The year 2019 is about to end. Less than one month left. A lot of websites already try to predict the future in their posts about upcoming cloud computing trends for 2020. This article provides an overview of some of these trends. Big disclaimer here, no one can really predict the future. But at least, we can anticipate it. The list of trends is never complete, but exiting things are about to come. Read on to find out about what (hopefully) will be hot the next year.
A fast moving target
A lot of topics and technologies which are cloud related evolve very fast. It’s no surprise that developers and companies have a hard time keeping up. It’s difficult to constantly align with the latest trending technologies and new tool versions. The number of new cloud native services which cloud providers offer grows very fast.
Some patterns can be seen in this light. Patterns like the increased popularity of open source software for example. Or what about DevSecOps which is getting more popular by shifting security left (into the hands of developers for self-service consumption). With the rise of containers, workloads are more portable than ever.
All of these patterns accelerate cloud computing possibilities. The “target state” is evolving faster than ever! Maybe that is a trend as well? 🙂
No one can deny the popularity of Kubernetes nowadays. In 2019 a lot of large enterprises focused their attention to use Kubernetes for their application workloads. In 2020, it’s expected to continue. Kubernetes also plays a vital role in their multi-cloud strategy. With Kubernetes as the standard orchestration tool for running containers, it’s easier to swap the infrastructure layer for another one. From EKS to AKS to on-prem. Tools like Rancher combine all of these. In the end it’s “just Kubernetes” for all of these platforms. This trend makes the underlying infrastructure becoming less important. Focus can and probably will be put elsewhere.
We see companies utilizing Kubernetes from multiple cloud providers. Rancher provides a way to manage these as well as Hashicorp Consul. Companies dare to expand their workloads to multiple cloud providers and pick the best from both worlds. Looks like Kubernetes provides a lot of (portability) freedom and flexibility for companies when it comes to running containers and handle the integration points as well.
2. Role and influence of CNCF
The Cloud Native Compute Foundation is setting the goals and direction for Kubernetes and other projects like Envoy and FluentD. Cloud providers incorporate open source initiatives, for example AWS uses Clair to scan container images in their container registry (ECR) but at the same time competes with other open source initiatives like the ELK stack. AWS provides and promotes their own solution called Cloudwatch in favor of the ELK stack.
By identifying these trends, it’s about consolidation and competition. It can create more silos since solutions are becoming “standard” and highly portable as well as “cloud provider specific” at the same time. Thus creating a vendor lock-in. An interesting, but conflicting trend compared to the cloud agnostic movement.
3. Security almost comes “out of the box”
With more and more companies moving to the public cloud, the increased pressure on organizations (especially in the EU) about data protection and privacy, security becomes even more important. A lot of exiting things happened already in 2019: Twistlock has been acquired by Palo Alto. Harbor, the open source container registry has built-in vulnerability security scanning for container images as well as the option to digitally sign images and Microsoft promotes Azure Kubernetes Services (AKS) together with Azure Security Center.
Also in 2020 security will remain a top priority for organizations. A lot of vendors in the DevOps space and cloud providers offer options to secure applications. Container run-time protection is becoming more important than ever, also for workloads which have no clear infrastructure layer (e.g. AWS Fargate, Azure functions, AWS lambda) but in fact, that infrastructure layer still exists (in the hands of the cloud provider) and needs to be protected as well.
One of the leading security companies is Sysdig. This security specialist put together an extremely useful webpage which is all about run-time security from the perspective of containers, Kubernetes and serverless. Organizations who face the challenge of securing cloud native services like Elastic Kubernetes Services (EKS) from AWS as well as Kubeless can benefit from it.
With all of these trends, security comes more and more “out of the box”, developers and security teams can utilize whatever is there to help them protect valuable data and applications. These trends put the “shift left principle” into practice. Theory is over, it’s about implementation now.
4. Serverless becomes even more popular…but
Serverless computing is about abstracting the infrastructure layer from the application layer. Developer teams do not have to maintain the infrastructure layer like VMs or containers anymore. Instead, this is realized by cloud providers which offer serverless compute features as a cloud native service. Organizations pay on a “per use base”. For example the time a function needs to execute a piece of code, the amount of CPU power and memory which is used for that function, etc.
Even “hybrid/mixed” solutions are there to help with the migration from traditional VMs to containers to serverless. For example: Knative provides serverless features on top of Kubernetes.
There are a number of articles and papers which mention the growing popularity of serverless. For example the article on the website of CACM, and the article called Why Serverless Is the Uber of Infrastructure. Reading these articles, you would expect it’s only pros which serverless brings over traditional infrastructure. For example, serverless is mentioned as #1 in the top 5 of the article on the website of mombileappdaily.
Despite the fact that serverless is getting more popular every day, there are also trends which indicate companies are not so eager to jump in. Some people state that monoliths are still very beneficial since development and deployment patters are clear, it has less moving parts compared to a big number of containers, etc. Also a lot of organizations face quite some challenges to adopt the principles of serverless instead of VMs and containers. It’s hard sometimes to predict the true costs of serverless infra/components so these factors slow down adoption a bit. Yet another trend which is moving in multiple directions.
5. Focus on application development
The trends mentioned before are great improvements which help DevOps teams to speed up release cycles for their applications. From an application development perspective, this is all great, but still eats away a lot of time to actually learn and implement these features. Even if it really benefits the organization in the long run.
At the same time we see the following. Containers are not new anymore, some argue containers are boring :-). That means, organizations are used to it and know how to find their way in this world.
Perhaps one of the biggest advantages of this trend is the amount of effort developer teams need to spend on the configuration and maintenance of infrastructure (components). When focusing on containers and serverless, less code and less (valuable) time is needed to setup and maintain the infrastructure layer for applications. Code bases become smaller, less room for (human) errors, improper configuration, misalignment (with other teams) and reducing the attack surface since less customized code is needed. Developers are no infrastructure security experts, it’s better to have them develop applications instead.
With this in mind, there is more time to actually deliver business features. Applications with meaningful business features matter most, since they add real value for the end customers.
No and Low Code solutions (like those from BettyBlocks, Outsystems and Mendix) will democratize software development as a trade, making it available to citizen developers to create business applications without too much hassle.
Taking all of this in mind, the focus is back on application development and less on other aspects which do not contribute significantly to an organization.
As mentioned before, the list of future trends is and will never be complete. I hope this article has inspired you to keep an eye on features which might come and further evolve. Things which you can use to grow your organization. Stay tuned, since 2020 for sure will have more exiting cloud computing trends to explore.