 So, today so we have been actually discussing lot of development, quality, pre-production. So, what I have been talking mostly is once agile ends and when production starts. So, I have been a production, operations, engineer, architect, concerted for last 16 years. So, once the code moves to production what happens? So, I would present that picture and I would be a continuation to show it like show the whole thing of the pre-product. So, now mine is pure product what happens is pure product. So, feel free to stop me in between if you have any questions. So, that is about me what I do yeah. So, like I have been working with local clients for June 500 clients across geographies and across industry verticals finance, banking, retail, services, energy. So, whatever industry verticals you can think of I have covered all. So, that is what is my experience all about. So, currently emerging technologies I have a very keen interest on security, DevOps, AI, IoT, blockchain analytics. So, you will feel that what a production operation guy is doing in all these emerging technologies. Because end of the day once in developers and QAs, once the developer develops that cutting edge, bleeding edge technology and once the QA tests it properly finally, it is going to get hosted on production. So, that is where I need to be aware of like what is coming, what we finally get hosted on the production which I have to take care which I have to architect for. So, that is where I have keen interest. So, my hobbies are I am a active cyclist, I read a lot of books both technical, non-technical and also I do a lot of traveling. So, that photo that you see there was a solo travel in Orissa. There is a place called Gattak, there is a place Puri. So, I just had a attiva and I was traveling solo. No friends, no mobile, no internet, nothing. So, just cashing hand, buying and my contact gets it, nothing else. So, that is the photo from there. So, typically when you are in a group, you have group dynamics, somebody wants to do something. When you are solo, you do whatever you want. But again, you do not want to act like that. And if you happen still, you are not. Next. So, that is my LinkedIn link, anybody wants to connect. So, this is a very interesting video, whether it is play or not, I am not sure. Ok, it does not play. So, actually I will just tell you briefly about the video. So, this video actually compares how formula 1 race used to be when formula 1 started. Like a pit stop for a formula 1 car back when it started in 1960-70s, used to take a lot of time. And currently the formula 1 pit stop takes like few seconds. So, that is what the video is all about. So, again it will not play. So, that is what is agility. So, how the world is becoming agile from all perspective, not only software and hardware, even the normal lives have become agile, that is what the video is doing. So, that is what pace of change is becoming in our economy is accelerating very fast. That is what everybody has to pace up. So, when the developer is developing fast and when the QA is also testing fast, then the production team also has to be ready to deploy it into production fast. So, that is the way to where we are also getting into the agile way. Or else all the cold, all the latest cutting edge code you develop sits in pre product. It does not be production, it is not ready for production. So, same thing. So, as you may have noticed in the last like maybe 10-20 years there has been lot of changes. So, at some point in time in IT. So, currently all the people who are here. So, how many of you are on cloud? So, rest are not on cloud. So, how many of you here have a Gmail account? Anybody who does not use Gmail? So, let me just tell you, everybody knowing you or unknowing you is on cloud. So, the Gmail that you use, the Facebook that you use, anything everything in our life hits cloud. In some way or the other. How many, one more poll I want to do. How many people here feel is mainframe not dead? Or mainframes will be dead in the next 5-10 years? Like containers will take over. Anybody who feels mainframe will be dead in next 5-10 years? Rest are free, mainframe is going to be there? Ok, a reality check. Currently, any 19 transactions you do, any credit card you process, 95% of those transactions are processed by a mainframe somewhere in the world. So, mainframes are still there, they will remain. Ok, so it is not going to die anytime soon. So, IBM still exists and still also all the people who work on mainframes, it is not going to die soon. It is just a reality check. Ok, software is hitting the world, everybody knows this guy, I do not know him. So, we are in 2018. So, mostly what we do is, whenever we present, we give data of 20-25. And 20-20, 20-30, if I am not doing that, I am putting 20-18, we are in 20-18. So, we can actually check whether this is correct or not. So, here it says that one third of the top 20 market share leaders will be disrupted by the third platform. So, what is the third platform? So, this is the third platform. So, all of this is impacting anyways, isn't it? All of the companies. So, Uber everybody knows how it is disrupted the taxi industry. And all taxis now are forced to join Uber, there is no choice. Or else there is no business. So, does Uber own any taxi? So, means just a hypothetical situation. Let us say Uber decides to close down in India. What will be the exact time frame of Uber? Any idea? Any guesses? Any guesses? Means how will it take one year, two years, five years, six months? Any guesses? If Uber decides to close down in India, I will give an answer. It is just a one thing. All the legal procedures done, they will switch off their servers which are on the cloud. The app stops working game over. They don't only need taxis, all their offices are rented, their office space are rented. So, that's it. The exit is so easy because of the third