Modernization of Data Infrastructures – where to start

Ein Blogartikel von:

Christian Gilcher

Christian Gilcher

Christian ist als Gründer und CTO Takt- und Ideengeber in Personalunion bei der ValueCloud. Er hat als IT Manager mehrere Jahre hinweg die Untiefen klassischer Enterprise-IT kennen gelernt und nutzt sein tiefes Verständnis von System Design und komplexen Software Architekturen, um moderne, Cloud native Lösungen zu konzipieren, die hohe Leistungsfähigkeit mit Robustheit und Sicherheit kombinieren.

Modernization of Data Infrastructures – where to start

The digital landscape has seen massive shifts over the last decade, with the rise of big data, AI, and cloud computing. Consequently, modernizing traditional data infrastructures is no longer a luxury for enterprises but a necessity. It enables the organization to glean insights from their data and use it to drive business decisions, improve operational efficiency, and gain a competitive advantage. But where does one start?

Modernization of data infrastructures is not a one-and-done kind of deal. It's a journey, a series of small, incremental steps. Each step, though small, delivers business value and sets the foundation for the next. This blog post will take you through three approaches that have gained significant attention lately: Event Mesh, Data Lakehouse, and Streaming ETL.

Typical Patterns and Approaches

Event Mesh:

An event mesh is an architectural layer that allows events from one application to be dynamically routed and received by any other application no matter where these applications are deployed—no matter which cloud, no matter which vendor. It's a real-time move away from traditional point-to-point integration and the silos it created. Its capabilities in exchanging data from distant systems in a loosely-coupled fashion also make the Event Mesh shine in Hybrid- and Multi-Cloud-Scenarios.

An Event Mesh typically consists of a central Event Streaming Hub and additional components, such as a Schema Registry, Source- and Sink-Connectors, Event Processing and mechanisms for strong Encrpytion & Authorization:

To integrate an event mesh into your enterprise, start by focusing on specific areas where real-time data is most needed. For instance, your customer service could significantly benefit from real-time data to respond to customer issues swiftly. By creating an event mesh, you decentralize data delivery and can react to changes in real-time. An Event Mesh as a distributed communication backbone between various IT-Systems is suitable to act as a self-service data platform, encourages domain-ownership of your companies data and can be used to enact federated governance principles.

Remember, the initial objective is not to create a comprehensive event mesh but to establish a functional section of the mesh that delivers immediate business value. Once this section is operational and delivering value, you can gradually expand the mesh to include other areas of the business.

Data Lakehouse:

A data lakehouse combines the best features of data lakes and data warehouses. It supports both structured and unstructured data, provides the strong schema-on-read and write capabilities of a data warehouse, and the scalable storage capabilities of a data lake.

Start small by identifying data silos within your organization and begin breaking them down. You can implement a data lakehouse on a single, critical data source that needs better integration and analytics capabilities.

Once implemented, you will find it easier to ingest and store data from various sources and also perform analytics at scale. Your initial data lakehouse implementation could be as simple as creating a unified view of customer data. The benefits of this are immediate and multi-fold: better customer insights, personalized marketing, and superior customer service.

Streaming ETL:

Traditional ETL (Extract, Transform, Load) processes are batch-oriented, and they can't handle the speed and volume of data that modern businesses generate and consume. Streaming ETL processes, on the other hand, can handle and process data in real-time, providing businesses with immediate insights.

Start by identifying a single data pipeline that would benefit from real-time data processing. This could be a pipeline that feeds your customer service dashboard or real-time analytics dashboard. Once you've upgraded this pipeline to use streaming ETL, measure its impact. If you find it beneficial, which you likely will, start transitioning other data pipelines to use streaming ETL.

Business Considerations

Open source software plays a crucial role in the modernization of data infrastructures. By leveraging open source technologies, enterprises can access cutting-edge solutions, foster innovation, and reduce the risk of vendor lock-in, all while ensuring greater strategic flexibility, scalability, and multi-cloud portability.

Strategic Flexibility:

Open source components provide strategic flexibility by enabling enterprises to tailor solutions to their specific needs. They are inherently modular and can be pieced together in various ways to construct a custom solution. This flexibility allows your organization to start small, focusing on areas that would benefit most from modernization.

Consider Apache Kafka, an open-source stream-processing software, which plays a crucial role in creating event-driven architectures, like an Event Mesh. By implementing Kafka, you can start with a small, manageable stream of data, and gradually expand as you see the need and value.


Open source solutions are designed for scalability. You can start with a small-scale deployment, and as your data requirements grow, the open-source solutions can grow with you. For instance, you might begin your Data Lakehouse implementation with Apache Hudi or Delta Lake on a small subset of data. As your data grows and diversifies, these solutions can scale to meet your needs, handling petabytes of data efficiently.

Multi-cloud Portability:

One of the significant advantages of open-source is that it’s platform agnostic. You can deploy open-source solutions in any environment - on-premise, cloud, or even a hybrid setup. This neutrality extends to multi-cloud setups as well, allowing you to avoid vendor lock-in and choose the best services from various cloud providers.

Let's take the case of Kubernetes, an open-source platform designed to automate deploying, scaling, and managing containerized applications. Kubernetes ensures that your Streaming ETL process, running in Docker containers, is portable across different cloud environments. This portability allows your business to take advantage of the best offerings from different cloud providers without the hassle of rearchitecting your applications.


Modernizing your data infrastructure does not require a colossal, disruptive overhaul. Instead, a series of small, incremental steps focusing on an Event Mesh, Data Lakehouse, and Streaming ETL can lead to significant improvements.

Each step is tangible and delivers real business value. Remember, the goal is not to implement state of the art technology for the sake of it, but to actually enhance your organization's problem-solving capabilities.

Modernizing your data infrastructure with open source components offers strategic flexibility, scalability, and multi-cloud portability. It allows you to create a data infrastructure that’s tailor-made for your business needs and future-proof. By starting small and scaling over time, your enterprise can stay agile, minimize risks, and deliver continuous business value.

About ValueCloud:

ValueCloud is a European leader for open source data- & AI platforms outside the traditional US-hyperscaler based infrastructure.

Lassen Sie uns wissen, wie wir SIE unterstützen können: