Session Information:

Cloud Strategy for the Next Decade

Session Transcript:

I have a global expert with us. Our next guest is ram via hace and Ram is coming to us directly from Bangalore in India. So, Ram, go ahead and toggle your camera on and please join us.

Great to have you here. So, ram has over 28 years of experience in IT services offerings, including, offering management, solution architecture, enterprise architecture, hybrid cloud solutions, cloud transformation and adoption programs and orchestration. Pretty sales, IT. Consultant transitions, and delivery. Ram is presently working for entity as the National Head for Multi Cloud Solutions. He has worked with large Fortune 500 companies around the world and consulted on cloud on the cloud adoption journey, service improvement programs, and IT transformation projects. Ram. It's a real pleasure to have you directly from Bangalore, India, to the world, sharing your expertise, very thankful to have you with us.

Thank you, just say, Hello, everyone.

Good morning.

Oh.

I'm going to talk about the cloud strategy for digital transformation And all of us, pretty much across all the geographies are older factors go into the term and the Kuwait situation with thus far for quite some time, 20 20 and beyond during these tough times, the cloud strategy or the true transformation, major focus areas. So, I want to spend some time on OPM up on this journey to cloud.

I'm basically going to talk about one-on-one level for different type of cloud that we are saying. What are the key drivers for our cloud adoption.

one of the workloads that are running on cloud are suited for cloud.

The architecture decisions are best practices that we see in this area. And since everybody's talking about the journey to cloud, we start with the migration of workloads as well. So, I wouldn't touch the migration aspects for Jacob Blob.

And some emerging trends that we see, a lot of traction in cloud.

How to manage cloud, when your cloud gets complex. And so how do you, because you have you will have private cloud, public cloud, multi-cloud. So how do you manage your workloads, you applications? when it's, like, beyond control sometimes. So that's one problem.

So I'm just going to start with some statistics, and I'm hoping most of your nano lifting Tesla. But let's talk.

So, Gartner said more than a year back, that nowadays, the surveys that were done involving more than 2000 companies across the world, it's 75% of the companies have to allow a strategy.

Which means like, when you're launching any application menu or launching any marketing campaign, when you are launching any offering, new software packages, et cetera, so ideally, the companies are looking at cloud.

Which is a platform for this new application of new software package.

The second point I really like, last week, you feel follow the economic times on Wall Street Journal, et cetera. So IBM, CEO or wind, you mentioned that glow is $1000 billion market capacity.

So it's very huge.

one a Gartner or this year, we are talking about more than $250 billion in cloud revenues from different companies and different services.

So lot of companies obviously have made an investment in cloud and there are me, which are coming into glob, the one which are already there that expanding and now getting into the new avenues and new services going up the value chain.

So, when we talk about cloud in one, but there are different flavors. There are different variants. I just wanted to be on the same page, so it has different connotation.

So, if you are having, uh, know, your applications in either your data center or a third party, or located, etcetera, that is also a cloud.

However, like, you know, you're not that everything there is a closed subset, but a lot of them, over a period of time. Customers have standardized system, since you're not a lot of heterogeneous systems, et cetera, to help standardize on cellular, their candidate, Combustion will virtually, at the kind of done the whole data center, yesterday's software defined data center. So with standardization with localization, you can get the benefits of Glow. And you can offer the services, which are posted on top of this way out Kaggle.

Then, once you realize some of the benefits of cloud, but most of those things still work on CapEx model capital extending expenditure.

So, clients may want to move workloads onto public.

Why? Because you can align to the systems, you can launch your application, no matter of minutes, minutes. You need connectivity there. And you are correct a cycle in the current data center is taking that.

If you're a developer, won't likely you're not in good terms with your IT admin team because they are taking time. And that are always the problem the China all in-between sometimes.

So that is the use case for hybrid cloud, where you will start extending your private cloud into, uh, public love.

Then there are integration use cases in this space, which are monitoring. Once you are on public dollars, what marketing tools you want to use? Can you leverage your current IT service management systems? Do you want to create new networks in the public cloud?

Or do you want to stretch your existing network policies, or v-lans, et cetera, onto public cloud? So those options are also available. Then we see the use case for cloud. So this is where, like, now, for example, you have somewhat groups on Earth now, Think some of these can be all the new workload from data center can be best suited for a job.