platform. So, they own nothing. Only the app is this. Okay. This is the odd sea or so. I need not say more. So, it's talking more about DevOps. So, I don't want to spend much time on DevOps because we have been listening to DevOps since yesterday. Everybody knows what is DevOps. So, I will focus more on Ops. Okay. So, typically my presentation was mostly for the Ops crowd. So, I speak in conferences where there are more like storage experts, network experts and cloud experts. So, they want to know more about the depth. So, here I will talk more about Ops. So, that's how everybody is aware. That's how is the industry leader, third party like analysis company. So, they have come up with a very nice model. This is called a life model. So, currently only if you are new to its startup, then you are completely on cutting reading edge technology. Rest if you are a large enterprise with a lot of legacy and a lot of technical net. So, you would be running my model. So, your old applications which are still running good, you will not re-factor them. You will not take your application which is running nice without a single day or single minute downhill. You will not convert it into a container because there is no business sense out of it. Because it is working fine. So, let it work. Whatever new I have developed today for the social platform, for the artificial intelligence platform, that will be on the new platform. So, that is what is my model. This is the model that all enterprise in the company. So, this is very very critical. So, you can read whatever is written. Yeah. So, reliability and compliance is what ITIL is. So, ITIL is a predominant is operation framework, which is the operations world for us. Development and best environment don't follow ITIL. So, DevOps is what the development and QA follow. So, this is the start difference between both. So, reliability compliance and rapid development. So, here adding on to the DevOps. There is a very critical component security. So, since today morning, I think so if some of you attended the hacking or the hacker event where a Bluetooth device can be hacked, IOP device can be hacked. So, security is a very critical component. So, I have been very active in many security forums. So, what I predominantly feel is that the developer and the QA community are not that much exposed to security aspects of any forum. So, what I would urge all my developer and QA friends out here also have security in mind. Because if your code has vulnerability, those will be exploited very easily. So, because definitely once you need productions, there are security experts who will check the security aspects of things. But however, code vulnerabilities are tough to detect. So, that is what I feel that a security aspect should be there to DevOps. So, I know everybody knows this. I will not discuss this further. What is DevOps? So, DevOps everybody has their own perspective. Depends on what background you are coming from. So, that is an elephant for all the people. So, this is the actual scenario. So, when we say DevOps, DevOps are never friends. So, anybody here who, how many people here from Dev, Dev QA? How many people ops? Only one or just a few. So, do DevOps like each other? Do they love each other? Do they go on outing and not the company paid outing? Personal outing, do they go together? Do they go like on Friday night, drink parties and all? They hate each other to the core. It is the company which keeps them together. That is what is DevOps. So, I have been ops. I always fight to the death. Your code cannot be deployed. This is the problem. This is the challenge. This is the security problem. This particular code is open. This is in something or the other. Something latest not approved. This version not approved. So, all that is the constant fight going on. So, this is the actual scenario. So, difference is that developers want freedom. They want to try everything new. Whatever comes on open source, whatever comes on GitHub, they want to incorporate everything into their code. And operations, they want to everything. They want standardized, secure and managed environment. They do not want any troubles. No downtime. So, that is the reason. Operation guys always run n minus 1, n minus 2. Whatever is the latest release, it will be always n minus 1. You will wait for the patches to come, for the update fixes to come. Then only we will deploy. So, same things I will do one. Conversations. So, basically we have to enable communication between both. Both have to talk to each other. That is very, very critical in DevOps and with trust. In the developer should trust the operations. That an operation should also trust the developer. That both are working for the same application. They are not like two different teams. So, this is what is the photo shows. And we have to develop a pattern of playing together. So, when we are in a sprint, we are working for the deadline for the sprint. So, again we have to strike a balance. And finally, this is like ideal state when you are sharing and you know each other's pain and you are helping each other resolve the problems. And then comes honeycomb. So, this is the what you call as soviet criteria told that is a dream state. So, this is I do not know which companies actually have it. In reality. So, on paper everybody is doing it, but on reality I do not know. Now, we want to upside. So, majority of the budget in operations is spent on operations which is keeping the lights on, maintenance, keeping the lights on. That is 80 percent of the budget is built in. So, there is no budget left for any innovation or trying out new things in operations. So, how IT evolved? So, here anybody has worked in Waterfall? You people, good. So, then some people know what is the difference. So, when there was Waterfall, all applications were working here and here in physical servers and all