Solution – RKE Cluster MetalLB provides Services with IP Addresses but doesn’t ARP for the address

I ran in to the the same issue detailed here working with a RKE cluster

https://github.com/metallb/metallb/issues/1154

After looking around for a few hours digging in to the logs i figured out the issue, hopefully this helps some one else our there in the situation save some time.

Make sure the IPVS mode is enabled on the cluster configuration

If you are using :

RKE2 – edit the cluster.yaml file

RKE1 – Edit the cluster configuration from the rancher UI > Cluster management > Select the cluster > edit configuration > edit as YAML

Locate the services field under rancher_kubernetes_engine_config and add the following options to enable IPVS

    kubeproxy:
      extra_args:
        ipvs-scheduler: lc
        proxy-mode: ipvs

https://www.suse.com/support/kb/doc/?id=000020035

Default

After changes

Make sure the Kernel modules are enabled on the nodes running control planes

Background

Example Rancher – RKE1 cluster

sudo docker ps | grep proxy # find the container ID for kubproxy

sudo docker logs ####containerID###

0313 21:44:08.315888  108645 feature_gate.go:245] feature gates: &{map[]}
I0313 21:44:08.346872  108645 proxier.go:652] "Failed to load kernel module with modprobe, you can ignore this message when kube-proxy is running inside container without mounting /lib/modules" moduleName="nf_conntrack_ipv4"
E0313 21:44:08.347024  108645 server_others.go:107] "Can't use the IPVS proxier" err="IPVS proxier will not be used because the following required kernel modules are not loaded: [ip_vs_lc]"

Kubproxy is trying to load the needed kernel modules and failing to enable IPVS

Lets enable the kernel modules

sudo nano /etc/modules-load.d/ipvs.conf

ip_vs_lc
ip_vs
ip_vs_rr
ip_vs_wrr
ip_vs_sh
nf_conntrack_ipv4

Install ipvsadm to confirm the changes

sudo dnf install ipvsadm -y

Reboot the VM or the Baremetal server

use the sudo ipvsadm to confirm ipvs is enabled

sudo ipvsadm

Testing

kubectl get svc -n #namespace | grep load
arping -I ens192 192.168.94.140
ARPING 192.168.94.140 from 192.168.94.65 ens192
Unicast reply from 192.168.94.140 [00:50:56:96:E3:1D] 1.117ms
Unicast reply from 192.168.94.140 [00:50:56:96:E3:1D] 0.737ms
Unicast reply from 192.168.94.140 [00:50:56:96:E3:1D] 0.845ms
Unicast reply from 192.168.94.140 [00:50:56:96:E3:1D] 0.668ms
Sent 4 probes (1 broadcast(s))
Received 4 response(s)

If you have the service type load balancer on a deployment now you should be able to reach it if the container is responding on the service

helpful Links

https://metallb.universe.tf/configuration/troubleshooting/

https://github.com/metallb/metallb/issues/1154

https://github.com/rancher/rke2/issues/3710

External Pi-hole with IPv6 – Setup a secured Pi-hole DNS service on Docker using Linode/AWS

Let me address the question of why I decided to put a DNS server (Pihole) exposed to the internet (not fully open but still).

I needed/wanted to set up an Umbrella/NextDNS/CF type DNS server that’s publicly accessible but secured to certain IP addresses.

Sure NextDNS is an option and its cheap with similar features, but i wanted roll my own solution so i can learn a few things along the way

I can easily set this up for my family members with minimal technical knowledge and unable to deal with another extra device (Raspberry pi) plugged into their home network.

This will also serve as a quick and dirty guide on how to use Docker compose and address some Issues with Running Pi-hole, Docker with UFW on Ubuntu 20.x

So lets get stahhhted…….

