"Data silos are good" according to Splunk CEO Doug Merritt
- 25 October, 2019 09:39
Doug Merritt (Splunk)
The CEO of machine data specialist Splunk is doubling down on his belief that data silos are not the enterprise bogeyman they are often made out to be, as he positions Splunk as an agnostic platform for collecting and making sense of machine data, whether that resides on-premise, in the cloud or even in your SaaS environments.
Talking to the press after his keynote during Splunk's .conf in Las Vegas, CEO Doug Merritt said: "One of the aspects that we've talked about today is this belief that silos are bad. I think, a lot like the belief that you must schematise data to use it – which we blew up with Splunk Enterprise – we've got to get past that silos are bad. Silos are good."
Splunk started life as a big-data analytics platform for IT professionals looking to index and make sense of their logs and machine data, but has since pivoted into security and other business areas.
Although data silos – pockets of data spread across a business – are often positioned as something that need to be broken down, Merritt is keenly aware that most of his customers will live in some form of a hybrid world, where data resides on-premise, in the cloud and increasingly within SaaS tools like Salesforce, Workday and ServiceNow.
"You've got a whole host of distributed components that need to talk effectively amongst each other, and they're going to all have their own data sources, you're not going to data lake the entire thing," he said.
A great deal of the Splunk product roadmap, then, has been focused on "allowing silos to exist, because different departments are going to go through digital transformation at different times, and with different desired outcomes".
Picking up on this thread, Splunk CTO Tim Tully added that he has been focused on developing monitoring and alerting software that is infrastructure agnostic.
"What I think I've learned, probably just in the last few months is: I don't really think about on prem or cloud anymore, or hybrid, for that matter, it's just all the same software for us," he said.
For example, the newly announced Splunk Data Stream Processor tool for streaming data analytics has been built on a foundation of Docker containers so that it can be deployed wherever a customer desires.
"It's totally agnostic in terms of where it runs. We run that thing in the cloud in our own Splunk cloud platform, but we're targeting for on prem customers first right now, so we're shipping everything in a massive binary – that's just Docker containers," he said.
As this is Splunk's biggest conference of the year the vendor took the opportunity to announce several new products and features, with multi-cloud functionality a recurring theme.
The vendor announced that it was adding VMware vSphere and multi-cloud monitoring (in beta) to its IT Service Intelligence (ITSI) tool, as well as enhancements for Windows, Unix and Linux monitoring, regardless of if those operating systems reside in physical or virtualised environments.
This focus is reflected in Splunk's recent M&A activity, where it acquired SignalFx, which specialises in real-time cloud infrastructure, microservices and application monitoring, for $1.05 billion earlier this year.
It also acquired Omnition to boost its footprint when it comes to microservices and the open source community. SignalFx has since been integrated with the Splunk Cloud monitoring tool.