Thoroughly Modern: Clearing Up Some Cloud And IBM i Computing Myths
June 14, 2021 Timothy Prickett Morgan
We hear a lot of talk about cloud these days, but who is walking the walk? The coronavirus pandemic made a lot of companies take a hard look at how they were doing business, and hardware wasn’t excluded from this conversation. Businesses are now saying, with so many cloud options available, where and how can they start getting their feet wet?
We certainly all know that the IBM i community is dramatically different from most other platforms in the IT space, and has been since Day One. For household names such as Amazon Web Services, Microsoft Azure, and Google Cloud, third party applications make up the majority of back office software hosted on cloud infrastructure. In our community, however, businesses rely on highly customized apps that have flourished on IBM i and its predecessors and that has provided competitive advantage over the years – and continues to do so. Companies seem to be hesitant to move to the public cloud: Security, compliance, ownership, and cost are some of the top areas of concerns and big unknowns, and even some customers using the cloud are not really clear on how the value proposition really plays out for the public cloud.
I recently caught up with Fresche’s Nick Hampson to learn what he’s seeing in terms of cloud adoption on IBM i, the benefits of moving to the cloud, and common objections that he’s encountered.
Timothy Prickett Morgan: Welcome back, Nick. Not everyone knows who you are, so tell us a little about yourself before we dive in.
Nick Hampson: I am the product manager for digital transformation at Fresche Solutions. I’m responsible for our solutions that help clients transform high-value, monolithic applications into other languages and, more importantly, leverage the value that they have in them.
TPM: What is it that people get wrong when they think about IBM i and the public cloud?
Nick Hampson: Everybody wants a simple answer to a complicated question. In its most simple sense, cloud can mean that someone is managing your hardware and operating system for you in a data center. But cloud isn’t a distinct thing that forces people into a specific route. It’s a broad continuum that has options for most of today’s IBM i clients. I think that’s the bit that often seems scary – a client might look at one option that doesn’t suit their needs, but it’s just one of many approaches you can take, depending on what you need.
TPM: Can you give us some examples where you have seen IBM i customers leveraging the public cloud successfully?
Nick Hampson: Steve Will, the chief architect of the IBM i platform, recently led an informative session about IBM i and the cloud for the Power Systems Virtual User Group. I highly recommend watching the replay for an overview of what’s available from IBM.
For the purposes of our discussion today, I’ll give you three examples. The first one would be using the private cloud that IBM offers, along with a combination of your IBM i as it is today (logical partitions with your RPG, your DDS, 5250 green screens). This combo approach allows you to have on-premise infrastructure alongside some cloud-based capabilities, where you can employ HA/DR, containerization and DevOps in an environment that’s entirely under your control.
You always have control with cloud, but I think some people are still rooted in a mentality that makes them feel as if they must have physical control. This combo approach is low-cost and low-risk because it allows you to take advantage of horizontal scalability and DevOps without having to look at whole-scale change in your business.
The next example would have you move the logical partition to a cloud-based provider, like IBM Cloud or a certified IBM i cloud provider like Skytap. This approach enables you to divest yourself of the hardware and just pay a monthly fee. The fees for small companies are quite reasonable. With this second option, you can rest assured that the business application is going to run safely, and it’s going to be kept up to date. You also don’t have to worry about power outages and Internet connectivity.
My third example, where you’re modernizing your application and delivering it in an enterprise environment. You might be transforming your monolithic RPG or Synon application into a true multi-tiered, containerized application that will horizontally scale.
At this point, we’re only seeing the big customers take that approach. They’re used to that infrastructure already running Red Hat’s OpenShift Kubernetes container platform and have the people in place for it. They’re also pragmatic about the cloud, and they understand that there’s no magic button that will give them a microservices architecture. They’re looking at that as the largest step forward they can take with automation, getting themselves into the right environment for it, and then they’ll deal with the parts of their systems that they want to microservice-enable afterwards. In our experience, that’s often 10 percent to 20 percent of an application.
TPM: How much real cloud adoption have you seen within the IBM i community?
Nick Hampson: What we’ve seen is low so far. Up until this time last year, it was a stated future direction for quite a few of our clients, but there weren’t direct incentives to get there. I would say that COVID-19 has accelerated that shift, but there are still quite a few clients who have not looked at it or are not doing it because they don’t think they can (often for security, compliance or cost reasons). Unfortunately, this is often due to a lack of information.
TPM: What kinds of objections are you seeing from companies in terms of security, compliance, data sovereignty, and cost?
Nick Hampson: I’ll start by saying that there are no real reasons not to move to the cloud in some capacity. A lot of clients will say that they can’t put their systems in the cloud because of security and regulatory compliance. They’re often the same clients who have their customer data in cloud applications like Salesforce.
It’s important to also remember that you don’t have to expose your system to the public. There are now multiple secure ways of accessing your own infrastructure, so moving into the cloud doesn’t mean you have to have a publicly accessible login to your system.
