Tech Radar

FinOps helps quantify the TCO of the operations in the cloud era, It introduces an operating model which enables organisation’s to understand cloud costs and make tradeoffs.

With the the growing consumption of managed services FinOps adoption will become a standard similarly as DevOps revolutionised development by breaking down silos and increasing agility, FinOps increases the business value of cloud by bringing together technology, business and finance professionals with a new set of processes & tools.

FinOps is built like any data driven processes and requires 3 stages Inform -> Optimise -> Operate. The Inform phase consists of transparency of all operational processes and their Total Operational Cost, this data helps drive Optimisation of processes and eventually define how new business initiatives are measured based on the Operational KPI’s.

There are many OpenSource project which offer a good starting point for FinOps worth noting ArgoCD which now indicate the workflow cost projects like KubeCost etc.

FinOps helps quantify the TCO of the operations in the cloud era, introducing an operating model which enables organisations to understand cloud costs and make tradeoffs. With the the growing consumption of managed services FinOps adoption will become a standard, in the same way that DevOps revolutionised development by breaking down silos and increasing agility. FinOps increases the business value of the cloud by bringing together technology, business and finance professionals with a new set of processes & tools. FinOps is built like any data-driven process, and requires 3 stages: Inform -> Optimise -> Operate. The Inform phase consists of transparency of all operational processes and their Total Operational Cost. This data helps drive Optimisation **of processes and eventually defines how new business initiatives are measured based on the **Operational KPI’s. There are many open-source project which offer a good starting point for FinOps, for example Argo Workflow and KubeCost.

Data is becoming a first-class member in most of the projects today. AI and Machine Learning endeavor emphasized the need for accurate and repeatable data, especially when building and tuning AI models. For that purpose, many tools and frameworks are built, which allows better control over the data and its various versions. Not only schema versioning, but also the data itself needs to be verified, tested, and reviewed. Much like code, there are tools for CI, CD, and data monitoring.

In the 3rd volume of the tech radar that was published in early 2019 we identified remote work as a trend, seeing more and more development organizations adopting the methodology of remote work. Today, post COVID-19 we’re seeing a new reality of hybrid remote. This new situation brought organizations to change their entire perspective and start thinking and behaving differently, embracing and adapting to the new reality. Hybrid remote means combining work in the home environment, synchronous and async processes, creating many new situations and dilemmas that we haven’t faced before. There are a number of different ways to implement hybrid-remote: regular office days, employees that come to the office every day versus those that work only remotely. Therefore, there is more focus on results versus the number of hours invested. Onboarding of new employees is still a big challenge, and there is still a state of mind that creativity and brainstorming belong to the on-life versus online world. There is also an open question regarding the sustainability of all of this - what will happen in “the day after” COVID-19. 

The battle over leading the container orchestration market seems to be dominated by Kubernetes and its fast-growing ecosystem and standards. Alongside the standards of OCI, CNI, CSI and others we now have a platform that encourages extendability. In the past 5 years, we’ve witnessed Operators and Controllers which take Kubernete’s core values and solve many long-living problems such as stability, extendability, and most importantly maintainability. A few key examples inspired by CNCF projects like: - Workflow and pipelines like Argo and Tekton, that can be used for data pipelines and CI/CD - ClusterApi and Crossplane for the creation and management of infrastructure

• ArgoCD, Flux GitOps implementation as Kubernetes deployment tools
• K3s and others for edge-computing
• Kube Bench for security checks
• Kube Cost for FinOps   These Are just a few examples of the “Kubernetes platform” trend.  

As we can see that Kubernetes become the standard for running workloads of microservices. Almost every product or service runs on Kubernetes. Due to the standardization applied by CNF and the extendability mechanism of is APIs (including CRD’s and Kubernetes reconciliation loop that implements the desired state), we see an enormous wealth of tools and frameworks based on Kubernetes in lots of domains.

One product, one codebase is the latest trend of the mobile app solutions market right now. Most people believe that single codebase applications are of lower quality and performance as compared to native apps. But by using the latest technology, we can make single codebase apps that are high on quality and performance like native apps such as Flutter and React Native.

We have a lot of advantages when using cross platform frameworks:

• Easier Implementation for both Android and iOS devices. And non mobile apps like Web, Windows, Mac, Linux etc…
• Uniformity across all devices - gives a uniform feel to the user across all the devices, standard UI and a very good UX
• Reusable code - developers will not have to write unique code for every platform
• Reduced development time and costs - you only need 1 team to write 1 code instead of 2 teams writing the same code on different platforms.
• Maintenance is easy because we have a single codebase for all devices
• Greater Revenue - the more platforms you can reach without spending more money on development is a great value for business!

There are also some disadvantages in using cross platform frameworks:

• Vendor Lock-In → many cross-platform frameworks depend primarily on companies like Google and Facebook.
• Lack of Updates → each time there is an update or a new feature released for the native platform the framework will have to update as well
• Integration Problems → because of cross platform frameworks immaturity we might have difficulties using third party services