Open source is free right?

Being part of the Cloud Engagement Hub team in IBM, I focus on collaborating with clients on their journey to cloud. The clients I have worked with to date, all have different journeys and they are all in different places with those journeys.

Since IBM closed on the Red Hat acquisition, I have heard from people that they just don't get why IBM acquired a company who primarily supports open source software that anyone can download for free. I applaud all those Do It Yourself (DIY) or Roll Your Own (RYO) technical people out there, and I often view myself as a DIY person as well for many things, but I do draw the line on various things. While I work on my car at times, I chose not to work on my brakes as I view needing some professional support for that.

There is no doubt that with the right training, the right number of people and the right collection of open source software, businesses can put together an environment to run applications on a cloud. The current view of many is that the answer for cloud based application landing zones is made up of Linux, Containers, and Kubernetes, all of which can be downloaded for free from the open source community websites.

A single Kubernetes cluster or two can be stood up fairly easily, and some applications can be created and placed on these clusters. With a little effort, your cloud application can quickly be up and running as an MVP in many cases. It starts to get a little more complicated as one determines they need to add operational features to the clusters like monitoring, logging, metrics aggregation and viewing, DevOps pipelines, image registries, and maybe some persistent storage. There are many great open source components available as free downloads which will address each of these areas, but how do I chose the right ones and then weave these various components into a solution? It can be done, but it takes people, time, and training get to an application platform evolved from of a set of building blocks.

As soon as you have built your application platform, along comes new open source releases for some of the components you used, and in addition, security updates are coming out and someone has to determine how critical it is to get each particular security update implemented. This starts the whole life cycle maintenance effort which must exist especially if your business is expecting to run production or mission critical applications.

The DIY team is assuming the responsibility and liability for keeping the entire ecosystem secure and compliant. We have all read about breaches that continue to happen and the exposing of thousands or millions of people's personal data which they entrusted to a company to protect. As your business scale from 1 or 2 clusters to 10 or more this gets really painful to manage effectively, including maintaining high availability and rolling updates without an outage.

There is no doubt businesses can hire and train the right people to execute all these functions, but it is hard to retain them as their skills are in high demand and the Public cloud providers, along with other businesses, who provide are running production Kubernetes, are growing quickly and they are continually looking to hire these people away.

Doing DIY is simply a tradeoff a business makes in investing resources, time and money to build, maintain, and operating an application platform versus simply buying a managed service. Skilled developers are critical to improving client experience and is your business willing to invest some of your top developers in support of an ecosystem or would you rather have them developing new applications and services to improve your client's experience with your company?

IBM's acquisition of Red Hat recognizes that for many businesses there is a need to ensure their Kubernetes ecosystem is a commercialized offering for production applications. They need: extensive quality assurance activities with regular updates tested, hardware and software certifications, and they need enterprise level maintenance backed by global support with SLAs. This is what Red Hat Openshift and Red Hat Enterpise Linux provides.

In the end, while the open source software is free to download and use, the people, their time and training is not, and it simply boils down to a choice each business has to make.

#opensource #openshift #cloudcomputing