company had their own data centers. And all Dev and QA had their own server farms. Nobody can touch their servers. The servers keep running for years together. There is 0.1 percent activity of that server when the test is done, the rest of the time it is just powered on. That is where I am coming from. So, I have had those servers as well. So, that is the reason I have put it. So, that is why I came to Ajay. Ajay made it multi-tier, virtual servers and from your own data center you went to a co-located data center. Like somebody else in the data center hosted. So, you host on there and you pay rent. Like how you pay your rent for your house. You pay rent for a data center. That is called hosted. Then came DevOps and microservices and containers and so on. So, currently we are here. However, there are many companies, many applications which are still in the top one in the middle one and they are transitioning. So, once those applications die then the new applications will follow the latest trend. So, same thing pretty much. So, these are few like sample architectures from operations once they are deployed. So, first one is just a waste with the app. Second one has virtual machines and third one has got containers. So, three types. So, typically what happens is previously everything was a silo. Like I have an application. I have my server, my storage, my network. So, I have budget for my application. So, I will buy stuff for mine. So, that used to be the case. So, even if I am using it 20% unless 80% I will not share it in any other application because this is my budget. This is my server. That used to be the case type. So, then came the demand with agile DevOps, elasticity, flexibility, rapid deployment and to reduce complexity. This was the demand. So, as I told you, server, storage, networking, power and traditional IT then came virtualization and converging infrastructure. So, I will be discussing converging infrastructure going forward. And then came flash storage. So, everybody can understand what is a flash? Who doesn't understand? Anybody knows the solid state device SSV? So, flash is the operations term for SSV. It is capable of high performance. And then you also have hyperconverged infrastructure now when what happens is the compute storage and the network can be on a single rack. In a data center you have racks. So, on a single rack you get a compute storage. That is a hyperconverged. And then also there is orchestration and automation. So, orchestration and automation will actually reduce manual effort. Will reduce system administrator, database administrator, network administrator who are manually configuring new systems, updating patches. So, all those aspects can be automated. You do orchestration and automation. And then anyways agile infrastructure is the end state where you are truly agile. So, this is typically the architecture that you see. That is like a very complicated architecture. Everything now can be done on a single rack. That is what I wanted to do. And then there is a very interesting concept network function virtualization and software different networking. I know there was a session before me. If anybody attended that. So, physical network devices are getting converted into virtual machines. The virtual machines will actually do the network router, network switch, load balancer, firewall, all these activities will be virtualized. So, that is what is software different networking and network function virtualization. That will bring in a lot of cost optimization because hardware devices from proprietary vendors are very very expensive. This will bring in a very cost effectiveness. So, again I will skip through. Because I don't have much time. I won't be able to explain the whole architectures. So, benefits I already covered. That lower capex and optics. Capital expense and operating expenses in operations. So, virtual machines, how they were. And containers. This everybody knows. So, I will not spend any more time on this. Everybody knows what is the difference. So, main benefit of containers in a operations world is that wherever they find a Docker environment, they can run. That's the. And they can run on pretty much everything. Any of these environment which can run a Docker layer can post a container. That's the benefit from operations point of view. For a container. Yeah. So, Lidards, Windows, Mainframe, Amazon, Azure, any more. So, whichever offers a container service, they can run there. That's the beauty. So, you have can go anywhere. They are not hide down to a particular like it has to be on a Lidards box. It has to be on a Windows box. So, all those dependencies are lost. So, you can move between infrastructures. Okay. That is the beauty that I see with containers. But however for containers applications have to be containerized. In the sense it has to be Microsoft services based and your traditional SAP Oracle, they cannot get into containers. All that still problems still exist. So, scale up, scale up everybody knows. You increase the memory CPU storage or you increase the number of nodes, put it on a load balancer and scale up, scale down or scale up, scale in. Then comes automation as I told. So, automation, what happens is every team is doing their own automation. Database team has got their own scripts and server team has got their own scripts. Everybody is running their own automation. So, those are called like islands of automation. Then we have those new automations where multiple team automations are tied together using orchestra. So, they are like of this flow. So, that is the idle script, orchestrate. Where everything is interlinked and everything moves in a flow and gives you the desired result without any manual automation. Full automation. So, again I will skip through. This is showcasing the various areas which can be automated together using orchestrate. So, what are different