from San Francisco it's the queue covering Red Hat summit 2018 brought to you by Red Hat hey welcome back everyone here live in San Francisco at Moscone West of cubes exclusive coverage of Red Hat summit 2018 I'm John four with mykos John Shroyer founder of tech reckoning advisory and on community services firm our next guest is Jason O'Connell openshift platform owner of mark McQuarrie group welcome to the Cuse let's get it right that's right well the retail bank of Macquarie so thank you and financial services thanks for coming on so bas lead banking is pretty hot big-time early adopter of all things tech yes and you doing a lot of work at kubernetes tell us about what you're doing take a minute to explain your job what your focus is some of the some of the environment DevOps things you're doing that's the basic I'm head of the container platforms team at Macquarie Bank so basically my team manages open shifts on AWS we do the architecture on there but we also focus a lot on the value add on top so we don't just give our our customers for my team are the developers and the development teams we don't just give them a blank platform we do a lot of automation a lot of work on top of that basically because we want to make sure that the idea of a platform as a service is that we do as much as possible to make developers lives easy tell about the journey when did you start on this effort Asli Amazon's great cloud we use it as well other clouds are coming on you got Google and Microsoft and others but when did the open shift conversations start happening where were you what year was it how long have you been using it it's gone through some great changes I want to get your experience on that opened she have to journey I mean somewhat of an early adopter I mean we started looking at this two years ago so that was openshift 3.1 a lot of the basic features weren't even there but it took us a year to both build it out as well as migrate about 40 applications to production so there's only a year ago that we've been in production so it's evolved like so rapidly during that time so 40 applications migrating right that enough in and of itself in a year is is a pretty heavy lift can you talk a little bit about are you just replied forming the applications obviously probably not rewriting at this point the open shift has been a good home for the applications that you started out with it sounds like I mean one of the reasons to choose openshift was docker and it was about that migration path I mean part of the migration was ensuring that developers could get everything running locally get these legacy systems we did a lot of micro services running locally on docker containers on their laptop then the migration was was easy from there but we deliberately didn't want to do like a lift and shift we wanted to rethink how we delivered software as part of this project okay what's the biggest challenges you had in doing this I mean as you go but she has got some great movements you could burn aces a good bet and kubernetes is looking like a really awesome way to move workloads around and manage containers and clusters so you know what's what are some of the things we've learned what are some of the complexities that you overcame can you share a little bit about some of the specifics I think I think the newness is is probably the biggest challenge I mean going back to two years ago there were some very basic components that weren't there at the time when we knew were coming and even now there are pieces of work which we just don't tackle and we do a very quick fix because we know it's coming later I mean it's just moving and evolving so quickly you know we're waiting a lot for sto which is coming in the future so we're holding back on investing in certain areas because of that so it's always a constant challenge yeah I still looking good and the service mesh is hot as well how has OpenShift helped you but what's the what's the if you had to kind of boil it down what's the been the the impact to you guys where's the where's that coming from I mean before we even selected OpenShift we had we're looking at our objectives from a business perspective not a technology perspective I'm the biggest objective we had with speed to delivery you know how could you get a business idea a product idea into production as fast as possible or even if you look at a minor fix to something something that should be easier develop it takes a data ride why does it take a month to release the production so speed of delivery was one of the key objectives and I can tell you more about how we we delivered that in detail but just going back to the objectives we also looked at developer experience you know sometimes the developers are not spending enough time coding and doing if they want they get bogged down in a lot of other pieces of work dinner I'm really delivering business value yeah so again we wanted the platform to handle that for them they could focus more on their work and this is the promise of DevOps and the whole idea of DevOps is to automate away the hassles and I mean my part to Dave a lot that calls a rock fetches no one likes to do all that work it's like can someone else just handle it and then when you got now automation that frees it up but this brings up the thing I would love to get your reaction to because one things we've been covering and talking a lot about in the cube is this is been happening around us it's not just what we're doing but this new modern way to deploy software you look at like some of the big things that are happening in with cloud native and you mention SEO is to do this awesome dynamic things on the fly that are automated away so it changes the how software is being built how are you guys embracing that what's the thought oh so you've got a team that's got the mindset of DevOps yeah I'll see embracing this vision and if everything else is probably substandard she'll you look at you know waterfall or any kind of non agile what is the your view of this modern era of writing code and building applications what I mean for people who don't aren't getting it how are you how do you explain it you know I think it's I mean it's an unbelievable time that we're in at the moment I mean the amount of automation that we're doing is huge and part of our openshift is that it's an automated bull platform so I've got a few junior guys in my team they're like two graduates and in turn they do a lot of the automation yeah it's that easy now everything's got API so we can connect everything so I do find when we interface with some of the older school teams in different parts of the bank that aren't doing this level of automation they used to manual processes and manual ways of doing things and now we look at everything where everything can be automated that's thing you really truly feel now opened up that you could automate absolutely everything I mean the developer productivity one is key you know state of mind is another I mean the mood is better okay people are in a better mood more productive yeah and I think if you look at interestingly in like security and risk teams and governments teams where we're finding look they can improve security risk and all this by automating you know they're the one set and now we've got SEC offs movements and things like that so speed of production is is does not prohibit better security and in fact with Sec ups the amount of automation we do you got a far greater amount of security because we now know everything that's deployed we can continually