The main goal of this workshop is to introduce various aspects of systemd to all of you who haven't used it extensively yet.
Our main requirement is having Linux distribution with systemd version at least 229. It might be installed on some virtual machine on your laptop. Just make sure it works and running following command returns required systemd version:
$ systemctl --version systemd 229
You may install Fedora 24 (server or workstation edition) - it already has systemd 229 installed.
Second requirement would be having installed Python in the 2.7 version (that's ok to have 3.x also, but it'll be up to you to make our code work on 3.x; we try to keep codebase as simple as possible in order to focus on what's really important). It'd be also very convenient to have virtualenvwrapper, but that's not a really requirement as long as you're ok with installing system - wide libraries on your laptop (or VM).
Third requirement is git. That's ok if you're not very fluent with it - we'll show you what to do.
How to register?
Registrations starts 17th Nov 2016 at 18:00 on http://systemd.evenea.pl
We've got 25 slots.
Who's the organizer?
That's actually community event. Organizer would be Anonymous Admins group
That will be Lumesse office in Kraków
Lumesse-Guest / Lumesse4488
26th November 2016, 11:00
Who's gonna be the teacher?
During this part we'll discover various binaries provided by systemd and use some ot those.
Unit files are a basic configuration tools in systemd which define behaviour for the whole system.
We'll learn what types of unit files systemd provides, what is the purpose of them
and how to correctly write a new one. Thanks to that knowledge both creating quick ad hoc
and fairly complicated chain of
services in separate
target shouldn't be a problem.
Here be dragons. We'll dive into specifics of managing logs with journald. Starting with reading logs and finishing on creating very specific queries. If there's enough time maybe we'll try to push logs to central repository?
We don't need Docker to play with containers, systemd-nspawn will let you run both system and aplication containers providing stable workflow and predictable behaviour. We'll learn how to create containers, restrict their capabilities and resources and write custom unit files for them. We'll also try to run Docker container under systemd-nspawn supervision to understand differences and strong/weak points in both sollutions.
We'll learn how to integrate Python (and maybe C and Java?) applications with systemd making systemd aware of applications' state. Thanks to that application operational aspects like automated failover, restart and others will be done without manual intervention and in a clear, and proper way.