tools from disparate vendors. So, no enterprise has a single vendor policy that I will only have IBM or Red Hat or HP or whatever. They have products from all over. So, orchestrators which can actually automate things across the board are required for effective automation. That is what is being highlighted. So, this is a used case. Again I have very less time. I won't run through it. It is an end-to-end automation for a process where a person used to take maybe few hours to do it. Now it will be executed in a matter of seconds. So, with this is the infrastructure problem solved? No, because if you are building your own infrastructure, own server, own network, own storage, very complicated as you can see. So, this is just a diagram. If you see a data-centered architecture diagram, you will fall off. It is so complicated. So, you need some guy who understands everything big. And on that also there are chances of errors. If there is any error, the entire performance latency everything goes in for a time. Ok, it is very complex. So, then comes cloud. So, now everybody knows that cloud is like solution to all problem. All problem in this world, any issues you face just put it on cloud. And cloud gonna take care. So, just a reality check again just like the main frame. So, yeah, so there is no single cloud which can meet like all other requirements. Clouds, there are different types of cloud, different varieties of cloud based on the different workloads, based on the different use cases, industry verticals. So, that expertise is where a cloud consultant architect practitioner brings in on the table. He knows it. He can help you like it. So, this everybody knows they are not covered IS, SPAS and SAS. So, here comes different types of open source and some vendor cloud platforms. So, I will just skip through. So, which cloud to go to? Is it Amazon? Is it Azure? Is it Google Cloud? So, that is the question. So, why you want to go to cloud? And what is the workload that you want to go to cloud? All that questions have to be asked before we can decide which cloud suits you. There is no single answer. So, developers might prefer Amazon cloud or whatever. If it is a Docker developer, he will go to Azure cloud. If he loves Google, he will go to Google cloud. But operations doesn't work that way. Operations have different parameters to use the cloud platform for deploying production apps. So, these are various types of cloud computing patterns for production. So, how the burst comes, what is the load. So, accordingly we take a call, we decide. So, which workload goes where? This is a very, very critical component in cloud architecture and cloud server. So, SAP can go where? Oracle can go where? Because enterprise apps are typically not, they have got a vendor component involved. Because there is a lot of stringent SLA sign, there is a lot of penalty clauses those are signed between vendors and clients. So, hence these applications has to be decided. Which particular cloud platform is this? So, traditional IT, cloud data, already I have covered the buy model. This is into more details. So, benefits of cloud everybody knows, I have not covered. So, there is no capping, no upfront buying of anything. You pay as you go, like your mobile bill, like your electricity bill. But you have to be careful, because the bill can be very high as well. Scale up down, you can scrap rate I think. So, once you boosted your application on cloud platform, how much time will it take to move to a different cloud platform? Let us say they offered you 50% discount. Is it as simple as your mobile event things? I can move from attempt to whatever you said to us. This is okay. So, depends on your application landscape, depends on the count of servers, virtual machines, containers, complexity. It can take years for large enterprises to move from one cloud platform to another. It is a huge, huge effort. So, serverless, everybody tell me about serverless? Many people don't know. So, serverless, I just added few points here. Because current needs to trend now that all ops guys will be fired, and ops won't exist any further. There will be only developer and servers. That's the new thing coming in. So, I don't have much time to go detail into it. So, serverless is more like a function call. So, if your application can do only function calls and rely on function calls only, then you don't need any to manage any server, anything, neither in Amazon nor in Azure, nor in private cloud, nor anywhere. You just consume the function calls, pay for the function calls, that's it. So, your virtually means the ops is eliminated. No more fights, no more fights between developers. So, these are few of the benefits. Again, I have last few minutes. But the challenge in vendor working, do you feel that once you move on to a particular cloud platform, you have vendor work? Or do you feel that now you are free to move out immediately? Function as a service, our serverless is tied to that particular like vendor. You cannot immediately port your port, your port doesn't work. So, all these security is a major challenge, enterprises is going to move. So, again from operations point of view, high availability is very, very good. We sign legal contracts on high availability, scalability and monitoring. So, this is infrastructure, it's already covered. So, this is all DevOps stuff, it's covered. These are few of the sample like architectures for few industries, just like a 10,000 feet away. That's a, this one was the insurance industry on Oracle. This is healthcare, VDI and this is for online vendor work. So, that's it. So, any questions? Then please feel free to ask. Anything on operations? What has happened? Okay, thank you. Lightning talks will be happening in campus through this place in 5 to 10 minutes. Please don't forget to tweet. Please do put in your feedback and then we'll come to the side.