Top 4 infrastructure and operations trends for 2023

1. Platform Engineering will advance IaC/DevOps collaboration

Enabling developer self-service is not a new trend. However, enterprises have realized that maintaining internal products to manage diverging IaC and cloud platforms is both expensive and difficult.  For this reason, there’s growing discussion on how infrastructure and operations teams can collaborate better with developer teams while increasing reuse and control of the systems they manage.  

This creates a paradox for enterprises: how do you increase developer autonomy without creating an open-ended infrastructure management challenge? We believe Platform Engineering teams will separate developer UX to allow many options while operator UX will consolidate the backend control plane.  This also keeps DevOps effort from being overly opinionated.

Watch for Rob’s Platform Engineering discussion on The Cloudcast #682!  Also, check out related Cloud2030 discussions: Are Platform Teams Good?.and Platform Engineering Makes You Angry?

2. CRDs redefine Kubernetes

Custom Resource Definitions (CRDs) emerged as a way for Kubernetes to store and manage objects that were not formally defined in the Kubernetes specification.  What started as an API convenience has blossomed into a full blown secondary use-case for Kubernetes nearly as important as its core container management functionality.  

CRDs are quickly becoming the new configuration and abstraction APIs for Cloud Native applications and infrastructure.  This also reinforces the declarative programming approach that heavily influences IaC automation and is central to Digital Rebar architecture.

Related Cloud2030 discussions Explaining Kubernetes Controller Architecture and What We’re Watching At Kubecon

3. CXL disrupts hardware planning

We’d expected that SmartNICs would make more of an impact these last few years and they still may; however, momentum with the Server OEMs has shifted significantly towards servers using CXL and ARM.  The industry is finally ready to collaborate around CXL, aka Computer Express Link, as the next generation PCIe bus.  

CXL provides a standard that allows customers to break the chassis barrier when they design systems.  Specifically, customers can use CXL to combine resources (RAM, NIC, Disk, GPU, etc) from different chassis dynamically into new server configurations.  This effectively creates rack-scale composable infrastructure using a multi-vendor standard.  

This could dramatically impact how enterprises plan server commodities assuming (queue RackN theme music) there is an accompanying dynamic infrastructure management system.

4. Can edge applications break free from single-use cases?

Progress on turning edge infrastructure into IT managed infrastructure remains slow.  Part of the challenge is that edge, by definition, lacks the surrounding support infrastructure, elasticity and adjacent services of cloud. 

That makes it difficult for current IT tooling and management systems to operate effectively without an underlying IT DevOps platform (like Digital Rebar); consequently, most edge applications are single use-case focused (aka operations technology or OT).

Related Cloud2030 discussion Is Edge Impact of Digital Twins and Serverless the ultimate Edge OpEx platform?

What is RackN Doing in 2023?

Our 2023 roadmap directly supports customers’ initiatives in these areas.  We’re adding features and integrations that help Platform Teams empower self-service with less toil. Here is a peek at what we’re working on:

• Implementing  a Digital Rebar Machine CRD to enable Bare Metal Kubernetes
• Making it easier to add custom hardware types without RackN engineering
• Refining our Day 2 operations capabilities with improved orchestration and blueprints

If you want to stay informed about these topics and the RackN roadmap for Digital Rebar, please subscribe to our newsletter.

Want more predictions?  SIgn up for our newsletter!  We’ll post some more in our January newsletter.