The best part of my job is talking to really smart DevOps and Ops/SRE engineers who understand Kubernetes and the surrounding ecosystem well. They understand that running Kubernetes in any meaningful way requires a significant investment in the right people, the right tools, and time.
But even the smartest teams can fall into the “we can build this inhouse” trap, which leads to companies wasting time experimenting with OSS packages before either abandoning projects, or eventually looking at commercial products to get back on track. Broadly speaking, the following steps need to be taken for you to successfully build a Kubernetes management and operations platform inhouse:
- Hire: You have to find the budget to hire and task additional Developers, DevOps, Operations personnel along with Product Managers who can manage the development process for the new platform.
- Train: Each of these people will need to have a deep level of understanding of the problem you are trying to solve at scale, and this will cost you time and training dollars.
- Build: With teams hired and trained, you’ll now start building a platform. Has leadership considered the opportunity cost of the initial rollout?
- Document: The tribal knowledge your teams will acquire will need to be captured and communicated to others.
- Sustain: Even if you build the perfect solution, you’ll need to keep it up to date and improve upon it. A team of engineers will be required for care & feeding purposes.
- Retain: Kubernetes and microservices skills are in significant demand and our customers are telling us turnover is 20-30% per year.
The harsh reality is that even though some teams could certainly implement an elegant, fully automated platform given enough time and resources, most companies will never be able to allocate enough time and resources to build such a platform.
Kubernetes may be critical to your business, but it is:
- Likely not core to your product offering, and
- A wheel that has already been invented as a consumable service by Rafay.
Don’t waste precious engineering bandwidth on non-core technologies. Whether you end up engaging with Rafay or not, it would behoove every company exploring Kubernetes to evaluate 3rd party options to:
- Accelerate your time to market,
- Avoid technical debt, and
- Eliminate the need to hire teams of expensive hard to retain employees
Not doing so would be fiscally irresponsible.
Rafay delivers a managed Kubernetes service that automates multi-cluster management and containerized application operations across cloud regions, data centers and the Edge. Companies can leverage Rafay to get containerized applications operational today, leading to a faster time to market. And because Rafay also offers a white glove support model to customers, companies can operationalize containerized apps today while partnering with Rafay to internally build a Kubernetes-focused skillset.