Apt-cacher

First, and simplest, I set up apt-cacher to save time and bandwidth. Nice and easy, especially with the tip in the last couple of comments on that article about setting apt (on all the clients) to use a proxy, which saves changing every line in sources.list. Just so I don’t lose it, here it is: create a file in /etc/apt/apt.conf.d called whatever you like (I used 90apt-proxy-local) with this line in it:

Acquire::http::Proxy “http://cache-host:3142″;

Obviously, use the actual hostname of the machine running apt-cacher there. Just as obviously, I just cut and pasted from the article linked above.

Puppet

Second, I started setting up Puppet. Since I’m using Xen on my new hardware, I’m likely to have more virtual machines kicking around than I want to manage. Automation is the obvious answer, and puppet looks pretty good so far. Having a nice version-controlled description of all the machines I’m running, which then gets enforced, is very appealing. Here’s what I’m hoping to get out of puppet:

• Documentation. I think the puppet manifests (being declarative) will serve nicely as documentation of my setup.
• Enforcement. The tool takes my documentation and makes it real.
• Backup. Keeping the complete documentation for my whole setup in a versioning tool gives me an extra layer of safety beyond normal backups.
• Ease. I want to write down, just once, how I want my machines set up. Then I want a tool that will make sure it happens, and stays happening.

We’ll see how much of that I get with puppet.