Work Tech Blog Issue 2019.46

For all of our readers out there (if there are any 😜), I would like to start a weekly blog series about our work tasks and what technical news have moved us during the last week. You may ask why I am going to write such a blog series? For me, there are mainly two reasons. The first reason is, that due to this weekly blog format I hopefully will enforce myself to write on a more regularly basis and the second reason is, that there are so much changes and topics going on that it might be interesting for readers to get some insights about working in a technological permanently changing field.

The “W” and the “T” inside the blog logo represents the words “work”, “technology”, “weekly”, “team” or something like that. Just pick what fits best for your 😂! The date which is written with a 8 bit font represents the year and the number of the week inside the year. The logo itself should be a tribute to the connection of the analogue and digital world.

And now, after this short introduction, lets start with the highlights of the last week!

We hit Let’s Encrypt’s rate limit

😂 Yes, we hit the rate limit of Let’s Encrypt. As you might know, Let’s Encrypt limits the creation of new SSL certificates. Due to the usage of GitLab pipelines and the very neat creation process which was invented by Bernhard, it is easy for our colleagues to request SSL certificates for their new applications and domains. This means, that we are creating more official certificates than ever before and more and more sites are becoming SSL encrypted.

Docker sold it’s enterprise business

One headline of this week news was the acquisition of Docker Inc.’s enterprise business by Mirantis. In my opinion something like that was predictable over the last two years. But there is no need to panic because a software, especially a Open Source software does not stop to work for this reason. Inside the Docker Community Leader Channel (which I am part of) there were some smaller changes. One of them was that one of our CL-Guardians, Karen Bajza, will join Mirantis. Also, some of the colleagues working at Docker will move to Mirantis within the next days. Well, we will see if the focus on the development and desktop experience of Docker will bring the shift. Docker is still a force in the container world as the Stackoverflow survey 2019 shows - Docker ranked third in the platform category.

AWS Route53 Serverless Backup

Nico Schilling from Nordcloud provided a cool skeleton for a AWS Route53 Zone backup with a serverless AWS Lambda function. We brought in some small changes and now everyone could use this solution. If you are interessted, head over to the GitHub repository and take a look! It’s always nice to work as team together! 🚀😊

Elasticsearch logging improvements

Since we are now running ECE on-premises, we are able to use a lot of new features from Elastic out of the box now. We did some research this week about what can be replaced or even optimized by using the Elastic Metric Beat or the Elastic File Beat more widely. We think that we can achieve some quick-wins for the logging of our load balancers (several hundred). In combination with our friends from the development department, who are using Elastic to monitor their single page applications, we are pretty sure to establish a very informative dashboard in the future.

RedHat open sources Project Quay

This week RedHat open sources Project Quay which also includes the Quay.io registry including the Clair scanner. The different reporters on the internet are not sure why this step has been done now but some of them think that RedHat hopes that more developers will join the project and that therefore the project will be used more broadly. The responses on the internet forums and the mainstream tech-news pages were restrained. Well we will see.

GitLab job runner inside AWS

We also did some research this week about GitLab runners in AWS with on-premise connection to speed up the automatic GUI tests for the developers. Together we stitched together a main plan about what we would like to do. Roughly described, we are thinking about the possibility to have around 50 GitLab runners (EC2) in stopped state inside AWS and after a central service receives a webhook GitLab job call, it starts one of the waiting AWS runners. The startup time of the instance takes around 30 seconds but this is OK because the GUI test job runs for about 7 to 8 minutes than. After the job completes, the instance will be shut down. This would enable us to use instances with more performance for a short period of time.

Closing

That’s it for Week 2019.46! Please let me know if you are interested in such kind of a weekly blog! Write me an email or get in contact with me by one of the possibilities below ⬇⬇⬇⬇! Thanks!