Acquisition gives HPE access to Nimble’s secret analytics sauce

HPE is buying Nimble for US$1b

Solid state storage manufacturer Nimble Storage, which is being acquired by HPE, is nominally, a storage manufacturer. However, its key differentiator, its ‘secret sauce’ according to the head of Nimble Storage Australia is its data analytics capabilities, which have applications wider than storage support.

Bede Hackney, ANZ managing director for Nimble Storage, speaking to Computerworld prior to HPE’s acquisition announcement, said: “I like to think of Nimble as a big data company that today just happens to sell great storage.”

When HPE announced it was buying Nimble for US$1 billion the company said it planned to incorporate Nimble's InfoSight predictive analytics platform across its storage portfolio, to “enable a stronger, simplified support experience for HPE customers.”

“For example, InfoSight automatically detects 90 percent of all issues within a customer's infrastructure, and resolves over 85 percent of them. This dramatically reduces the amount of time and effort a customer's IT team spends on support activities,” HPE said.

However Hackney said Nimble had already been looking at expanding the application of InfoSight beyond storage to other parts of the IT infrastructure. He explained that analytics had been a key component of Nimble’s product strategy since its inception in 2009.

“We hired our customer support team a year before we shipped a product, when we had no customers. From version zero of our operating system they were embedding telemetry and sensor points every few lines into the code so it could report back to the support team. It is very hard to go back and do that when you get to version six or version seven of the operating system.

“Now we have 10,000 customers, some with multiple arrays, and every one of our units reports between 30 million and 70 million data points back to our support centre every day.

“We have been collecting that data from day one and we now have the biggest repository of information about storage in the data centre. Nobody else collects data at that level of frequency and granularity. We have been collecting data from day one and applying analytics to that data.”

Hackney said the application of analytics to this massive volume of data had enabled Nimble to massively automate product support. “We have no level one or level two engineers. … Our number of support engineers per customer is much lower than any of our competitors.

“If a customer calls they speak to a level three engineer who has 13 months of data with one minute granularity about the performance of that array at their fingertips. We resolve issues in 41 minutes on average.”

He added that retaining historical data from every unit enabled the company to “inoculate” other units when it discovered a bug and to run scenarios on the impact of design changes. “When we hit a problem we can usually identify the piece of code responsible in minutes by doing a traceback on all those sensor points.

“We have had cases where we have identified a bug, created a patch within a day, gone to the analytics engine, searched it for other customers with the same scenario and proactively inoculated them before they have an issue.”

Hackney said Nimble had always seen its role as being beyond storage and he speculated on how InfoSight could be applied to other parts of the IT infrastructure.

“We defined our mission statement as resolving application delay in the data centre. So we defined our problem domain as everything under the operating system. That includes storage, host, network and even hypervisor. … At present we are applying InfoSight to storage, but it is interesting to consider what would happen if we applied all that data we have about people’s data centres to other challenges in the data centre.

“A year ago we took InfoSight and pointed at vCenter and started collecting information about vCenter and we were able to apply the same analytics and data science approach to analysing the VMware environment data,” Hackney said. “And we are on record as being focussed on developing similar capabilities for HyperV and even different application stacks.”

Hackney was unavailable to make any additional comments post announcement of the acquisition, but noted that Nimble was possibly one of the world’s biggest users of Vertica analytic database management software acquired by HP in 2011 and now owned by HPE.

Join the newsletter!


Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.

Tags storagebig dataNimbleHP Enterprise

More about ANZHPNimble Storage

Show Comments