It is often said that event-based systems are particularly powerful because events are ideal to exchange information about technical events between various systems and thus integrate them. In fact, however, a completely different picture is often clear: each system uses its own version of phenomena, the format is not compatible, and the integrations expected through events fail quickly. The clearly missing here is a standardized format – a common standard that defines which data should have an event, how it is structured and how it is transmitted.
fortunately, Cloud domestic computing foundation (CNCF) accepted this challenge and created such a standard: Cloud eventsThis subject is very relevant for every developer who works with an event-based systems. Therefore, I would like to give you more detailed insight into this blog post.
Golo Roden is the founder and CTO of the native web GMBH. It deals with web and cloud application as well as conception and development of APIs with web and cloud application focusing on event-powered and service-based distribution. Their guiding theory is that software development is not an end in itself, but always has to follow an underlying professionalism.
What are the events anyway?
Maybe you first wonder what I mean by the word “event”. An accurate answer will go beyond the structure at this point. Therefore, I recommend my blog post on event sourcing, which was published a few weeks ago. It is worth acquainting to read itself before and only again.
If you already know what events mean, you will probably be able to know the problem: Each event-operated system defines its own format for events, which makes them more difficult to exchange and integrate between different systems. The theoretical advantage of a common form of communication through events is often unused in practice, as the formats do not fit only together. As long as there is no binding standard, the benefits of such events are limited for system integration.
This is exactly where the concept of cloud events comes. The idea behind it is simple: it is about creating a standardized format for events to enable the difference between different systems. As I said, this standard cloud was developed by native computing foundation, a non-profit organization, supported by almost all major cloud providers such as Microsoft, Amazon and Google.
Recommended editorial material
With your consent, an external YouTube video (Google Ireland Limited) has been invited here.
Always load youtube videos
What are cloud events? // German
Meaning and purpose of cloud events
The goal of cloud events is to define the same metadata for events to make them compatible at least their external structure. In addition, platforms and transport layer-individent structures are to be made possible- for example, JSON must be useful through https in the form of XML via MQTT. Standard defines central words such as “ID”, “source”, “type” and other essential areas.
Why should you rely on this standard? Clear advantage lies in a significantly better gap between the system, along with reducing conversions and pursing efforts. In addition, the cloud event supports various protocols and can be moved in a flexible manner – regardless of the specific transmission path. It is relevant not only for applications, but also for routers and brokers, as events can be analyzed, filtered and distributed due to standardized metadata. In addition, the structured details of events improve traceability, for example in relation to the origin and time of the event. On this basis, generic tools for events can be felt-one aspect that is particularly important for microsarvis, serverless functions and comparable architecture. And if you continue this idea, the cloud event will also be well suited as a firm format for event sourcing.
What does Cloud speaks against events?
Of course, the question arises as to which challenges or objections may speak against the use of cloud events. A central problem is that many systems have not yet supported this standard. This means that there are currently only a few applications that are directly compatible. Current, proprietary event format should so at least be supported. This causes extra effort that you really want to avoid. This is not only applied to small systems, but also applies to AWS to Amazon Eventbridge or Microsoft Azure to services like event grid.
Another point is that the events of the cloud only standardize the metadata. User data – ie Payoad – especially remains for the related application. This system is -in -case because the technical content of an event cannot be defined in the generic. In addition, through their structured size, the events of the cloud can cause more overheads than simple ownership formats, especially if all the desired areas are not used.
Nevertheless, I think the cloud event is a big step in the right direction. Without clear rules, this doubles is unnecessarily complicated for phenomena-based systems. A common terminology seems to be very intelligent here. It is also positive that the cloud event can be combined with other standards – for example Partial Or OpentelemetryIn all, the standard is designed to simplify as possible and facilitate integration in existing systems.
How this implementation is successful
This is the result of this question how to implement cloud events wisely. First of all, you should clarify whether the use for your own application makes sense at all. According to my individual assessment, the answer is clear: as soon as you work with events and as soon as you think of pairing or integrating your software with other systems, you should consider cloud events. Of course, you can also define your own format – however, you will win very little. Only the “disadvantage” is that you have to familiarize yourself with the cloud event standard. In my experience, it is very possible: standard is understandable and manageable.
If you choose to use a cloud event, you should do it correctly. Above all, it means dealing with the meaning and format of individual areas. It will not be very useful to formally support the format, but to use incorrect values- especially if such errors appear only months later in the event of integration. Fields like “source”, “subject” or “type” require special attention. It is not complicated, but needs to read carefully Specification,
If you already use an event-based system, which is not yet taken into account the cloud event standard, then there is no reason to reject the subject. With comparatively low effort, you can use cover, converter or middleware to translate between existing formats and standard. This is especially meaningful in the long run, as they install a similar exchange standard in this way and reduce the requirement of individual adapters. It becomes particularly relevant. If you also use the cloud event, as the basis of event sourcing – events are already standardized.
Quo Vadis?
Finally, the question is whether cloud events will set themselves as a standard in the long term. I am optimistic in this regard. Cloud providers’ support is increasing, as well as distribution in open and close sources applications. In addition, cloud events already serve as the basis of other standards, for example CDVentsThis especially focuses on continuous distribution processes. This development shows that cloud events are considered a viable and durable basis.
So I can only advise you to deal with the cloud event intensively. The standard is helpful, well thought out and, above, practical. The more they work with an event-based architecture, this concept becomes for you in development, infrastructure or in other areas-more relevant.
If you have already gained experience with cloud events or experience to deal with it, I look forward to your response. And if you typically want to know more about events-based systems: In addition to introduction to the pre-mentioned event sourcing, you will also get more information as the basis of modern, flexible and scalable application architecture to CQRS in blog posts.
Continuous delivery is good, progressive distribution is better-Online conference is dedicated to this motto Gitops On 30 April 2025, both infrastructure as well as the challenges of applications can be consistent and reproductive.
This program is aimed at software developers: Inside, devops engineers and platform teams who take responsibility for software take responsibility for successfully developed and provided.
Highlights from the program:
By 2 April Tickets with initial bird discount Safe!
(May)