For them, it might be something like, let us say, Active Directory synchronization, Federation, better who might be better than it's just as an example. So you start creeping back having workloads in different clouds, AWS, or Google, or Alibaba and IBM Cloud, et cetera. So it's quite common that we see customers having more than by clubs, along with their on prem, private cloud. So that's like multi-cloud. If so nowadays, there is this concept of, build your application once and run. any risks with Docker Kubernetes picked up.

A tremendous traction feminist. So then, in that case, you can actually run this workflow whenever actual.

Depending on some of the parameters, like cost scalability, provisioning time, or management, eads, so you can run your application and realize benefits.

By looking at this, obviously, there's no one size fits all the requirement that no one language can everything as a day than teacher solution.

So, obviously, they didn't know one solution, which fix the client. It is all based on the requirements.

It's all based on the cloud strategy cloud roadmap that is agreed by the service provider and service Duma.

I have a chart that will talk about it.

So, these are the cloud adoption DIBELS, why you want to go to Cloud, so, first one is the agility and scalability.

So, for example, if I need a VM, I will get in a minute's in any of the.

If I want to increase my workload during port range, I can do that by setting up a policy to play.

If I want to reduce during the weekend, wear them on and off my employer working.

In general, let us say phi ref server I need one reps on the weekend, I can do those capability just managed by defining the scaling policy. So, cloud gives me that ability.

Next one, if DT will learn about pet snakes. So, you can actually spin up instances when you have a need.

For example, if you have a bad job that runs in the night, you can actually run your workload for a couple of hours in the night and then close that VM or close that task and dominate all the resources. So, that's the beauty of cloud.

So, you're completely on OpEx.

I'll do another example.

Like, if you have to run a data warehousing squaring, you are Data center.

It means like somebody had to come to the design of that data warehousing solution in your data And get your requirements. Design some dashboards, then you are ready to run your queries and get some benefits. So, in a cloud, you straightaway go with your SQL queries, rambo Square is, and you just pay for the amount of time that it takes.

It takes, let's say, NaN you pay for NaN, and that also meant that we run out, spend one million dollars to do all this.

Yes, the Healthy right Software design, software development life cycle, to have that data warehousing solution, then. now.

Since a data center, ultimately the Cloud.

So there are aspects which you can do via data refresh, Aldo moro Madison, by using Allow capabilities which are offered by player like they are HP, et cetera.

So they help us pick workloads for big data, naturalization platform or VDI, virtual desktop, or, if you're running the city, so that's also possible.

Nowadays, organizations are going cloud native, which means like, you are basically converting your legacy applications into microservices. So yeah, pick up where you can scale up, scale down. These decoupled micro services according to the requirements, and you can have faster needs. So, with your legacy monolithic application layer it is to take let us say two years or one release. nowadays.

We see multiple companies doing resistivity, 1 or two weeks, It's quite common.

Another example of adopting cloud is a global distributed workforce.

Like, if you have companies in different parts of the geography, having the cloud footprint, for example, or DNS load balancer center worse than the CDN solution.

For example, these are some of them. So, it's quite expensive.

So, cloud can give you, all of those benefits, wherein you can connect you to the nearest location, nearest region near the celebrities on where you are. And you can post what workloads in. Those locations can do it if it keeps on.

For that, you have implement, in terms of the network connectivity.

Is the latency also stone.

The last one is provisioning cycle high productivity. So ... software development life cycle has come down now, with the Agile methodology.

You really don't have one year cycles or anything product. So which means in touch, it also has to take care of all the business requirements.

And in fact, check should be provision, along with the ... off the streets. So sometimes you provision and then be provision, you have a test enrollment, you go in. There are good as some of them day. After the work is done of the testing is done, that environment needs to be decommissioned. So cloud helps with those requirements. And nowadays, again, the last point here in about the API economy.

So, let's say you're a retail company, You want to connect your systems to credit card company, credit card companies, in turn Mournful. Connect with Social Security or credit for companies.

So, all that thing is happening via API, and you basically open up your API services and put Request API requests, postal, get with the third party, trust it, and then get responses and go with your business, so that I'm not businesses which are API.

