The next BriefingsDirect Voice of the Customer use case discussion explores how a European gaming company adopted new cloud management and governance capabilities with developer productivity as the prime motivator.
We’ll now learn how Magellan Robotech puts an emphasis on cloud management and control as a means to best exploit hybrid cloud services to rapidly bring desired tools and app building resources to its developers.
Here to reveal the journey to responsible cloud adoption with impressive payoffs is Graham Banner, Head of IT Operations at Magellan Robotech in Liverpool, England, and Raj Mistry, Go-to-Market Lead for OneSphere at Hewlett Packard Enterprise (HPE), based in Manchester, England. The discussion is moderated by Dana Gardner, principal analyst at Interarbor Solutions.
Here are some excerpts:
Gardner: What are the drivers in your organization for attaining faster cloud adoption, and how do you keep that from spinning out of control, Graham?
Banner: That’s a great question. It’s been a challenge for us. One of the main problems we have as a business is the aggressive marketplace in Europe. It’s essential that we deliver services rapidly. Now some of our competitors might be able to deliver something in a month. We need to undercut them because the competition is so fierce.
Going from on premises into virtualization and on-premises cloud was our first step, but it wasn’t enough. We needed to do more.
Gardner: Speed is essential, but if you move too fast there can be risks. What are some of the risks that you try to avoid?
Banner: We want to avoid shadow IT. We’ve adopted capabilities before where the infrastructure team wasn’t able to provision services that supported our developers fast enough. We learned that the developers were then doing their own thing: There was no governance, there was no control over what they were doing, and that was a risk for us.
Gardner: Given that speed is essential, how do you bring quality control issues to bear when faced with hybrid cloud activities?
Banner: That’s been a challenge for us as well. There hasn’t traditionally been central payment management from across multiple cloud interfaces so that we could ensure that the correct policies are being applied to those services.
We needed a product that ensures that we deliver quality services to all of our customs across Europe.
Gardner: Raj, is this developer focus for cloud adoption maturity a main driver as HPE OneSphere is being evaluated?
Mistry: Yes, absolutely. The reason OneSphere is so good for the developers is we enable them to use the tools and frameworks they are accustomed to — but under the safety and governance of IT operations. They can deploy with speed and have safe, secure access to the resources they need when they need them.
Gardner: Developers probably want self-service more than anyone. Is it possible to give them self-service but using tools that keep them under a governance model?
Mistry: Some developers like the self-service element, and some developers might use APIs. That’s the beauty of HPE OneSphere, it addresses both of those requirements for the developers. So they can use native tools or the self-service capabilities.
Gardner: We also have to consider the IT operators. When we think about where those new applications end up — it could be on-premises or in some number of clouds, even multiple clouds.
Simplified Hybrid Cloud Management
Mistry: HPE OneSphere is very much application-centric, with the capability to manage the right workload, in the right cloud, at the right cost. Through the ability to understand what the workload is doing — and based on the data and insights we collect — we can then make informed decisions on what best to do next.
Gardner: Graham, you are on the operations’ side and you have to keep your developers happy. What is it about HPE OneSphere that’s been beneficial for both?
Banner: It provides great insights and reporting features into our state. When we deployed it, the feedback was almost instantaneous. We could see where our environments were, we could see the workloads, we could see the costs, and this is something that we did not have before. We didn’t have this visibility function.
And this was a very simple install procedure. Once it was up and running, everything rolled out smoothly in a matter of hours. We have never seen a product do this before.
Gardner: Has having this management and monitoring capability given you the confidence to adopt multicloud in ways that you may not have been willing to before?
Banner: Yes, absolutely. One of the challenges we faced before was we were traditionally on-premises for the entire state. The developers had wanted to use and leverage functions that were available only in public clouds.
One of the challenges we faced before was we were traditionally on-premises for the entire state. But the developers wanted to use and leverage functions only available in the public clouds.
But we have a small operations team. We were wary about spending too much training our staff across the multiple public cloud platforms. HPE OneSphere enabled us to onboard multiple clouds in a very smooth way. And people could use it with very little training. The user interface (UI) was fantastic to use, it was very intuitive. Line of business, stack managers, compliance and directors, they all could go on and run reports straight away. It ticked off all the boxes that we needed for it to do.
Gardner: Getting the trains to run on time is important, but the cost of the trip is also important. Have you been able to gain better control over your own destiny when it comes to the comparative costs across these different cloud providers?
Banner: One of the great features that OneSphere has is the capability to input values about how much your on-premise resources cost. Now, we have had OPEX and CAPEX models for our spend, but we didn’t have real-time feedback on what the different environments we are using cost across our shared infrastructures.