Scope

  • Setup Pi-hole as a docker container on a VM
  • Enable IPV6 support
  • Setup UFW rules to prune traffic and a cronjob to handle the rules to update with the dynamic WAN IPs
  • Deploy and test

What we need

  • Linux VM (Ubuntu, Hardened BSD, etc)
  • Docker and Docker Compose
  • Dynamic DNS service to track the changing IP (Dyndns,no-Ip, etc)

Deployment

Setup Dynamic DNS solution to track your Dynamic WAN IP

for this demo, we are going to use DynDNS since I already own a paid account and its supported on most platforms (Routers, UTMs, NAS devices, IP camera-DVRs, etc)

Use some google-fu there are multiple ways to do this without having to pay for the service, all we need is a DNS record that's up-to-date with your current Public IP address. 

For Network A and Network B, I’m going to use the routers built-in DDNS update features

Network A gateway – UDM Pro

Network B Gateway – Netgear R6230

Confirmation

Setup the VM with Docker-compose

Pick your service provider, you can and should be able to use a free tier VM for this since its just DNS

  • Linode
  • AWS lightsail
  • IBM cloud
  • Oracle cloud
  • Google Compute
  • Digital Ocean droplet

Make sure you have a dedicated (static) IPv4 and IPv6 address attached to the resource

For this deployment, I’m going to use a Linode – Nanode, due to their native IPv6 support and cause I prefer their platform for personal projects

Setup your Linode VM – Getting started Guide

SSH in to the VM or use weblish console

Update your packages and sources

sudo apt-get update 
install Docker and Docker Compose

Assuming you already have SSH access to the VM with a static IPv4 and IPv6 address

Guide to installing Docker Engine on Ubuntu

Guide to Installing Docker-Compose

Once you have this setup confirm the docker setup

docker-compose version

Setup the Pi-hole Docker Image

Lets Configure the docker networking side to fit our Needs

Create a Seperate Bridge network for the Pi-hole container

I guess you could use the default bridge network, but I like to create one to keep things organized and this way this service can be isolated from the other containers I have

docker network create --ipv6 --driver bridge --subnet "fd01::/64" Piholev6

verification

We will use this network later in docker compose

With the new ubuntu version 20.x, Systemd will start a local DNS stub client that runs on 127.0.0.53:53

which will prevent the container from starting. because Pi-hole binds to the same port UDP 53

we could disable the service but that breaks DNS resolution on the VM causing more headaches and pain for automation and updates

After some google fu and trickering around this this is the workaround i found.

  • Disable the stub-listener
  • Change the symlink to the /etc/resolved.conf to /run/systemd/resolve/resolv.conf
  • push the external name servers so the VM won’t look at loopback to resolve DNS
  • Restart systemd-resolved
Resolving Conflicts with the systemd-resolved stub listener

We need to disable the stub listener thats bound to port 53, as i mentioned before this breaks the local dns resolution we will fix it in a bit.

sudo nano /etc/systemd/resolved.conf

Find and uncomment the line “DNSStubListener=yes” and change it to “no”

After this we need to push the external DNS servers to the box, this setting is stored on the following file

/etc/resolv.conf
#     DO NOT EDIT THIS FILE BY HAND -- YOUR CHANGES WILL BE OVERWRITTEN
# 127.0.0.53 is the systemd-resolved stub resolver.
# run "systemd-resolve --status" to see details about the actual nameservers.

nameserver 127.0.0.53

But we cant manually update this file with out own DNS servers, lets investigate

Cartoon of a detective investigate following footprints | Premium ...
ls -l /etc/resolv.conf

its a symlink to the another system file

/run/systemd/resolve/stub-resolv.conf

When you take a look at the directory where that file resides, there are two files

When you look at the other file you will see that /run/systemd/resolve/resolv.conf is the one which really is carrying the external name servers

You still can’t manually edit This file, and it gets updated by whatever the IPs provided as DNS servers via DHCP. netplan will dictate the IPs based on the static DNS servers you configure on Netplan YAML file

i can see there two entries, and they are the default Linode DNS servers discovered via DHCP, I’m going to keep them as is, since they are good enough for my use case