Then, like A pre 20, 20, and in 20 20, we are in this situation.

So, what's happening now with the advent of fluid 90? So, a lot of companies, we see that they are changing the financial motor on CapEx to OpEx.

So flower.

It's run on a one-off, extended capability, which is that is the major drag, lower class, 10 other strategies.

Then we're also seeing the workforce working from home, Suddenly, like we started seeing millions and millions of employees on thousands of companies, started working from home, it has to be enabled, so really lower the best platform to do that.

Then collaboration productivity tools, besides connecting the system. That's not only use kids, you want to have a conference call. You want Whiteboard Obsession with your colleagues you want Chalk.

You have different things for your great, productive hitting file, et cetera. So things like Microsoft Teams are Cisco, WebEx, so they're very relevant in these times.

Right.

So this is where in fracture at the core, DevOps, aggravated approval, automation, is a big, big team.

And by default, everything is permitted in Glow. When you define any deployment, it will tell you, do you want to download this, defund planned? We want to download this template.

Which is a core, whatever you are doing the portal or CLI is basically that candidate pool. So that is very important nowadays, and you can make changes.

By downloading that end, deploy it, and stay on, you know, the core can be used.

We're also seeing industry specific solutions here. So just you want an example.

So, for example, in the retail industry, so what they're doing is, or any ... or any buyer who wants to go into the retail shop.

They will have an ID, can be there in the phone itself, and then one needs to download that app and then log into that shop. Then one gets appointment.

Go to that shop.

Then if there is a Q it will tell, you go to a 10 minute. Once that person is let inside the shop, then he or she will do the shopping, then do the online payment and analytic solutions.

You can get details on who made what purchase, what is getting more traction, which product, how to position this button like this product, is, put them online shopping. And both offline shopping, there are solutions that are emerging.

So, this can be developed in blood.

Workload consideration. So when you have any applications, obviously are developed, along with the IT admin, needs to assess the requirements, whether it is to raise the compute me, when requested bare metal. If it is like a GPU type of a workload cellular by itself.

Then there are security requirements on encryption.

The keys, which are the secret certificates.

And then there are IAM policies, identity access management, then they were based access Temple to walk in the Cloud Wire.

Then they said, Oh, Federated login, for example.

If I have a web app or mobile app, I would let my user login via his or her Facebook ID.

Google I slash O, that's very common nowadays.

So I can facilitate federated third party authentication, so clower will help with any of those aspects.

Then audits like some companies, they are very keen to get inspection ISO 27,000 talks, etcetera. So if it is facility infection, etcetera, how cloud is not really a good candidate, because they won't let you does it devastated and by default everything is shared that the Internet chess.

The audit compliantly, the Audit and Compliance Needs need to be very limited funding, then our workloads, where you can have them, whatever. For example, you feel somewhat blogs on prem, and you can extend them onto Cloud. Once you think, cannot put data, you can then get back to on prem.

You can also leverage your own licenses, the green on prem, and public Cloud.

All those options are available.

They did this concept of pets and Cadence.

So basically, what it says in an actually like, talking about the new table, So instead of the problem with cetera on a workload, it's easier in cloud.

to kill that workload, terminate that workload and bring it back and put your application there in a matter of minutes. And get going.

Rather than spending to our DLLs debugging, what's happening on that.

So that's the catch, The novelty Pepsi is basically you have let us say an application like I said the H RMS employee salary type of obligation, which is critical, tend to do that. You can treat it like a pet, and then run it on prem.

You can manage it.

You can also make sure that no data is left outside data center. We can say that, I can do facilities fiction. I can comply with all the requirements because I manage it and monitor it.

So there are two, it is this area.

The next one is the security policies. Again, like a lot of companies, like especially banks, angel answers, like insurance companies, they have in-house applications in how security policy, how they should be managed, etcetera.

And before you talk about Lean that go into Cloud, those policies also need to be addressed inflow. most of the time.

Colored, you have coupled applications. You have decoupled application, legacy applications, cloud native applications.

That's how it looks when you how no basket of applications here.

So obviously, decoupled applications, cloud native applications. The advocates, OK, talk loud and legacy applications with App modernization, with containerization.

They can be easily ported onto Cloud. Of course, there is some three factor or engineering effort we did.