Getting this information back from HPE OneSphere was essential for us. We can now look at some products and say, “You know what? This is actually costing x amount of money. If we move it onto another platform, or to another service provider, we’d actually save costs.” These are the kind of insights that are generated now that we did not have before.
Gardner: I think that economics trumps technology, because ultimately, it’s the people paying the bills who have the final say. If economics trumps technology, are you demonstrating a return on investment (ROI) with HPE OneSphere?
Mistry: One of the aims for OneSphere is the “what-if” analysis. If I have a cloud workload, what are its characteristics, its requirements, and where should I best place it? What’s best for that actual thing? And then having the capability to determine which hyperscale cloud provider — or even the private cloud — has the correct set of features for that application. So that will come in the not too distant future.
Gardner: Tell us more about Magellan Robotech and why application quality, speed, and operational integrity are so important.
Banner: We operate across Europe. We offer virtual gaming, sports, terminals, and casino products, and we have integration to other providers, which is unique for a bookmaking company. A lot of gaming providers operate just retail platforms, or maybe have an online presence. We do everything.
Because we compete with so many others, it’s essential that our applications are stable, scalable, and have zero downtime. If we don’t meet these requirements, we’re not going to be able to compete and our customers are going to move elsewhere.
As a service provider we sell all of these products to other vendors. We have to make sure that our customers are pleasing their own customers. We want to make sure that our customers have these value-adds as well. And this is where HPE OneSphere comes into play for us.
Simplified Hybrid Cloud Management
Gardner: As the third largest gaming enterprise in Europe, you’re in multiple markets, but that means multiple jurisdictions, with multiple laws about privacy. Tell us about your security and compliance needs and how HPE OneSphere helps manage complexity across these different jurisdictions?
Banner: We deal with several regulatory bodies across Europe. Nearly all of them have different compliance standards that have to be applied to products. It’s unreasonable for us to expect the developers to know which standards have to be applied.
The current process is manual. We have to submit applications and spin-up machines on-premises. They have to be audited by a third-party, and by a government body from each country. This process can take months. It’s a long, arduous process for us to just release a product.
We needed a tool that provides us an overview of what is available out there, and what policies need to be applied to all of our services. We need to know how long it’s going to take to solve the problems before we can release services.
With HPE OneSphere, we are gaining great insights into what’s coming with regards to better managing compliance and policies. There will be governance panes, and the capability for line-of-business staff members to come in and assign policies to various different cloud providers.
And we can take this information to the developers and they can decide, “You know what? For us to go live in this particular country, we have to assign these various policies, and so we are going to need to change our code.” And this means that our time-to-market and time-to-value are going to be much higher.
Gardner: Raj, how important is this capability to go into different jurisdictions? I know there is another part of HPE called Cloud28+ and they are getting into different discrete markets and working with an ecosystem of providers. How much of a requirement is it to deal with multiple jurisdictions?
Guided compliance, vigilance
Mistry: It’s very complex. One of the evolving challenges that customers face as they adopt a hybrid or a multicloud strategy is how do I maintain my risk posture and compliance. So the intellectual property (IP) that we have built into OneSphere, which has been available from August 2018 onward, allows customers to look at the typical frameworks: FIPS, HIPAA, GDPR, FCA, etc.
They will be able to understand, not just from a process perspective, but from a coding perspective, what needs to occur. Guidelines are provided to the developers. Applications can be deployed based on those, and then we will continually monitor the application.
If there is a change in the framework that they need to comply with, the line-of-business teams and the IT operations teams will get a note from the system saying, “Something has happened here, and if you are okay, please continue.” Or, “There is a risk, you have been made aware of it and now you need to take some action to resolve it.” And that’s really key. I don’t think anybody else in the market can do that.
Gardner: Graham, it sounds like you are going to be moving to wider adoption for HPE OneSphere. Is it too soon to get a sense of some of the paybacks, some of the metrics of success?
Guidelines are provided to the developers. Applications can only be deployed based on those, and we will continuously monitor the applications in production.
Banner: Fortunately, during the proof of concept we managed to get some metrics back. We had set some guidelines, and some aims for us to achieve during this process. I can give you an example. Traditionally we had a very old-fashioned ticket system for developers and our other customers.
They turned in a ticket, and they could wait for up to five days for that service to become available, so the developer or the customer could begin using that particular service.