If you want to use your own servers here – Follow this guide

 Lets change the symlink to this file instead of the stub-resolve.conf

$ sudo ln -sf /run/systemd/resolve/resolv.conf /etc/resolv.conf

Now that its pointing to the right file

Lets restart the systemd-resolved

systemctl restart systemd-resolved

Now you can resolve DNS and install packages, etc

Docker compose script file for the PI-Hole

sudo mkdir /Docker_Images/
sudo mkdir /Docker_Images/Piholev6/

Lets navigate to this directory and start setting up our environment

nano /Docker_Images/Piholev6/docker-compose.yml
version: '3.4'
services:

   Pihole:
    container_name: pihole_v6
    image: pihole/pihole:latest
    hostname: Multicastbits-DNSService
    ports:
      - "53:53/tcp"
      - "53:53/udp"
      - "8080:80/tcp"
      - "4343:443/tcp"
    environment:
      TZ: America/New_York
      DNS1: 1.1.1.1
      DNS2: 8.8.8.8
      WEBPASSWORD: F1ghtm4_Keng3n4sura
      ServerIP: 45.33.73.186
      enable_ipv6: "true"
      ServerIPv6: 2600:3c03::f03c:92ff:feb9:ea9c
    volumes:
       - '${ROOT}/pihole/etc-pihole/:/etc/pihole/'
       - '${ROOT}/pihole/etc-dnsmasq.d/:/etc/dnsmasq.d/'
    dns:
      - 127.0.0.1
      - 1.1.1.1
    cap_add:
      - NET_ADMIN
    restart: always

networks:
  default:
    external:
      name: Piholev6
networks:
  default:
    external:
      name: Piholev6

Lets break this down a littlebit

  • Version – Declare Docker compose version
  • container_name – This is the name of the container on the docker container registry
  • image – What image to pull from the Docker Hub
  • hostname – This is the host-name for the Docker container – this name will show up on your lookup when you are using this Pi-hole
  • ports – What ports should be NATed via the Docker Bridge to the host VM
  • TZ – Time Zone
  • DNS1 – DNS server used with in the image
  • DNS2 – DNS server used with in the image
  • WEBPASSWORD – Password for the Pi-Hole web console
  • ServerIP – Use the IPv4 address assigned to the VMs network interface(You need this for the Pi-Hole to respond on the IP for DNS queries)
  • IPv6 – Enable Disable IPv6 support
  • ServerIPv6 – Use the IPv4 address assigned to the VMs network interface (You need this for the Pi-Hole to respond on the IP for DNS queries)
  • volumes – These volumes will hold the configuration data so the container settings and historical data will persist reboots
  • cap_add:- NET_ADMIN – Add Linux capabilities to edit the network stack – link
  • restart: always – This will make sure the container gets restarted every time the VM boots up – Link
  • networks:default:external:name: Piholev6 – Set the container to use the network bridge we created before

Now lets bring up the Docker container

docker-compose up -d

-d switch will bring up the Docker container in the background

Run ‘Docker ps’ to confirm

Now you can access the web interface and use the Pihole

verifying its using the bridge network you created

Grab the network ID for the bridge network we create before and use the inspect switch to check the config

docker network ls
docker network inspect f7ba28db09ae

This will bring up the full configuration for the Linux bridge we created and the containers attached to the bridge will be visible under the “Containers”: tag

Testing

I manually configured my workstations primary DNS to the Pi-Hole IPs

Updating the docker Image

Pull the new image from the Registry

docker pull pihole/pihole

Take down the current container

docker-compose down

Run the new container

docker-compose up -d

Your settings will persist this update

Securing the install

now that we have a working Pi-Hole with IPv6 enabled, we can login and configure the Pihole server and resolve DNS as needed

but this is open to the public internet and will fall victim to DNS reflection attacks, etc

lets set up firewall rules and open up relevant ports (DNS, SSH, HTTPS) to the relevant IP addresses before we proceed

Disable IPtables from the docker daemon

Ubuntu uses UFW (uncomplicated firewall) as an obfuscation layer to make things easier for operators, but by default, Docker will open ports using IPtables with higher precedence, Rules added via UFW doesn’t take effect

So we need to tell docker not to do this when launching a container so we can manage the firewall rules via UFW

This file may not exist already if so nano will create it for you

sudo nano /etc/docker/daemon.json

Add the following lines to the file

{
"iptables": false
}

restart the docker services

sudo systemctl restart docker

now doing this might disrupt communication with the container until we allow them back in using UFW commands, so keep that in mind.