So always all these workloads will have shared services.

So when you move something to cloud, make sure to share so with this requirement is known to you, for example, Electro Directory, DNS, DHCP entity. So it's not that you have to take the whole thing glyph Cloud but then how we work on cloud.

Point two will shared services if they're posted on prem unless you are moving them to love which is possible.

How do you start the journey to cloud.

So, you basically start with the advisory services.

So, led to how all the skills in your company, a lot of companies in, on system, integrators IT service he will come and help you with advisory services, where they come, take a look at your workload to applications, holiday tupper, how complex they are, David Eating, etcetera. So, there is this sheet sobs, tipple right mentality, where application is evaluated against.

So the chips are basically, retain, retire, lift and shift, the engineer re platform, etcetera. So those need to be evaluated. Then, once that roadmap is available, which clients are three along with the service provider, and my patient, I was struck, by then, you have your target, You have a landing zone below. Then you can, again, migration those CPL. 10 workloads. Roadmap is different. If you have one thousand workloads. The Roadmap will be complexity. Different approach to migrate out will be different.

So that's how migration happens. And if it is tolerant of let us say VMs and hundreds of obligation, they are normally done via a wave plan.

In advisory, they distort the workload. Then you go to different ways. For example, the first way you migrate applications, which do not have any dependency, which are like low hanging fruit, which are totally copyright. So you will migrate them. Next, we'll look at, say, 2, 3. What laws, which are there in the same data center, they're talking to, talking to the same subnet. That you can migrate together.

And then the third, to address the complex workloads, where, let's say, you have database in some other location, web in some other location, business logic, in some other location. So they need to be treated differently and complex.

Some of them are hard coded. So they come in the last wave.

So in some cases, like, you also need to build services on Cloud. For example, set up networks, set up security policies, set up, VMs, load balancers, firewalls, and the application.

Battles, NAD. There are only two components that need to be big.

Netflow so, Before the conditions have migrated, so that's needed, and the last one is now. We have assume that your event, the migration period, is also ready. Applications are running live, so somebody needed to, it cannot lose applications in the cloud.

So that's managed services and resources aspect, so this is the kind of journey to cloud thing, and this one, which are not shown here, Optimization, or ..., also known as CSI Continuous improvement.

So, continuous implemented.

You look at the security posture, look at the cost, you look at the availability, look at the security aspect, So there's always some room for improvement.

Let's give me, like, uh, now, it's always the continual transformation, that's what this is.

I put to Philadelphia.

one is, like, What business are, we standing must think when it comes to club, and what services provider has to do, so, he's brought up yet, just considering that it can be a system integrator. It can be IT service provider.

Who is helping the client?

So, what client is thinking? So, in case of client, like, they are talking about the business transformation roadmap. So, what is good for my business, where actually invest my money?

So, that's the Roadmap discussion, which happens at CXO level.

Your CIO here for a second, Then, there are different lines of businesses can present their own business clarity, before a roadmap, in this definition.

So, each of the lines of businesses will have their own products, that campaigns, etcetera. So, there'll be product specific roadmap, that David have, in case of a, finance, et cetera, so, each of them will have them. Then they will have a high level mapping out what capabilities are needed for my but, they cannot do the IG session here or they will also look at their competition in this industry for the finer American. Something they want, the best part was something.

So, uh, that evaluation is needed, then kills they begin our bill. They have 10000 people to do all these things that business expects. In most of the cases it is the law and in some capacity this depending on the size of organization the maturity of that organization.

So now on the right side if we should focus this becomes the services specific to the integrator or IT service provider. They want to understand what are the requirements of display.

They send the requirement they can actually provide the ID design of IdeaScale lambskin.

Then it also means they will say that, you know, I can provide you the skills I can also get skills with another et cetera and family business that this service provider can do it.

In some cases, they didn't like multiple service providers who will do it end to end, sort of the big thing.

Then, most probably IT, the service can do more when we have multiple service providers.

These IT services also play a state wide communicate, because everyone wants to make a change.

Those kids that are available, since they work, with different companies, the same domain that it is banking, retail, insurance, manufacturing. Right? So, they can actually provide the industry best practices to this Duma.

Which is a real benefit.