With HPE OneSphere, and the self-service function which we provided, we found out that the time was no longer measured in days, it was no longer hours — it was minutes. This enabled the developers to quickly spin up machines. They can do iterative testing and get their products live, functioning, and bug-free faster. It frees up operational time so that we can concentrate on upgrading our platform and focus on various other projects.
We have already seen massive value in this product. When we spoke to the line of business about this, they have been pleased. They have already seen the benefits.
Gardner: Raj, what gets the most traction in the market? What is it that people perk up to when it comes to what OneSphere can do?
The data-insight advantage
Mistry: It’s the cost analytics and governance element. Deployment is a thing of the past. But once you have deployed it, how do you know what’s going on? How do you know what to do next? That’s the challenge we are trying to resolve. And that’s what’s resonating well with customers. It’s about, “Let me give you insights. Let’s get you the data so you can do something about it and take action.” That’s the biggest thing about it.
Simplified Hybrid Cloud Management
Gardner: What is it about the combination of product and support services and methodologies that are also helping to bring this to market?
Mistry: It’s about the guidance on application transformation. As people go digital, writing the new cloud-native stuff is easy. But like with Graham’s organization, and many organizations we talk to, they have a cloud-hosted, cloud-aware application that they need to be able to transform to make it more digitally friendly.
From a services perspective, we can guide customers in terms of what they should do and how they should introduce microservices and more cloud-native ways of working. Beyond that, it’s helping with cultural stuff. So, the beginnings of Agile development, leading to DevOps in the not too distant future.
The other side of it is the capability to build minimum viable clouds, both in the private and the public clouds with the IP that we have. So, the cloud thing can be had, but our effort is really to make it very easy.
Gardner: That strikes me as a huge next chapter, the minimum viable cloud. Is that attractive to you at Magellan Robotech?
Banner: Absolutely, yes. From an on-premise perspective, we want to go forward into the public cloud. We know we can leverage its services. But one thing we are very wary of is the cost. Traditionally, it has been expensive. Things have changed. We want to make sure we are not provisioning services that aren’t being used. Having these metrics is going to allow us to make the right choices in the future.
Gardner: Let’s look into the crystal ball. Going to the future, Graham, as a consumer, what would you like to see in HPE OneSphere next?
Public core and private cloud together
Banner: We already have the single pane of glass with OneSphere, so we can look at all our different clouds at once. We don’t have to go in multiple consoles and spend time learning and training on how to get to these reports from three or four different providers. So, we have the core, the core is there. We know that the public cloud and private cloud have different functionalities.
On–premises can do certain things extremely well; it can handle all our current workloads. Public cloud can do this, too, and there are loads of additional features available. What we would like to see is a transition where some of these core functionalities of the public cloud are taken, managed, and applied to our private cloud as well.
There are compliance reasons why we can’t move all of our products into the public cloud. But by merging them together, you get a much more agnostic point of view of where are you going to best deploy your services and what features you should have.
Gardner: Ultimately, it may even be invisible to you as to whether it’s in a public or private cloud architecture. You want your requirements met, you want your compliance and security issues met, and let the automation of the underlying tool to take over.
Simplified Hybrid Cloud Management
Banner: Absolutely, yes. We would like to abstract away the location completely from our developers and our application guys. So, when they deploy, it gets put in the right place automatically, it has the right policies assigned to it. It’s in the right location. It can provide the services needed. It can scale. It can auto-bounce — all of this stuff. The end-user, our applications team, they won’t need to know which cloud it’s in. They just want to be able to use it and use the best available services.
Gardner: Raj, you just heard what the market is asking for. What do you see next for providers of cloud monitoring and management capabilities?
Mistry: Our focus will be around customizable cloud reporting, so the capability to report back on specific things from across all of the providers. Moving forward, we will have trending capabilities, the what-if forecasting capability from an analytics and insights perspective. Then we will build more on the compliance and governance. That’s where we are heading in the not-too-distant future. If our own developers do well, we will have that by the end of the year.
You may also be interested in:
- Ryder Cup provides extreme use case for managing the digital edge for 250K mobile golf fans
- How Norway’s Fatland beat back ransomware thanks to a rapid backup and recovery data protection stack
- How hybrid cloud deployments gain traction via Equinix datacenter adjacency coupled with the Cloud28+ ecosystem
- HPE and Citrix team up to make hybrid cloud-enabled workspaces simpler to deploy
- Citrix and HPE team to bring simplicity to the hybrid core-cloud-edge architecture
- New strategies emerge to stem the costly downside of complex cloud choices
- Huge waste in public cloud spend sets stage for next wave of total cloud governance decisions, says 451’s Fellows
- How HPE and Docker together accelerate and automate hybrid cloud adoption