acquisition coordinates provided by forward air controllers on the battlefield to name but a few – so too does the smooth running of a data center.

The advancement of computer analytics means that it is now possible for machines to do the mathematical heavy lifting in the facilities charged with meeting society’s insatiable demand for connectivity; to draw data from across a facility’s grey and white space, calculate permutations in a split second and suggest a roadmap to optimisation.

I am yet to meet an individual capable of doing the same and given that to err is human, it is easy to see how a DCIM can become an invaluable ally. Capable of generating automated, high-fidelity work orders that can signpost engineers and technicians directly to a point of failure or suggest the optimal configuration for new assets, it can dramatically reduce the chance of human misadventure.

The evidence for preserving the status quo does not stack up with research conducted by Jerry Williams and based on his work in the UK energy sector, highlighting there is a 50 percent chance of something going wrong when a person is required to perform an unfamiliar task at speed and under pressure.

While this figure does drop dramatically to 0.04 percent if the task is a familiar one, the advantages of having a DCIM guide in the event of a crisis is clear – when dealing with downtime, “events” are not the type of odds that promise peace of mind. 

An alternative to mitigating human hiccups would be to again borrow from military best practice and ‘drill’ data center staff repeatedly in the hope that they can more instinctively respond to critical situations, no matter how rare. Such intense training, however, would come at a considerable cost – both in terms of time and money.

A DCIM’s ability to plug any gaps in the understanding of the finer nuances of the day-to-day realities of data center operations could also prove crucial to fending off any regulatory risks.

The sector is coming under increasing Government scrutiny over power consumption, albeit while contradictorily being asked to facilitate faster networks and greater capacity, so being armed with accurate facts may assist operators in fighting their corner. 

A DCIM configured to report power usage effectiveness against ISO Standard metrics and key performance indicators could create a clearer picture of causal links and help ensure any regulations are introduced for the right reasons.

Finally, and perhaps most importantly, the adoption of an analytical autopilot can eliminate any threats to a data center’s fiscal fitness. Whether ensuring an operation is not haemorrhaging money as a consequence of unwittingly under-utilising capacity or delivering savings through unprecedented levels of synchronicity in the consumption of space, power and cooling, a DCIM is a worthy wingman when it comes to evading risks to profit margins. 

In an ultra-competitive market, those data center operators with designs on being the sector’s top gun that don’t yet have an automation ace as part of their crew, should “feel the need…”. 

LinkedIn

With over 25 years experience, Mark is a specialist in data centre operations, concentrating on the delivery of business critical services from highly reliable, world class data centres with 24x365 availability expectations. Mark is a respected data centre leader. He has extensive international experience and solid technical skills combining data centre facilities design, IT and facilities operational management with technical consulting.