Menu
Menu
The big data equation: Coming out of the dark

The big data equation: Coming out of the dark

Every enterprise move generates data - but they are not necessarily being tapped to provide better business outcomes, says HP.

Every action creates a unique digital footprint from within the organisation or outside. This could range from a small tweet to a customer invoice. Yet, much of this data is not analysed today, says Mohan Krishnan, vice president and general manager technology consulting, technology services for HP in Asia Pacific and Japan.

Krishnan uses the term “dark data” for corporate information that it not harnessed for predictive analytics.

He asks, “Why is this data not analysed even if it results in benefits to the customer?”

Large complex information is difficult to process and analyse, he says. Data is both structured and unstructured, with 80 percent of the latter comprising total data and growing two or three times faster than structured data.

“Organisations want to accelerate the monetisation of big data, but do not fully comprehend the impact, strain and complexity involved in building an integrated, well-managed big data environment,” says Krishnan, who was speaking at the HP World Tour in Beijing.

Mohan Krishnan

The first step, he says, is to develop an integrated IT strategy for big data management. This starts with a ‘transformation workshop’ where both heads of ICT and line of business executives meet to define and architect an approach for big data.

The goal is to take data, analyse them, and co-relate them to provide better business outcomes in real-time, says Krishnan.

It is important to involve line of business leaders in this discussion, notes Daniel Aw, technology consulting sales director, HP, Asia Pacific and Japan. “You can’t do a big data strategy without understanding what businesses like.”

Once an enterprise embarks on big data, they start to understand what data they are generating and which are important now. In the future, they may want to do analytics on everything, says Aw.

Business analytics is the ultimate goal, he says, but big data is not just about analytics. We need to look at it as a holistic thing to do, says Aw. “There are a lot of things you need to do at the infrastructure layer, how do you store data?”

Aw says it is important to identify what kind of data the organisation has. “A lot of times, business people don’t know they can do analytics on certain information they have.”

For instance, in a retail store, getting real time information on a customer’s transactions will allow the store to make an offer that suits what the customers wants and needs.

He says among the most advanced customers in this space are in the financial services.

The finance sector has been doing business intelligence for the longest time, but around structured data. Their work continues in this space as banks transition into integrated multichannels, pulling out a lot of information on how they interact with customers

Daniel Aw

Krishnan, meanwhile, notes the very characteristics of big data are also the source of challenges to organisations.

Sheer volume is the first challenge, with estimates that by next year, companies with more than 1000 staff will have an average of 14.6 petabytes of data (a million gigabyte is petabyte).

Data comes in a variety of formats, and need to be processed across these different formats. Velocity or the speed in which data is processed is also a factor.

The design and implementation of systems to ensure availability, security and compliance of big data systems is critical. “You need to protect data from falling into the wrong hands,” says Krishnan, who says the average annual cost of information theft or cybercrime is US$4 billion.

 

Managing infrastructure trends

James Merritt, senior vice president and general manager enterprise group for HP in the Asia Pacific and Japan, points to the company’s focus on three big infrastructure trends impacting CIOs.

These are converged infrastructure, the cloud and the software defined datacentre

The first covers the different resource pools to help organisations reduce overall IR costs. The second involves public, private or managed cloud implementations to “achieve faster time to market or value”. The third trend is around customers using software to automate the datacentre and its workload provisions.

Mike Prieto, vice president and general manager HP Storage for Asia Pacific and Japan, tackles the next era of storage for a new style of IT.

“A lot of architecture was designed 20 years ago and can’t be modified to account for the changes,” he says.

He says legacy storage is not architected for current needs and workloads, and has led to an environment that is complex and expensive.

“Most companies spend 70 percent of time turning lights on rather than innovation,” he says. The aim is to simplify the complexity created over several years and which has resulted in storage sprawl and fragmented architecture.

 

Mike Prieto

Other announcements at the HP World Tour include:

 

Expanded converged storage portfolio: Features a purpose—built all flash HP 3Par StoreServe system that HP says delivers performance and low latency without compromising enterprise resiliency or adding datacentre complexity.

HP StoreOnce VSA software-defined storage solution for information protection: HP says this is a hardware-agnostic, software-defined storage system deployed as a virtual machine on existing industry-standard server infrastructure. HP claims it will reduce capital expense by up to 65 percent compared to dedicated backup appliances, and is the industry’s most scalable deduplication virtual appliance.

Scalable tape library for backup and archive in midsized organisations: HP StoreEver MSL6480 provides a“pay as you grow” pricing model that allows customers to buy additional tape drives to increase capacity and boost performance. The product is aimed at midsized organisations that may find it challenge to keep up with unpredictable data growth, unexpected spikes in demand for services, and variances in backup windows and retention periods, says HP.  As to the focus on tape, Prieto explains while tape is in decline, it is still the lowest cost option for having masses of data that organisations don’t need in a hurry. “[Tape] is going to be here forever but in a reduced capacity.”

3PAR storeserv storage all-flash array: This is a purpose built flash appliance that delivers enterprise class function but with mid-range price. It delivers over 500,000 input/output operations per second performance with extremely low latency to boost application service levels without deploying a separate storage architecture, says HP.

 

Divina Paredes attended the HP World Tour in Beijing as a guest of HP.

Join the CIO New Zealand group on LinkedIn. The group is open to CIOs, IT Directors, COOs, CTOs and senior IT managers.

Join the CIO New Zealand newsletter!

Error: Please check your email address.
Show Comments