Automatically updating Firewall Rules based on the DYN DNS Host records

we are going to create a shell script and run it every hour using crontab

Shell Script Dry run

  • Get the IP from the DYNDNS Host records
  • remove/Cleanup existing rules
  • Add Default deny Rules
  • Add allow rules using the resolved IPs as the source

Dynamic IP addresses are updated on the following DNS records

  • trusted-Network01.selfip.net
  • trusted-Network02.selfip.net

Lets start by creating the script file under /bin/*

sudo touch /bin/PIHolefwruleupdate.sh
sudo chmod +x /bin/PIHolefwruleupdate.sh
sudo nano /bin/PIHolefwruleupdate.sh

now lets build the script

#!/bin/bash
PATH=/sbin:/bin:/usr/sbin:/usr/bin
now=$(date +"%m/%d/%T")
DYNDNSNetwork01="trusted-Network01.selfip.net"
DYNDNSNetwork02="trusted-Network02.selfip.com"
#Get the network IP using dig
Network01_CurrentIP=`dig +short $DYNDNSNetwork01`
Network02_CurrentIP=`dig +short $DYNDNSNetwork02`
echo "-----------------------------------------------------------------"
echo Network A WAN IP $Network01_CurrentIP
echo Network B WAN IP $Network02_CurrentIP
echo "Script Run time : $now"
echo "-----------------------------------------------------------------"
#update firewall Rules
#reset firewall rules
#
sudo ufw --force reset
#
#Re-enable Firewall
#
sudo ufw --force enable
#
#Enable inbound default Deny firewall Rules
#
sudo ufw default deny incoming
#
#add allow Rules to the relevant networks
#
sudo ufw allow from $Network01_CurrentIP to any port 22 proto tcp
sudo ufw allow from $Network01_CurrentIP to any port 8080 proto tcp
sudo ufw allow from $Network01_CurrentIP to any port 53 proto udp
sudo ufw allow from $Network02_CurrentIP to any port 53 proto udp
#add the ipV6 DNS allow all Rule - Working on finding an effective way to lock this down, with IPv6 rick is minimal
sudo ufw allow 53/udp
#find and delete the allow any to any IPv4 Rule for port 53
sudo ufw --force delete $(ufw status numbered | grep '53*.*Anywhere.' | grep -v v6 | awk -F"[][]" '{print $2}')
echo "--------------------end Script------------------------------"

Lets run the script to make sure its working

I used a online port scanner to confirm

Setup Scheduled job with logging

lets use crontab and setup a scheduled job to run this script every hour

Make sure the script is copied to the /bin folder with the executable permissions

using crontab -e (If you are launching this for the first time it will ask you to pick the editor, I picked Nano)

crontab -e

Add the following line

0 * * * * /bin/PIHolefwruleupdate.sh >> /var/log/PIHolefwruleupdate_Cronoutput.log 2>&1
Lets break this down
0 * * * *

this will run the script every time minutes hit zero which is usually every hour

/bin/PIHolefwruleupdate.sh

Script Path to execute

/var/log/PIHolefwruleupdate_Cronoutput.log 2>&1

Log file with errors captured

Vagrant Ansible LAB Guide – Bridged network

Here’s a is a quick guide to get you started with a “Ansible core lab” using Vagrant.

Alright lets get started

TLDR Version

  • Install Vagrant
  • Install Virtual-box
  • Create project folder and CD in to it
Vagrant init
  • Vagrantfile – link
  • Vagrant Provisioning Shell Script to Deploy Ansible – link
  • Install the vagrant-vbguest plugin to deploy missing
vagrant plugin install vagrant-vbguest
  • Bring up the Vagrant environment
Vagrant up

Install Vagrant and Virtual box

For this demo we are using windows 10 1909 but you can use the same guide for MAC OSX

