Hi,
my name is Kaspar, and I’m the CEO of Humanitec. I’m writing to you from the table in my kitchen in Berlin, and I hope this message finds you well. I recently had a conversation with my friend Erik Muttersbach around how to set engineering teams up for success. He encouraged me to share some of the insights we gathered throughout the last years.
We sit on a ton of information and insight into the software industry. We’ve talked to hundreds of teams (over 560 teams in 18 months!) and we perform real-time analysis of all the conversations in the cloud-native tooling space.
This email is the start of a series for our “developer productivity” newsletter which we plan to send out every few weeks. Topics will range from DevOps automation hacks and engineering best practices to up and coming CI/CD tools or Kubernetes news. And these are just to name a few.
Enough said, let’s get started with these first topics. If you want to know more about Humanitec here are some of our upcoming Webinars.
Thank you for being here with us,
Kaspar
CEO of Humanitec
+++The Developer Productivity Newsletter+++
Insights we picked up in the industry
A question that constantly bugs us: what is the future of Kubernetes and what does the world actually think about it?
On a global level, we currently register around 200 conversations on the keywords Kubernetes and Helm charts a day. And where these conversations take place are in blog posts, forum post discussions, or in groups. Daily total impressions on these conversations exceed 170,000. And of those conversations, 26.1% are positive, 9.1% negative, and 64.8% neutral. Altogether, the attention towards this topic has grown 13% in the last quarter.
Interestingly enough a lot of conversations reflect the early state of this technology and the pace at which it’s evolving. The general distribution and adoption of best practices and the general understanding of Kubernetes is still comparable poor. From what we see this leads to failures in implementation as teams choose a self-maintained K8s setup and often for the wrong use-cases. That makes them miss out on using a lot of great tools along the way. In the end, Kubernetes was built to build platforms on top, but only a few teams can afford to run it themselves. The most interesting trends can be found in these platforms on top of Kubernetes as they combine all the value but without the downside. Below are a few linked articles that explain this thought further.
There was a super interesting discussion going on in the /DevOps sub-threat on Reddit whether one should split CI and CD in different tools. While there is a convenience aspect in keeping both sides in one tool there are a good amount of reasons against.
Marli Thumbarle, portworx CEO, wrote a great article for NewStack on the state of Kubernetes, and it’s worth checking out. He details how CIOs are looking at this trend and fleshes out some of the mistakes lined out above.
We’re in love with...
Harbor’s (the image registry) new update Harbor 2.0. This release makes Harbor the first OCI (Open Container Initiative)-compliant open source registry capable of storing a multitude of cloud-native artifacts like container images, Helm charts, OPAs, Singularity, and much more. We believe Harbor is the #1 registry, and no, we don’t hold a special relationship with them.
InfraApp is another great tool that has come to our attention. It’s a tool developed by former Docker employees. It allows you to manage and monitor K8s from your Desktop with a wonderful flow, well functioning, and slick. It’s just really well done.
Some cool content from us
We love this piece by Chris Stephenson about environment management in Kubernetes. He also gave a talk on the topic. In short: from 5+ services on, really make sure you think about environment management properly. Otherwise the dependencies and communication in your team grow and you enter microservice hell. Not a place you want to end up in.
Antoine Rougout made a good amount of noise with his piece on the “challenges managing environment variables”. Antoine is a hard-core user of Humanitec and shares his insights managing variables as a team with hands-on examples.
Our colleague Nils is a huge fan of the Robot Framework. He published this piece on why the framework is key to improving testing. Testing and QA in general is one of the areas that make it so vital to do your container and Kubernetes set up correctly from the start. Think of all the time you lose if you haven’t parameterized your environment variables but want to dynamically test subsets of features…
