Seite wählen

NETWAYS Blog

DEV Retreat 2018 Recap

von | Sep 27, 2018 | ,

We had a lovely time somewhere in the „Franconian Alps“, away from daily development business. Our dev retreat always includes one day of „looking back, looking forward“, the second day is reserved for fun stuff.
After a relaxed breakfast on Friday, Marius invited us into the icebreaker game. Rock-Paper-Scissors combined with catching the other team on win, really a challenge and discussions on failure afterwards. Before diving into a strategic outlook with Icinga and NETWAYS development, Flo made a shout-out for the first round of beer pong. Another group was like „hey, what’s cards against humanity with monitoring?“ and just started playing, that was pure fun 🙂
Later on, we did a full retrospective of the past year in our development team. Everyone shared their current mood, with then collecting three important topics (good or bad) from the past year. This entire timeline with many milestones was put in contrast with everyone’s mood over time. Reflecting the not-so-good tasks towards the „we really love this project and technology“ we created our own team „graph“.
Having shared this many feelings and milestones, we made it into another round – direct feedback on everyone’s important points, be it good or bad. We’ve then built a topic cloud from these and voted topics we would like to discuss and define solutions. The two of them were selected and we started to discuss, help each other, criticise specific points and agree on a solution together. Pure energy, and not even the BBQ being ready could have stopped us. Knowledge transfer, trainee practice, more direct feedback, interesting technologies and new projects coming up …
After all we enjoyed a lovely evening and night sitting by the campfire, playing another round of beer pong, chatting and enjoying life.
The next day our plan was to escape a room. 60 minutes, lots of puzzles and hints, to finally solve everything. It is hard to describe as everyone finds his/her role pushing the others to the next level and puzzle. And there’s discussion and „anger“ afterwards that you’ve lost 20 minutes on such a stupid question … but you failed together, and that’s a good feeling even 🙂
Next to getting to know colleagues and new trainees better, my personal highlights this year are these two quotes:

During our debugging sessions, I never had that feeling that I am treated as a „dumb“ trainee. My ideas and knowledge are taken into account, and I really like this in our team.

Your mission as Senior Developer or experienced developer is to share your knowledge and experience with your colleagues. Help them improve and learn from you, with the most detailed feedback you can give.

And on we go, we have to develop projects like Icinga DB, Reporting, NoMa, Mobile, support daily business with Icinga 2, Icinga Web 2, Icinga Exchange, Request Tracker, Pretix, Discourse, GitLab, Jenkins and so much more filling the timeline for next year’s dev retreat already 🙂

How css-tricks improved my work life

von | Aug 30, 2018 | , ,

My relationship with CSS is some kind of a love-hate one and I think that is a sentiment a lot of web-devs share with me.
My biggest issue with it is, that it is really counterintuitive and most of the time it feels a bit like solving puzzles to make things align the way you want it to.

and oftentimes it feels a lot like


Properties that mean the same thing being worded inconsistently – I’m looking at you…

vertical-align: middle;
text-align: center;

And there is also some questionable naming with ‚grey‘ (#808080) being darker than ‚darkgrey‘ (#a9a9a9)…
In general there are too many quirks you need to just know. Unless you work with it every day, you need to google every second line in order to write your stylesheets.
There you can find all sorts of guidelines, help, blogposts with a nice mix of bad-practices and outdated information.
What helped me a lot is the website css-tricks.
mehr lesen…

Feu Mourek
Feu Mourek
Developer Advocate
Feu verbrachte seine Kindheit im schönen Steigerwald, bevor es sich aufmachte die Welt zu Erkunden. Seit September 2016 unterstützt es Icinga zunächst als Developer und seit 2020 als Developer Advocate, und NETWAYS als Git und GitLab Trainer. Seine Freizeit verbringt es hauptsächlich damit Video-, und Pen and Paper Rollenspiele zu spielen, sich Häuser zu designen (die es sich nie leisten können wird) oder ganz lässig mit seinem Cabrio durch die Gegend zu düsen.
Lies mehr von Feu und triff unser Team

A peek into GitLab 11, the Web IDE and Auto DevOps with Kubernetes

von | Jun 28, 2018 | , ,

GitLab 11 was just released last week and introduces a bunch of new features. I’ve picked the most exciting ones for a short peek as our hosted production environment was already upgraded by Stefan 🙂
 

Web IDE

You can already add and edit files directly in your browser. Dirk mentioned this in his blog post, it motivates open source contributors to enhance the documentation. For larger changes we are used to fire up an IDE like Visual Studio, JetBrains Goland, PHPStorm, Atom, Eclipse, etc.
Sometimes you don’t need a full blown IDE, or you don’t have access to on your mobile device. It would be nice to have immediate results from build, compile and test stages, best in an isolated (container) environment.
The newly introduced GitLab Web IDE attempts to become a new player in the field. On the right you can already see the current pipeline job status from the latest commit. Once we’ve edited the file, we can commit (and push at once) and the CI will trigger a new job run. Just awesome!

