 Live from San Francisco, it's theCUBE, covering Red Hat Summit 2016. Brought to you by Red Hat. Now here are your hosts, Stu Miniman and Brian Gracely. Welcome back, happy to have on the program CIO and Innovation Award winner here at Red Hat Summit, Kersi Tawadia, who's the CIO of the Bombay Stock Exchange. Thank you for making the long trip and welcome to theCUBE. Thank you very much. All right. Happy to be here. We read Innovation Award winners, I mean, you know, 10 to thousands of people, you know, so many people using Red Hat. You guys are in the infrastructure state. So can you tell us a little bit about obviously the stock exchange? I mean, I know that it's, you know, you know, high-frequency trading and, you know, latency is super important. You know, I mean, I think of, you know, proprietary hardware and lots of gear and, you know, networking there. So tell us a little bit about your infrastructure environment before and after, if you would. Prior to moving to the open source, we were working appropriately, mostly on the proprietary hardware. We were working on nonstop technologies. So everything was proprietary, not from the hardware, the OS, the application that we developed and any other supporting system that we need had to be proprietary. So there was nothing open to it. So we were running a challenge of making actually the high-frequency and the speed and the agility of what we wanted to change. There was a basic limitation that we were and we kept on innovating over there. Though we came from, say, we were earlier giving a response over 300 milliseconds, we came to a level where we came down to 10 milliseconds. But beyond the point, we thought that this hardware is not giving us the enough power that we wanted and our capacity was also, we were able to run only, say, 10 million orders per day. Okay, and when we were doing 10 million orders, our hardware capacity was reaching to 90%. So beyond the point, we thought, let's, there was no point in upgrading the hardware because if we continue with the same architecture, next two years, we'll again have the same problem. So we had to think about radically different of changing the entire architecture of the entire system. So we actually did the split about, everything was in a monolithic architecture previously. So we actually did all our periphery systems which are there. We first took out component by component from the open. So the main thing, we started putting it on our simple x86 systems. So we developed the entire periphery systems to an exchange, like in an exchange, you have the clearing system, the risk management system, the market data feed systems, and a lot of other systems which are there which requires matching engineers the heart. So what we did is first we created all the encillaries around it and then finally we replaced the heart. So when we replaced the heart, we took an open source, we went and entered the Red Hat stack. So our software was given by Deutsche Boards. So we took their software, we took the Red Hat and totally on an x86 box and we integrated the whole thing. And basically using our APIs, particularly what we professed as a microservices, instead of calling microservice, we actually had micro applications develop and connected everything through an API. So now we run almost 400 million orders a day with 10% of the capacity of the system utilization. So we have still got around 90% headroom available. Wow. So that's- 40x growth in how do you throughput there? And our latencies have come to six microseconds. Okay, so it's our acknowledgement. Can you bring us into, what was the board discussion? I mean, this is what you run your business on. It has to work, it has to be reliable and of course prices, usually prices almost thrown out the door usually when you think about that because- Not really, we're very, very cost conscious. BAC is very cost conscious because we had to be agile, smart and we had to spend our money smartly. So it was not that we had to throw our money, money was there, but we wanted to do intelligent investments and we have almost saved the money because the amount of money used to pay to the EMCs for all this proprietary infrastructure. In the same kind of cost, we have got the entire new hardware with three year warranty. So that is the kind of savings that we have done. So not only on the trading side, even on our big data engine, we used Hadoop as our basic platform for our big data and there also we have gone to 500 terabytes worth of data completely working on commodity hardware and totally open source and it's totally free for us and we are the largest exchange who actually made us live warehouse, big data, working on an open source. Most people are still trying it out or using this as more of a educational or research institutes. Being a financial institute, I think I'm quite proud to say that we have gone live, one of the first in India who have gone live at such a large scale. Yeah, what do you tell your fellow CIOs? What advice would you give them? A lot of them are saying, hey, we run critical applications just like you, but that kind of change is significant. The buck falls to you. What sort of advice do you give them? What's the thought process around going through that big of change, a big architecture modification? My advice is just to believe in yourself. You need to believe that open source works. I mean, that's the biggest, it's a mindset issue rather than a technology challenge, I would say, and when somebody else is successful, yes, you make a leader's role, you set a path for others to follow and I hope people will learn from our experience and follow the footsteps and they will come up because now people are waiting for somebody to be showing that it is successful, it can make it successful. I think BIC is a thought leader in most of these open source areas in India and we are leading and in a lot of areas, we are leading on the open source. I mean, we're doing a lot of innovations over there. Kersi, is BSE a contributor to open source? Very little, right now, we have just come into open source, maybe we were just two years back, but within two years we have done, we have covered a lot of road and a lot of success stories have been coming in, so we are still learning, so once we start, we'll start contributing also. It's absolutely no second thought about it. Yeah, can you give us any insight into just India in general as to the acceptance of open source and contribution? See, right now, the acceptance is a challenge, so only once it gets freely accepted, then people start contributing also. So right now, the mindset is a little skewed, in the sense people think open source is free, open source is unstable, open source is for learners, it's for labs, not for production, whereas here we are saying we are running a stock exchange, which is mission critical for the country, and it's considered one of those countries' most critical infrastructure, and if we can run the country's most critical infrastructure on open source technologies, any other person can run that on open source, so I think MEC has gone and set up, and now slowly it's catching up. People are looking at us as role models, as examples to be setting up within the country, and it's happening. You talked a lot about the technology change, what about the people side, how much did your team have to change over the last couple of years, what new skills did you have to integrate? See, learning open source is not very difficult, because most of our support that we got from our partners, and specifically we had a strong domain knowledge, because stock exchange is a very, very strong domain, so techies are usually very fast in learning technologies, but not the domain, and so basically what we did is we reskilled all our guys who were there on the old platform, we just trained them on the open source, and they are doing quite well, because bringing in a new, totally new talent from outside and working in a domain was a difficult choice, so we chose the other way around, we just reskilled everybody who was there, and it was a motivation for them also learning something new, getting on the new platform, getting more agile, so I think it's got very well in the organization. Chris, I want to give you the final word. Give us your thoughts as attending Red Hat Summit, being an innovation award winner. I think it was a very good experience. This is my first event, like I said, we've come into Red Hat just a couple of years back as an organization, and it really feels good, and it's happy to be getting honored. I'm feeling very happy that we came in so fast, and we never expected that our work would be recognized in such a large scale, so this recognition has given us even more confidence that even we can do even much better than we can freely contribute now, and we are feeling very proud to be associated with Red Hat. Thank you very much. All right, well thank you so much for sharing your story, and congrats to BSC on being an innovation award winner. We'll be back with lots more coverage here from Red Hat Summit 2016. You're watching theCUBE. All right, hello. My name is Stephen Keating, and this is August 18th. Round six p.m., and I...