Daniel Mierswa, Christoph Pegel, Leonard Dahlmann and Linux
Talking about dynamic DNS everybody associates those providers who provide free subdomains which you can update to point at your IP address like DynDNS and No-IP. However, there are many disadvantages when using those free services and they aren’t very flexible (DynDNS can be quite flexible when you pay for it though).
In this post we will discuss our very own way of providing dynamic DNS without depending on any of those free providers.
Booting of a root filesystem which is created or managed by a userspace utility which is not present in the kernel needs some extra configuration. It needs an additional image which is loaded right before entering the root filesystem. In this entry we will explain the boot process of the kernel, the meaning of a root filesystem and how all those things interact to get your system running. Hopefully by the end of this entry you can say your baseless fears goodbye and handle any userspace created root filesystem on your own.
As promised our first technical blog contains the setup of our filesystems. First of all we rebooted our server into the rescue mode to setup LVM (which is the first step on all of our machines anyway except for probably RAID and LUKS setup) which provides amazing flexibility when creating, resizing, backing up and managing filesystems.
This blog will contain ideas, tricks, security and configuration related issues and any kind of work we stumbled upon setting up services and other things on our shiny new root server which some people might find helpful when setting up their own server.