CBR Q: The latest Red Hat Storage Server is based on open source GlusterFS 3.6 file system and Red Hat Linux 6, what are its advantages versus proprietary solutions?
Ranga Rangachari, VP and GM storage and big data, Red Hat, A: If you look at the traditional storage industry, it’s essentially software that’s been wrapped with a proprietary piece of hardware. But underneath the covers, it’s just basically another x86 server with a version of BSD, Linux or something. And the vendors charge a huge premium for that integrated solution that they get and the users cannot take it apart. Yes, they do have intelligent software but they can’t buy the software away from the hardware. Our approach is dramatically different, which is take advantage of industry standard x86 servers that you would get from a HP, Dell or IBM, and lay it in our software. So the two aspects in how costs really come into play is that we write the commoditisation in terms of the price of these servers as they keep coming down. The other way that with our solution, you don’t have to pre-buy, so you could start with 500TB of storage and then as your workload continues to increase you continue to scale out. Whereas with the classic proprietary way, which is whether you use it or not, you still have to buy it in chunks of 100TB or 200TB. We don’t have that limitation. We give users flexibility in saying start where you want to and scale out as you grow.
So we primarily take advantage of the price point that x86 servers offer plus the value of our software. We don’t have perpetual licences or anything of that kind. Every RedHat product is sold on an annual subscription basis.
CBR Q: What advice would you give to CIOs looking to migrate from proprietary technologies over to Red Hat?
RR A: The advice I would give them is to say that the value of open software defined storage is here and it’s proven. If we use Red Hat as a data point, we have lots of customers with references around the world who are actually using our solution. Also, since it’s a new journey, I’d like to say, ‘think big, start small and scale fast’. For example, if you’re running an Hadoop workload or you’ve got a bunch of log files that all your machines are generating that’s an ideal first step. Try and get the economic advantages without compromising on control, and then as your data grows, continue to increase your footprint.
CBR Q: What kind of skills do companies need to deploy this solution?
RR, A: You don’t need any extra skills to do this because we’ve masked all the complexity. This is not a desktop product or an end user productivity tool. This is more a classic IT in the data centre type of a tool. But the basic skills centre on deploying cloud or deploying x86 servers.
CBR Q: To what extent has your portfolio grown since you acquired Gluster back in 2011?
RR, A: The emerging products, which everything that doesn’t include Red Hat Enterprise Linux, that’s actually growing faster, almost 60% year-over-year, across the board. So it’s an integral part of our product portfolio today, especially as customers start to move towards cloud infrastructure. Openstack is pretty much becoming a huge de facto cloud implementation today and Red Hat is leading the pack in terms of contributors, market leadership and others.
But storage for Openstack is absolutely key because as you scale out compute you want to scale out storage. Today, with the Gluster solution as well as the acquisition of Inktank (May 2014), we do have a full spectrum of data services running with an OpenStack, whether it’s object, block and file system storage.
CBR Q: You entered the market approximately two years ago. Where would you like to see yourself in the future in terms of market share?
RR, A: It’s very simply. The compute space industry 10 years ago was dominated by proprietary chip sets with proprietary operating systems, but what Red Hat and specifically Linux did was completely disrupt that whole market by allowing enterprises to take x86 servers and run Red Hat Enterprise Linux, and now today we have a huge market share on the operating system side of things. And it’s an exact same type of thing with what we want to do with the storage space, which is disrupt the market through community driven innovation and the volume commoditisation that’s happening with x86 servers. So we want to do to storage what we did to compute a decade ago.
CBR Q: So how do you find competing with those traditional players at the moment?
RR, A: The fact that those traditional hardware proprietary guys have been in the marketplace for 10 or 15 years, they do have an installed base. Our focus isn’t to go back and say, "Rip and replace and start moving your data to Red Hat storage portfolio". Our strategy is to augment what you have with a better, smarter and hopefully cheaper way to solve this. So it’s augmentation as opposed to replacement today and overtime I think that the customers will see the benefit of what we do and an new workloads start to come in, I think we’ll show enough value that they’ll continue to buy more workloads on our platform than the proprietary vendor platforms.
CBR Q: To what extent is your recent partnership with Cisco helping you to expand?
RR, A: We’ve created a reference architecture along with Cisco to help customers deploy a solution that includes the Cisco hardware, our software and Splunk software, specifically focused around cyber security. So the advantage for the cost, they can pack a lot more storage, can analyse a lot more data, and the more data you analyse the better your results are. So that’s an example of how you continue to build the ecosystem, to work together with ecosystems so that when a solution lands at a customer’s data centre it all works together, there’s a support model behind it, so customers don’t have to got to 18 places to get support.