And before they have done the IT journey together with a Tetra, then they can put together and create the digital assets which are which is the highest level of collaboration.

I want to talk about cloud on Cloud optimal workloads. So, first one is artificial intelligence, machine learning.

So beta much cloud native workloads. And every week, or every month, you will see new services get launch, whether it is GCP or AWS. So del Beta, solid platform in this. And if you want to do the same thing with the on prem or private club, I will not say that it is possible.

But I would say, it takes a lot of investment. Google does, and keeping up with the innovation that's happening elsewhere.

He's the most.

it's not, you cannot catch up with what's happening.

So either you ought to be too big, too rich. To do this, the best thing gets warm glow.

Second one is the big data, or data analytics cloud setting. We know how complex it takes to set up a Hadoop cluster on prem. So, you can always edit on Cloud, and getting going in a matter of minutes.

Data Warehousing, I give example of how we can run equity. And he'll pay for the time that you run the query, whether it's on GCP, with things like big query, or is there data warehousing solution, the solution?

DevOps, though, everybody is talking, every application developer is talking about DevOps, null, and slow use need to offer the envelopes functionality. So one is you can set up your DevOps platform in GCP, et cetera.

Or if you have samples, if you, if your next a developer, exposed to things like Bitbucket, Jenkins, or Bamboo Audra GitHub, you can actually do your DevOps on Cloud when integrating with those third party tools. So you can integrate Marketplace.

So finally, containers. So this is another workload. So nowadays, we see everybody modernizing their applications onto container sand.

All the cloud providers will be help you build applications on containers. And it's like matter of seconds, where you can either launch containers or move your self, manage those container, by owning the Kubernetes cluster that we want, the container cluster.

Do you want cloud provider to manage the cluster, or do you want to manage the cluster?

All the options that are available.

And with this, um, web apps, mobile apps, right, like no opposition figel, no SQL, databases, ..., So, for example, games, WordPress setup. So, no SQL solution available in all that long.

So, it's a very good platform. And, I don't know, how many have seen an article or video on YouTube, how the day in our launch, one million users, in a matter of a couple of months, or etcetera. So, we can go on to go by subject. So, if you The architecture of how you as a startup right, and can actually launch your application on cloud, and we can scale to billions of users.

Some of the pitfalls, lists.

I want to just put some point model is and, Hilary, that so, you need to evaluate the cost model, because on cloud, it's cost for a linear path for running container, the cost? For running a balance of the cost, for, let's say, managing a certificate.

Did it cost for data, which is going out of flow to your data center?

So, all those things, how do we evaluate it so that there are no surprises?

Then there's another thing that that for consideration is, the lock, and so, this is a topic, but just how some Saturday so that you don't lock into a cloud, and then it becomes a problem. We started valuable. It might be a problem. So be aware of it.

Then figuration of application, so when you want to migrate your applications, make sure that you have a roadmap, You have an approach. How do you migrate those applications?

Desktop application, then connect your really Google applications on Cloud.

It's always good to have multi-cloud model. So now, if you look at some of the future, whether it is VMs containers, plus the load balancers, they might look the same on all the cloud.

But if you double click, there could be some differentiated value in one cloud.

So take a look at that.

Governance for the cloud model is very important, but there is always name. So you want to make sure, when someone provisioning, you have the right workflows, approvals, et cetera, so that someone knows what is it possible to detect with that asset or resource. Then, it's also good to have tax mandate, make it mandatory for your developers.

You are writing, folks, so that, our tag to identify what is the ethic, which length of line of defense is building? Which group is using Mexico?

Then you can also how policies, So, for example, your development should not, Should not go for a high-end, federally funded department, which doesn't require Are you can say that I will not let any of my users run anything during the weekend after 10 PM in the night.

So, that's part of what you have to provide isolation by having different projects within them, within the organization, and make sure that one project doesn't interfere with another project, so that you avoid that isolation. You provide different answers for different projects. That's all possible.

This is my last job. So basically like what I have drawn here is ..., multi-cloud is the order of the day. So how do you manage this? So at the bottom, you will have different providers, if you have something in your data center, something, although you how I touch or in any update the gloves, just shown for IBM.

So are you on this internship, so you will actually provision your resources?

