Managed service providers (MSPs) have a complicated relationship with cloud service providers (CSPs). On the one hand, the massive number of workloads running on public clouds clearly helps drive demand for managed services. At the same time, however, CSPs have made it clear that they intend to deliver those managed services themselves whenever possible. MSPs typically do best when a customer is looking for help managing application workloads that are distributed across multiple clouds and on-premises IT environments that are typically beyond the effective reach of any single CSP.
Against that backdrop, the cost of deploying workloads in on-premises IT environments is declining. In fact, a new report crafted by 451 Research on behalf of Canonical, provider of the Ubuntu distribution of Linux that also engages in the delivery of a managed services dubbed Bootstack, claims the total cost of its managed service running in an on-premises IT environment has now achieved parity with public cloud service providers.
New report from @451Research shows that the total cost of running managed services in an on-premises IT environment has now achieved parity with public cloud service providers
At the core of that argument is a distribution of Kubernetes cluster running on top of the OpenStack cloud management framework that is deployed on top of Ubuntu. In general, Canonical claims that stack of software alone delivers 80 percent of the functionality of the OpenShift Platform-as-a-Service (PaaS) environment running on Red Hat Enterprise Linux (RHEL) at one-third of the cost and is half the cost of a comparable stack of software running on a commercial VMware software. Most of that cost savings is attributable to both reliance on open source software and the automation that Canonical has built into its distribution of OpenStack.
MSPs, of course, can replicate their own managed service using a combination of Kubernetes, OpenStack, and whatever distribution of Linux they like to achieve similar results either on a public cloud, in a third-party hosting facility or in a data center owned by their customer. The latter two options have become more attractive in the U.S. recently because changes to the tax laws now make it possible to write down the entire cost of capital equipment the moment it’s purchased.
What this means for MSPs
This means savvy MSPs can now present customers with an approach to deploying workloads that doesn’t require the end user to lock themselves into a specific platform. Most end users are now informed as to when a given application workload should be deployed across the span of its life cycle. Applications that were initially deployed on a public cloud can become much more expensive to run a year later as the number of virtual machines and the amount of data being accessed starts to increase. In fact, helping customers move workloads off public clouds has become yet another opportunity for MSPs.
Being able to provide that level of flexibility requires significant levels of investment in automation on the part of the MSP. The primary reason CSPs can offer managed services at a comparatively low cost is because the investments they’ve made in automation. Much of that automation technology, however, is also finding its way into multiple open source projects, so the gap between what a CSP can achieve compared to a third-party MSP will narrow.
In the meantime, MSPs should present themselves as a platform neutral advisor whenever possible. Public clouds are essentially becoming a logical extension of an extended enterprise. Becoming too attached to one platform versus another only serves to narrow the total addressable market at a time when most organizations are relying on more types and classes of platforms than ever.
Photo: ra2studio / Shutterstock.