Cloud infrastructure is built to be incredibly secure. Providers can’t afford to have issues with security! I’d be much more worried if I had an on-premise system, because I would have to remember to keep it up to date. With on-premise, you need to make time to install patches, or you need to dedicate someone’s time to keeping an eye on those things. This is not about getting rid of people, but rather putting their skills to better use. In terms of physical security, for a lot of companies, if you really wanted to go and steal their data, all you would need to do is throw a brick through a window and steal their IBM i. (If this happens it wasn’t me).
What I will say about cost is that cloud can be incredibly inexpensive for small clients with a few users. For the larger clients, it may end up being more expensive than just the hardware, but you need to do the apples-for-apples comparison. Who’s dealing with security? What’s in place to maintain resiliency and power and Internet connectivity? What does downtime cost you? Do you have bandwidth issues when you’re replicating offsite, whereas if it’s in the cloud using a different piece of bandwidth in between locations? What do outages cost you as a business if there’s a catastrophic failure?
TPM: I recall hearing some concerns from users around performance and scalability, but I would think that this would be an opportunity to improve those things? The noisy neighbor problem on the cloud cannot be as bad as still being on a Power6 or Power6+ machine, right?
Nick Hampson: The latest private cloud offerings from IBM do enable you to meter power. So you can switch that up, but that will never affect the fixed bandwidth that you’ve got coming into your system. That can be a bottleneck for people, unless they’re in countries with huge amounts of infrastructure in very small countries. It’s very different in the United States than it is in places like Japan. Obviously, the cloud does give you the capability to pick a system size and dynamically shift that up if you should.
As you move into the OpenShift containerized environment, you get the native capability to spawn more instances of that container to deal with load. That’s a much more dynamic system. But again, if you’re connected into other pieces of infrastructure, you’ll be in a high bandwidth, low latency environment as well.
If you’re using Microsoft Power BI, running IBM i on Skytap with Azure would give you a high bandwidth, low latency link. That’s just one particular example. There are all sorts of cases where being in a data center can give you access to multiple systems in that kind of high bandwidth, low latency environment. Many people don’t have good HA/DR setups, developer testing systems that can be fired up and shut down as you wish, the list goes on…
TPM: What do you think IBM i customers should really take a hard look at as it relates to public cloud?
Nick Hampson: DevOps is a huge advantage for companies that are doing active development. We see this internally with our own systems – it’s a game changer in terms of productivity, speed, reliability, resilience. For example, firing up development and test environments in a cloud environment takes seconds. You can have a template, spin up an environment while you use it and spin it back down again when you’re done. And you’re only paying for it for that short period of time.
TPM: As it relates to the IBM i platform, organizations are typically used to a comprehensive, integrated environment. We hear a lot about IBM’s Cloud Paks and how this business is booming for Big Blue in the wake of the Red Hat acquisition. Can you give us some insight into how Cloud Paks are the same and different from what IBM i shops are used to?
Nick Hampson: OpenShift is already the industry leader in making it easier to get to the cloud, and as a generalization, IBM’s Cloud Paks make some of the parts plug-and-play. For example, if you want to access the data, there’s a Cloud Pak for data. If you want multi-cloud management, where you have some loads in the cloud and some on premise, and you manage them through a single portal, it’s ready to go.
TPM: At Fresche, one of the things you do is help organizations build out their IT strategy on top of the IBM i roadmap. I would assume that the public cloud is part of these strategic conversations?
Nick Hampson: Absolutely. There’s tremendous value in the applications and systems that IBM i clients have built over the years. Clients often come to us with a technological problem and in most cases the answer is for business and IT to work together because they both fund and drive each other. We do see conflicting perspectives in our Discovery sessions where the business might be asking why they haven’t moved to the cloud, while the IT people have different thoughts on that.
Cloud is more of a future need for most of our clients. At this point, we’re primarily helping them build roadmaps and architectures to enable cloud in the future while maximizing the investment in their current systems. This might mean leveraging applications so they’ll run well on today’s IBM i infrastructure, but are portable enough to run in a containerized cloud-based environment when they want to do that.
At the end of the day, it’s our goal to help companies make the best decision for their needs by educating them in terms of what’s available and the pros and cons of each option. We can’t make the decision for them, but we try to ensure that when they do make that decision, they’re educated.
This content is sponsored by Fresche Solutions.
One last thing: Beginning tomorrow (June 15, 2021), Fresche is co-hosting a webinar series with IBM Canada about how organizations can go beyond keeping the lights on by realizing the full potential of the future while building on the value of their IBM i applications.
Each 60-minute session will focus on an area of opportunity, from roadmapping to quick project wins with web development, modernization preparation, cloud options and almost everything in between.
Attendees will also be invited to schedule a complementary two-hour IT strategy workshop with Fresche. You can register here.
Nick would love to hear what companies are doing with cloud and the services being used. Feel free to send him a message at firstname.lastname@example.org.
Nick Hampson is senior product manager for transformation at Fresche Solutions. Nick has extensive experience in design, UX, product management, innovation, and pre-sales. Focused on bringing the business value of design and UX to IBM i customers and the community at large, Nick is recognized as a UX expert on IBM i.