In its current implementation, the IDE is accessible from files view only, but the roadmap proves that there’s more to come on project creation. Just type t in the main project view and search for a specific file. Then pick Web IDE from the upper left bullets and rock on. There is syntax highlighting and code auto-completion included ❤️


 
 

Auto DevOps with Kubernetes

Our vision is to replace disparate DevOps toolchains with a single integrated application that is pre-configured to work by default across the complete DevOps lifecycle.

Auto DevOps enables your CI/CD workflow to advance even further with pipeline container environments built on Kubernetes. GitLab provides its own Kubernetes integration, accompanied with default templates for projects and build pipelines. This isn’t limited to build, test, deploy stages but also includes a variety of additional stages:

  • Build and Test
  • Code Quality checks
  • Static Application Security Testing (SAST), Dynamic Application Security Testing (DAST), Dependency and container scanning for vulnerabilities
  • Review Apps, deploy current state to temporary application environment in Kubernetes (optional)
  • Deploy to Kubernetes production environment (optional)
  • Monitoring for fetching deployment metrics from the Kubernetes cluster via Prometheus exporter (optional)


 
You can see the changes inside the menu where Operations was introduced and provides settings for environments and Kubernetes integration. This was previously found in the CI/CD section. A Kubernetes cluster can be assigned to a specific project, you can safely play and test without harming a global environment.
Nicole and Gabriel shared more insights into GitLab 11 features in their OSDC talk on „Git things done with GitLab!“. Check the archive for a live demo especially on the Auto DevOps feature, first time I fully understood its purpose 🙂
 

Notable Changes

In terms of roles, GitLab renamed „Master“ to „Maintainer“.

GitLab also decided to release the „Squash and Merge“ option into the Open source edition. This feature allows you to develop a merge request in multiple commits, and upon merge back to master, it will automatically squash the commits into a single one. Want to learn more about git squash? We’ve got you covered in our GitLab training 🙂


I really hope that GitLab will also open-source Merge request approvals in future releases. This is something which we use for code review and release managed on GitHub on a daily basis and would enhance our GitLab flow too.
Still not enough? I’d love to talk GitLab with you in our upcoming training sessions, including the best development workflows from our daily work 🙂

Releasing our Git and GitLab training as Open Source

von | Mai 24, 2018 | , ,

Development is super fast these days, there are many tools and integrations to make it more comfortable. One of the version control systems is Git, next to well-known SVN and CVS. Git got really pushed by GitHub and most recently GitLab which provide entire collaboration suites. Nowadays you not only commit code revisions/diffs, you can also test and deployment changes immediately to see whether your code is good on defined platforms or breaks. Issue and project management is put on top and makes it very easy to collaborate.
Continuous integration (CI) allows for deeper code quality too. Add code coverage reports, unit test results, end2end tests and actually build distribution packages with tests on many platforms included. In addition to CI, continuous deployment (CD) adds the icing on the cake. Once CI tests and package builds are fine, add a new build job to your pipeline for automated deployments. This could for example push updated RPM packages for your repository and immediately install the bugfix release in production on all your client hosts with Puppet or Ansible.
You can do all of this with the ease of GitHub or your own hosted GitLab instance (try it out in NWS right now!). Don’t forget about the basics for development and devops workflows:

  • Untracked files, staging area, … what’s within the .git directory?
  • What is a „good commit„?
  • I want to create patch for an open source project – what’s a „pull request„?
  • Workflow with branches?
  • A developer asked me to „rebase your branch against master“ and „squash the commits“ … what’s that?

Our Git training sessions have been renewed into a two day hands-on session on Git and GitLab. Many of us are using Git on a daily basis at NETWAYS, in addition to GitLab. Knowledge which we share and improve upon. The training starts with the Git basics, diving into good commits, branching, remote repositories and even more. Day 1 also provides your own NWS hosted GitLab instance.
Starting with day 2, you’ll learn about development workflows with branches and real-life use cases. Continuing with CI/CD and generating your own Job pipeline, and exploring GitLab even further. We’ll also discuss integrations into modern development tools (Visual Studio, JetBrains, etc.) and have time to share experiences from daily work. I’ve been working with Git since the beginning of Icinga more than nine years ago.
We have open-sourced our GitLab training material. We truly believe in Open Source and want make it easier for development and contributions on your favourite OSS project, like Icinga.
You are welcome to use our training material for your own studies, especially if you are an open source developer who’s been learning to use Git, GitLab and GitHub. For offline convenience, the handouts, exercises and solutions are provided as PDF too.
Many of the mentioned practical examples and experiences are only available in our two day training sessions at NETWAYS so please consider getting a ticket. There’s also time for your own experience and ideas – the previous training sessions have shown that you can always learn something new about Git. You can see that in the Git repository and the newer Git commits, where this feedback was added to the training material ❤️
See you soon at the famous NETWAYS Kesselhaus for a deep-dive into Git and GitLab!
Please note that the training material is licensed under Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International.

