Skip to content

Articles

Automating Mod Downloads from Curseforge for Servers

I like modded Mincraft. I was trying to setup a modded server from Curseforge and found out that there are a number of mods that are flagged to not auto download with the rest of the pack. During the server pack install a file called "MODS_NEED_DOWNLOAD.txt" will be created containing various mods that meed this criteria. The expectation being you need to download each one and copy it to the server pack mods directory.

Nomad and Traefik

Note

There is a lot here. I've done my best to make things correct, but I'm also still learning how a lot of nomad, consul, etc... work and their best practices. It's entirely possible I missed something or could have done something better.

The goal here is to be able to have a web application available at the address https://whoami.practicalkungfu.net. To do this we'll use Traefik, which is a very nice reverse proxy/load balancer that has integrations with a number of container orchestration platforms.

Consul and DNS

Note

There is a lot here. I've done my best to make things correct, but I'm also still learning how a lot of nomad, consul, etc... work and their best practices. It's entirely possible I missed something or could have done something better.

In the last couple posts I described how to setup a small(ish) Hashi Nomad and Consul cluster. Using Consul for service discovery is something that took me a little while to figure out.

Lets start by spinning up the whoami job if you haven't already.

Nomad Jobs

Note

There is a lot here. I've done my best to make things correct, but I'm also still learning how a lot of nomad, consul, etc... work and their best practices. It's entirely possible I missed something or could have done something better.

This is pretty easy once you get the hang of the job definitions. It's honestly not much diffent than docker-compose definitions.

Nomad, Consul, Gluster...Oh My

Note

There is a lot here. I've done my best to make things correct, but I'm also still learning how a lot of nomad, consul, etc... work and their best practices. It's entirely possible I missed something or could have done something better.

The goal here is to setup a 5 node nomad cluster (3 servers and 5 clients), a 5 node consul cluster, and a storage backend using a 3 node gluster cluster (I love saying gluster cluster). Gluster is mounted to a VIP provided by keepalived. I've found this gives a decent amount or resiliancy.

For the nodes in this cluster I use Debian 12, Armbian, or DietPi. Basically all Debian 12.

Poor Mans Connection Monitoring

This morning I had need to be alerted if I ping starts failing in an ad-hoc situation. The below script does a test ping and then beeps if the ping fails. It'll loop endlessly.

while ($true) {
    $ping = Test-Connection -ComputerName 10.120.10.100 -Count 1
    $ping
    Start-Sleep -Seconds 1
    If (($ping -eq "") -or ($ping -eq $null)) {
        [console]::beep(500,1000)
    }
}

Greenshot Intune Deployment

Greenshot is a very nice program. However there is one thing that REALLY annoys me about it. The hard coded browser poppup after install. Deploying Greenshot via Intune is not the most straight foward thing if you want to customize the install. Plus there is having to kill greenshot running under the deploy account (usually system) and also starting greenshot under the current logged on user after install.

For this I used this intune winget template. It consists of your usual detection, install, and uninstall scripts. Below I'll be going over the install script.

Configure Windows Time

AD members

This will reset a domain member to use NT5DS to set the time. This should be the default for "most" domain environments 

net stop w32time
w32tm /unregister
w32tm /register
net start w32time
w32tm /config /syncfromflags:DOMHIER /update
net stop w32time
net start w32time
w32tm /resync /rediscover /nowait