And there are workflows which will help you with provisioning, in terms of they can do that, requesting the resources and then the ...

engine to make sure that the resources provision approved and the end users are notified when the resource itself closure. Then, we also want to track which ethics?

Anyway, because, like, we will move to the liquidity of who is doing what we're doing with.

So, when you have these workloads crawling everywhere.

So that slash inventory, if they become very important, and you want to make sure that businesses are not swiping their credit.

They are bypassing the IT and I'm doing so with this on cloud.

Which one knows, right? Which no one knows, the Centrelink IT doesn't know and don't know what is the cost impact. So, the idea of shadow IT that it.

So, that you want to avoid now how artificial intelligence, which can detect some of the anomaly in your intellect and, know, make some recommendations. And some kids, some cases, ask the user to provide Mikey factor authentication from ocean security question, before allowing them to login.

Or, if one person, one ID trying to login geographies, for example, Latin America and south-east Asia, at the same time. So, that's an anomaly, right? So, those things need to be protected.

There, you are catering to different lines of businesses, so cost management, invoicing, or different lines of business.

They become important so it can be mended, MIT can add. So there are different solutions that are available.

So some of the native solution depicted. For example, since everybody's talking about DevOps, everybody's talking about microservices and containers.

So Google handles can actually provision and manage your container no matter where that can be in any of that input to that actual below and toss of the manager managing clusters wherever they are.

Vish, Tunnel wells with playing in this space for the solution which is to be known and pirated.

Microsoft is the rancher open.

This is an open-source solution, and they also have IOT cluster manager capabilities.

So, I've just gone one more level here.

So, you want to have your own automation, right, along to your workload, It can be on Private Cloud AWS, which you can do it via configuration management tools.

There are many notes as shown anticoagulant data, you can do that.

You also want to manage these resources and make sure if there is some problem with EC two VM no. Sorry. This one ticket incident indicate an IPC to take on LinkedIn to source. ServiceNow and remedy have just shown here, it can be.

So you also want to manage policies with tax paying the last leg. So then actually enforce policies when you deploy any resource anywhere, we want to have the same security policy, same provisioning policy.

Same hardening rules are applied? No microwave to solve, it all can be facilitated.

So, I'm basically talking about Orchestrators solution which are emerging in this area.

So that you have visibility And ensure compliance, you can push policies, then also do cost management, et cetera. So this is bulk, best practices in this space.

So with that, I'm pretty much done.

Rehm terrific presentation. Thank you for sharing that with us.

I'm gonna ask you to stop sharing your presentation, so that the audience can see our cameras, But you get very, very good presentation. Unfortunately, we are out of time for the live Q&A.

And what I'm going to ask you is, I'm going to ask the audience to go on to the Lincoln on LinkedIn and the impose your questions there for ram. And then we're going to ask him to go in there and provide some answers, some short answers to some of your questions, because we're out of time on this session and ready to start the next one at the top of the hour.

So there was a question, just so you know, ram from Ernest Villa ..., who is doing launching five G in Guam in the United States by NTT, docomo. And he asked if you have a relationship with, with them. And specifically, he asked about the business case for S C S E, which is Secure Access Service edge. Again, unfortunately, we don't have time to discuss that right now. But I'm gonna ask the audience, and maybe perhaps, ram, to, to stop by the LinkedIn for this conference, and then we can provide the answers there, REM. I know it's late in Bangalore. It will have really taken the time of your very busy day to share your expertise of our global audience. We're very appreciative of that, thank you so much for, for doing that for us today.

Sure, thank you.

My pleasure, thank you.

All right, ladies and gentlemen, this wraps up the second segment at the top of the hour. We are going to welcome Dina, who is a Google Cloud Specialist with Cloud peta, and do not want to miss her session. She's going to talk about Infrastructure modernization and how to choose the right modernization path for your specific environment. Brian Raffle, conference director is going to be introducing the next two sessions for for, for us, very capable hands and the M scene from Brian. And I look forward to continuing discussion tomorrow. And remember, you can go into the LinkedIn page and post your questions there. And we can have the speakers join us over there as well for additional Q&A. So Brian went back with you at the top of the hour. and Dina SOA, on, on modernization of infrastructure for IT, thank you very much.