Ansible, so einfach!

von | Jun 30, 2017 | , , , ,

Konfigurationsmanagement in Rechenzentren ist aus der modernen „DevOps“ IT nicht mehr wegzudenken. Puppet, Chef, Salt oder Ansible automatisieren Umgebungen von mittelständischen Unternehmen bis hin zu Weltkonzernen.
Wenn wir die verschiedenen Lösungen betrachten, bedienen sie sich alle einer eigenen Sprache, diese soll möglichst einfach unsere Infrastruktur abbilden. („infrastructure as a code“)
Da nicht jeder Admin im Tagesgeschäft programmiert, kann so eine Sprache ein Hindernis werden und Arbeitsschritte vielleicht verkomplizieren.
Seit einiger Zeit beschäftige ich mit Ansible, ein Tool welches ohne vorinstallierte Clients arbeitet und eine Konfiurationsprache basierend auf YAML nutzt.
Um Ansible zu nutzen muss lediglich eine SSH Verbindung, ein Benutzer mit Rootrechten und ein Inventarfile bestehen.
Das Sprache im YAML Format ist für jeden leicht lesbar und sofort verständlich.
Ansible kann entweder lokal auf dem Arbeitsrechner oder auf einem sogenannten „Managementnode“ über bereitgestellte Pakete installiert werden.
Unter „/etc/ansible/“ legen wir nach der Installation zusätzlich ein Inventar an.

$ cat /etc/ansible/hosts
host01.localdomain ansible_ssh_host=10.10.10.11
host02.localdomain ansible_ssh_host=10.10.10.12</pre lang="bash">

Wenn Ansible installiert und ein Inventarfile erstellt wurde, kann die Verbindung zum Server mit dem Modul „ping“ getestet werden. Hierbei versucht Ansible den Server per SSH zu erreichen und einzuloggen.

$ ansible host01.localdomain -m ping
host01.localdomain | success >> {
"changed": false,
"ping": "pong"
}</pre lang="bash">

Alternativ kann statt dem Hostalias auch „all“ gesetzt werden um alle Hosts aus dem Inventar zu prüfen.

$ ansible all -m ping</pre lang="bash">

Ansible bietet zahlreiche Module mit denen Pakete installiert, Dateien kopiert oder bearbeitet werden können. Hier findet ihr eine Übersicht aller Module
Tasks verwenden diese Module um die Arbeitsschritte in Playbooks oder Rollen auszuführen.
Beispiel eines Playbooks:

$ cat ansible_starter.yml
---
- hosts: all <- Führe die Tasks auf allen Hosts aus
  tasks:
    - name: say hello to everyone <- Name des Tasks
      debug:                      <- Name des Moduls
        msg: "Hello World"        <- Parameter des Moduls

– name: install ntp
yum:
name: ntp
state: installed
</pre lang=“yaml“>

$ ansible-playbook -s ansible_starter.yml
</pre lang="bash">
Diese Task werden der Reihenfolge nach auf dem Zielhost ausgeführt und damit haben wir schon eine kleine Automation geschaffen. Probiert's aus!

Da Ansible im Sturm mein Automationsherz erobern konnte war ich mit meinem Kollegen dieses Jahr auf dem Ansible Fest in London, einen Erfahrungsbericht der Veranstaltung findet ihr hier.

Thilo Wening
Thilo Wening
Manager Consulting
Thilo hat bei NETWAYS mit der Ausbildung zum Fachinformatiker, Schwerpunkt Systemadministration begonnen und unterstützt nun nach erfolgreich bestandener Prüfung tatkräftig die Kollegen im Consulting. In seiner Freizeit ist er athletisch in der Senkrechten unterwegs und stählt seine Muskeln beim Bouldern. Als richtiger Profi macht er das natürlich am liebsten in der Natur und geht nur noch in Ausnahmefällen in die Kletterhalle.
Lies mehr von Thilo und triff unser Team

Trainings

Web Services

Events