In The IBM i Trenches With: LightEdge Solutions
December 13, 2023 Timothy Prickett Morgan
The nature of the IBM i market is changing as the nature of the IBM i installed base is changing. With skills in short supply – and getting shorter by the day – and the work piling higher and higher as companies try to tackle application modernization and integration, different ways of acquiring and consuming compute and storage, and modern data analytics and artificial intelligence, we are entering an age where managed service and cloud suppliers will usurp the position formerly held by the business partners who worked downstream from IBM peddling and supporting on-premises Power Systems machinery.
Because variety is the spice of life, there will be myriad managed services providers with many different IBM i specialties as well as adjacent ones in the Windows Server and Linux platform worlds, which are also valuable to IBM i shops, most of whom do not run monolithic corporate computing environments. LightEdge Solutions, based in the datacenter hotbed of Altoona, Iowa, is expanding from its midwestern roots to cover the United States and is looking to be your cloud and MSP provider as you sort through your own infrastructure challenges.
We sat down with Sean Stenger, chief revenue officer for LightEdge, to talk about the company’s history and its expansion as it looks to make life a whole lot easier for IBM i shops under pressure to do so many things at the same time.
Timothy Prickett Morgan: The suburbs of Des Moines are where Meta Platforms and Microsoft run giant facilities, Google has one to the west of the Iowa state capital across the Missouri River from Omaha, Nebraska. But LightEdge started out a long time ago providing hosting and cloud services. Give us a little history of the company for some context.
Sean Stenger: As it stands today we are a national leader in highly secure and full stack private cloud and co-location managed services. If you go back in the history, LightEdge started as a business ISP based in Des Moines, and spent the better part of a decade focusing on the network connectivity to businesses. So the thesis of managed services or hosting started from that network core outwards to what services can be delivered over the network to the business. About 15 years ago, the kernel of our hosting and cloud strategy began. LightEdge started with purpose-built datacenters, offering co-location in Iowa that branched out to Kansas City and then into Omaha. And about seven years ago LightEdge was introduced to the market opportunity for hosting IBM midrange equipment.
One of our core principles is to deliver consistent hosting services in every datacenter we were in, regardless of the geography. And so when we launched our IBM Power cloud, we launched it right alongside a traditional private cloud for VMware-based workloads. Since that time, there has been a pretty good amount of growth – both organic and inorganic. We have grown through acquisition into Austin and into Raleigh and then into San Diego and into Phoenix. And there’s continued geographic expansion and the plan could be to reach out into markets like Minneapolis. We are focusing on the tier two and tier three geographical markets, and we want to provide highly secure physical datacenters as well as highly secure and compliant private cloud hosting platforms in each of those datacenters. We have always focused on uptime and mission critical compliance and security.
TPM: As an ISP, you could have started out just doing the network, or you could do the network or the hosting or both. Which did LightEdge do?
Sean Stenger: We did both hosting and network. LightEdge created the first Multiprotocol Label Switching, or MPLS, metro networks in Iowa, creating geographic ring metro connections for enterprise businesses in Iowa. We got the business capitalized and then we started offering suites of managed services over the network, starting with a hosted voice offering. From then on, the investment thesis for LightEdge was to find other services to deliver over the network connection that they already have with the businesses. Early on, our first cloud platform was IBM’s BladeCenter. We have just been iterating and offering from there, but the one thing that has stayed consistent throughout all of that time is an emphasis on security and compliance and delivering those services to highly regulated organizations, like banks and credit unions, healthcare companies, and so forth.
TPM: Do you have MPLS rings set up in each city, and do you connect the rings?
Sean Stenger: We do. And actually, we have rings that connect the datacenters, and we make sure that there are redundant paths between the carrier hotels or your main fiber backbones. The paths obviously go through there, but we also build paths around those as well. And then, because of that networking background that the company has, we can have a customer in our Austin datacenter and if the city of Austin loses Internet connectivity because of a major issue in Dallas, let’s say, we can actually give our customers Internet in that datacenter by going through Raleigh and back through our network. So the network is a foundational piece, and it’s part of that identity of that always-on uptime.
TPM: That sounds like a differentiating feature, and I now know I have to start asking other MSPs about their networks. . . .
How much of the business is driven by hosting and how much by cloud, and I realize that people sometimes use those terms interchangeably even though they are a little different – hosting is you rent a machine monthly and it is relatively static in configuration and cloud is you rent a machine hourly and can quickly turn capacity on and off. Anyway, how much of the business is cloud and hosting and how much is driven by the IBM platforms?
Sean Stenger: We think of hosting and cloud exactly how you are describing, and we are primarily hosting. So for us, it is about consistent workloads, consistent volumes, and consistent bills. Very few of our customers need scalability and elasticity and the platform stuff that your traditional hyperscale cloud would have.
Today, about 60 percent of our revenue is hosting or private cloud. The other 40 percent is co-location. We love selling managed services into the hosting and cloud side, and network is pretty evenly split between co-location and cloud. For the IBM business, it’s a healthy amount, not a huge percentage, probably 10 percent or so, and we have close to 100 customers at this point. Our IBM Power platform was launched only seven years ago, we’ve seen the adoption increase really meaningfully over the last four years. It took a couple of years for us to get a foothold.
