Big data no longer mythical, but it remains a challenge

Big data is no longer a mythical beast customers are grappling with – but as they become more reliant on intelligence driven from analytics on large stores of data, new challenges are arising.

That's the view of David Gorman, Brocade director of solutions marketing, who says resellers are increasingly the front line for big data as customers, who now understand the big data idea, seek to find ways to 'put it all together'.

"The big data conversation has changed. It's no longer this mythical 'how do I do it', it's 'we're going to do it, so where do I start?'," Gorman says.

"It has become more of a regular data center discussion.

"It's no longer that huge step function of something we don't do today and have to figure out," he adds.

However, Gorman says the need for an agile, real-time system capable of handling analytics at speed, is something many customers haven't fully thought about, with successful big data becoming 'to some degree a performance conversation'.

"The promise is putting together disparate data sets, laying your data on top and seeing what you can get out of it.

"It's a question of scale and methodology. Customer's understand the idea, it's just a question of how to put it together and they're going back to the advisors they're used to talking to – and that's going to be the resellers.

Gorman says: "Big data and big data analytics are a lot of things and everyone comes to the table with a particular perspective, whether it's the analytics tools are the most important thing, or the storage environment is the most important thing, or the file system performance is the most important thing.

"What we want to get across is that they are all important because the system doesn't work without any of those.

"If any one of those has not been thought about, has not been built the right way, if you didn't have your expectations set before you chose the tools, something is going to be the fail in that process.

Unsurprisingly, Brocade's key focus is on the network component, and more specificially, data fabrics.

The vendor works closely with partners, including HDS and EMC, who provide 'the big data solution or a data analytics solution', while Brocade 'provides the connectivity', Gorman says.

In the case of HDS and EMC, both have Brocade switches in their offerings.

"In a lot of cases, the network is assumed because companies have networks. But that network might have been built five or 10 years ago.

"It's a fundamentally different world today when you look at virtualisation running on top of the network.

Gorman says big data also takes the conversation into 'a storage network conversation' with all the storage network issues, including resiliency, performance, low latency, redundancy, serviceability and automation.

He cautions too that there is a new conversation for customers around hyperconverged infrastructure.

"If they think 'I have a converged infrastructure network baked in' that's true within a rack. But what happens when you have two, three or five racks?

"How do you put those together with those very dense high performance workloads with big data analytics?

"All of a sudden, it becomes a conversation about what the network is connecting all of these things.

"You put those very dense, high-performing flash enabled workloads in these racks, put a bunch on racks in a data center or multiple data centers, but if you haven't thought about what the connectivity is, what you're trying to get out of those, then the network is going to become the thing you have overlooked and have to correct for later.

"What we're looking at is building the right data fabric first.