scanning for vulnerabilities yeah what so Jason you talked about it being new we've talked a little bit about culture how much of this has been a training exercise how much is that it's a cultural shift within your organization as one of the leaders of it how are you approaching I mean we're lucky there within Macquarie Bank there was a large scale culture shift towards agile where the whole thing runs in that gel manner so that's helped us then feel in our technology and automation it complements that way of delivering so we've got some very unique ways where we've done automation and delivery which completely rethinks how we used to deliver before so example yeah for instance now if you think why were people scared of delivering something into production why was a small change scary change and a big part of it is the blast radius if something went wrong you know connecting through to our API is we've got our own channels we've got mobile apps got a website you've got a lot of partners there are the companies connecting through as well and so even if you did a small change if it costs an issue everyone's affected at once so a big piece of what we did to deliver faster is allowed targeted releases you know I could target a release and a change just to you we could target it to a percentage of customers monitor rolled out quickly if there's a problem dial it up if it's looking good good target to any channel it seems like there's a business benefit to that too oh it's massive here because you also can promise stability on certain channels if you want you can have faster channels that are moving quickly and in an API driven world we've got external companies connecting through to these api's you want to be able to say that we've given you a stable offering and you can upgrade when you want and then our channels we cannot move more fast so we've got mr. no-brainer I mean really the old way is completely dead because of that because you think about the blast radius you're pointing about blast radius the risk is massive so everyone's kind of on edge all these tests have to go in redundancies as if the planning is ridiculous all for the risk well that energy you're optimizing for a potential non-event or event here with micro-services and you and app can go down to the granular level the granularity is really amazing so when you go forward first of all it's a recruiting opportunity to get better engineers wait this is a way we work I'm going forward I want you to comment on your opinion as an industry participant and can clarify this because a lot of you'll get confused here automation they think jobs are going away administration is getting automated system admin type roles where junior people can now do more operating things but the operating roles not going away so talk about that that ops side because now the ops are more efficient the right things are audited me you but talk about that dynamic between the right things being automated and the right things that are gonna roll to operational service meshes or whatnot yeah I mean basically it's about getting people to do these higher-order functions so the people who are doing things manually and operating things manually you look at our Ops teams now morphing into like the classic SRE team you know the side reliability engineering teams where they're spending a significant amount of that time automating things you know looking at alerting and monitoring then Auto healing I mean it's actually more work to automate everything but with a far greater amount of quality and reliability when we go and the benefits are long it's worth it basically you do the work upfront and you reap the benefits and then variety of ways like writing rolling out software managing workloads talk about multi class here on Amazon multi cloud is a big focus to your hybrid cloud multi-cloud obviously we're seeing that trend how do you look at multi cloud as a practitioner what are some of the things that check our check boxes for you in terms of ok as we start looking for the next level there might be a multiple cloud scenario how do you think about that and how do you put that into perspective that's worth noting even two years ago and we selected open shifty it was with the idea that we could go multi-cloud you know that for the users for the developers they're not going to know the difference where we run it on so we're not locked into any provider I mean at the moment we're kind of just exploring Google cloud and we're looking at what it would look like so even we don't know yet some people have spoken about stretching your cluster across to clouds that means one cluster across two seems very difficult to me that a lot of latency issues potentially there's also cloud arbitrage you know can we get certain workloads on a card that's cheaper can we use spot instances can we spin things up and down we're on Google it's cheaper and then it also raises questions around okay do we need Federation and we know Federation has been talked about a lot with kubernetes how do we manage so many clusters and even on AWS now we have three production clusters you had multi clouds how am I gonna manage that what about the services layer of clouds right obviously the Red Hat platform gives you a services layer that could run anywhere but underneath that right AWS has its own services layer Google you know a lot of AI ml you know it could you be able to are you thinking about taking advantage or how are you thinking about those different offerings on different different places I mean this is the challenge I face and what we're exploring is that do some teams have the differentiating services the unique services that they want on Google especially for managing data machine learning we know those services are key for them some teams will have that but yet then can we call them over from AWS even oh do we have to deploy in in Google and have that in one data center can we go across with services so it's really like not just cloud AWS cloud Google but it's actually criss-crossing that's another thing we're exploring Jason thanks for coming on the cube really appreciate your commentary I've seen multiple red hats you guys have won awards you've been here before great job final question for you if you could boil down OpenShift into kind of like a sound byte for you what does it mean to you guys what's been the benefit what's been it it's been that what's been the role what's the benefit of openshift as you explore the cloud journey you know I could say speed I could say automation I mean that's huge but but really OpenShift and read how to pick the winner which is docker and kubernetes and a colleague of mine is in pucon in copenhagen last week he's constantly messaging me saying there's new tooling you guys can use this you can use that and a means that rather than us doing the work we're just getting tooling from the community so it's the de facto standards so that's that's probably the biggest benefit all the goodness is just coming right to your front door likely and I got to do my homework every night playing around with this technology so yeah great success story and again the great community open-source projects out there you guys can bring that in and productize it for the retail bank congratulations love open-source stories like this tier one citizen and again continues to power the world open-source softens the cube doing our part bring and use all the data from Red Hat summit 2018 I'm John fryer with John Tory we'll be back with more after this short break