TPM: You and every other cloud provider and MSP in the IBM Power market, as far as I can tell. It took everybody that same time. It took IBM time to get the cloud pricing on hardware and software, the license management and keys and virtual serial numbers to allow software to move around between on premises and cloud is only recently launched by Big Blue. So this is actually the right time to do a big push for hosting/cloud and managed services for IBM i customers. I think everybody’s facing that same scenario. What do you think?
Sean Stenger: I think it’s a pretty accurate depiction of what we see.
If you really get down to what the what the driving forces are, first, there are some legitimate applications that are great candidates for modernization, and modernization that can be done in a cloud native back end because there’s a feature or there’s a use case for it. But that represents maybe 35 percent of your traditional enterprise environment. So if I want to rebuild my application, make it more modern, so I can do things I can’t do today that give me a competitive advantage. And I’m using Amazon Web Services or Microsoft Azure and some of their cloud native proprietary features.
I think the reality is that the rest of the environment, the other 65 percent, doesn’t justify or warrant the premium pricing of those big clouds. And I do think that the traditional enterprise is super-concerned about security, and they don’t want to lose control of certain aspects of that security. And they want to avoid vendor lock-in for things that they don’t absolutely have to put on this cloud data platform.
So we think anywhere from 10 percent to 30 percent a traditional environment makes sense in the hyperscale cloud. And then the other 70 plus percent of it will stay on premises or be hosted or co-located and managed by someone who companies can trust to run it because it’s not a strategic part of my business to actually run it put eyeballs on it.
And contrary to popular opinion, those decisions are made by pretty rigorous cost analysis. And so, at the end of the day, if it’s cheaper for them to buy hardware, co-locate it, and manage it themselves versus the benefit of having it hosted they would do it. But that is not often the case.
And that is why our typical large customer has something in AWS and Azure, they have a pretty good SaaS software deployment, too, and they have a good chunk of applications hosted on our private cloud. And a number of our cloud customers also have co-location, and that might just be that they are co-locating their network and their security footprint. Or it could be that they are co-locating the pieces that are financially more feasible.
We did a recent market study on this to find out what drives things beyond cost, and it is this idea of control and security. There is a fear of sending services to a hosting provider that could get hit by ransomware, etc. Security is the primary driver after cost.
The IBM specific drivers are very interesting, because everyone always has a modernization strategy for these IBM applications. They always have a timeline of when they’re going to have it done and almost never do they get it done on time. And there’s always the decision to revert to what they had, which runs great, and they decide not to redevelop it. The modern IBM Power machines have gotten so dense on the computing power that, for instance, traditional manufacturing enterprises realized they did not need all of this horsepower. Not only that, all of that CPW is driving up their licensing costs. So we’ve seen people entertain hosting as a cost savings and they are not going to redevelop the application, and maybe they want help getting it managed because the IBM i team in-house is retiring.
TPM: We are living in a world where IBM i customers need maybe 3,000 CPWs or maybe even 10,000 CPWs in a system, but there is 25,000 CPWs in a single Power10 core and there might be 45,000 CPWs in a single Power11 core.
Here’s a pretty important question. There are lots of customers, according to my model, stuck on IBM i 7.2 and IBM i 7.1, and even some that go as far back as IBM i 6.1 and OS/400 V5R4. What can you do to help these customers?
Sean Stenger: That’s a great question. One of the things that we’ve done from the beginning of our IBM i hosting platform going back seven years was having it be capable of hosting outdated operating system versions. We walked into this business thinking that one of the reasons that hosting platform exists is for people to get a bridge solution for technical debt.
Early on, some of our best customers came from IBM Global Services, because the machinery and operating system a customer needed was so old they wouldn’t touch it. What we do is we provide the flexibility to bring those in the managed services to help facilitate some of the support pieces that are needed – those third party situations for extended maintenance and support, for instance. And then we partner with application service providers and managed service providers in the space to be able to help us move through that. We don’t do anything in the application layer, but we can stand up a new LPAR, load a new operating system version, and bring a partner in to install the database and see if there’s application compatibility challenges. If they are, they can plot a course to work through them. So we provide a staging space. No matter what, we’ll figure it out, and we’ll bring the right partners in to do that.
TPM: How old can you go? I mean, if I say I need an OS V5R3 partition, can you make it happen?
Sean Stenger: OS/400 V5R4 is the oldest thing we’ve touched. I know we have a couple customers that we will do annual tape validations for V5R4 right now. But I don’t think we have any V5R4 in production now. I would say the majority of them are IBM i 7.1, even though there are some older ones that are in there.
What we see is that multicloud is here to stay for a while, and we do know that security is a significant driver today. And we know that folks are just looking for help from their hosting providers. So we are bullish about the prospects for LightEdge, and we’ve got a lot of investment going into the business, fueling both organic and inorganic growth. And the IBM Power platform is a central part of that strategy. It is our fastest growing cloud or hosting platform, from a new business standpoint, last year and we’ve maintained pretty good momentum this year on that platform.
RELATED STORIES
Join Us For The IBM i On The Public Cloud Webinar
NGS And LightEdge Provide New Entry To The Cloud