Windows

Download Vagrant and virtual box and install it the good ol way –

https://www.vagrantup.com/downloads.html

https://www.virtualbox.org/wiki/Downloads

https://www.vagrantmanager.com/downloads/

Install the vagrant-vbguest plugin (We need this with newer versions of Ubuntu)

vagrant plugin install vagrant-vbguest

Or Using chocolatey

choco install vagrant
choco install virtualbox
choco install vagrant-manager

Install the vagrant-vbguest plugin (We need this with newer versions of Ubuntu)

vagrant plugin install vagrant-vbguest

MAC OSX – using Brewcask

Install virtual box

$ brew cask install virtualbox

Now install Vagrant either from the website or use homebrew for installing it.

$ brew cask install vagrant

Vagrant-Manager is a nice way to manage all your virtual machines in one place directly from the menu bar.

$ brew cask install vagrant-manager

Install the vagrant-vbguest plugin (We need this with newer versions of Ubuntu)

vagrant plugin install vagrant-vbguest

Setup the Vagrant Environment

Open Powershell

to get started lets check our environment

vagrant version

Create a project directory and Initialize the environment

for the project directory im using D:\vagrant

Open powershell and run

mkdir D:\vagrant
cd D:\vagrant

Initialize the environment under the project folder

vagrant init

this will create Two Items

.vagrant – Hidden folder holding Base Machines and meta data

Vagrantfile – Vagrant config file

Lets Create the Vagrantfile to deploy the VMs

https://www.vagrantup.com/docs/vagrantfile/

The syntax of Vagrantfiles is Ruby this gives us a lot of flexibility to program in logic when building your files

Im using Atom to edit the vagrantfile

Vagrant.configure("2") do |config|
     config.vm.define "controller" do |controller|
                  controller.vm.box = "ubuntu/trusty64"
                  controller.vm.hostname = "LAB-Controller"
                  controller.vm.network "public_network", bridge: "Intel(R) I211 Gigabit Network Connection", ip: "172.17.10.120"
                    controller.vm.provider "virtualbox" do |vb|
                                 vb.memory = "2048"
                  end
                  controller.vm.provision :shell, path: 'Ansible_LAB_setup.sh'
   end
   (1..3).each do |i|
         config.vm.define "vls-node#{i}" do |node|
                       node.vm.box = "ubuntu/trusty64"
                       node.vm.hostname = "vls-node#{i}"
                       node.vm.network "public_network", bridge: "Intel(R) I211 Gigabit Network Connection" ip: "172.17.10.12#{i}"
                      node.vm.provider "virtualbox" do |vb|
                                                  vb.memory = "1024"
                     end
              end
        end
end

You can grab the code from my Repo

https://github.com/malindarathnayake/Ansible_Vagrant_LAB/blob/master/Vagrantfile

Let’s talk a little bit about this code and unpack this

Vagrant API version

Vagrant uses API versions for its configuration file, this is how it can stay backward compatible. So in every Vagrantfile we need to specify which version to use. The current one is version 2 which works with Vagrant 1.1 and up.

Provisioning the Ansible VM

This will

  • Provision the controller Ubuntu VM
  • Create a bridged network adapter
  • Set the host-name – LAB-Controller
  • Set the static IP – 172.17.10.120/24
  • Run the Shell script that installs Ansible using apt-get install (We will get to this below)

Lets start digging in…

Specifying the Controller VM Name, base box and hostname

Vagrant uses a base image to clone a virtual machine quickly. These base images are known as “boxes” in Vagrant, and specifying the box to use for your Vagrant environment is always the first step after creating a new Vagrantfile.

You can find different base boxes from app.vagrantup.com

Or you can create custom base boxes for pretty much anything including “CiscoVIRL(CML)” images – keep an eye out for the next article on this

Network configurations

controller.vm.network "public_network", bridge: "Intel(R) I211 Gigabit Network Connection", ip: "your IP"

in this case, we are asking it to create a bridged adapter using the Intel(R) I211 NIC and set the IP address you defined on under IP attribute

You can the relavant interface name using

get-netadapter

You can also create a host-only private network

controller.vm.network :private_network, ip: "10.0.0.10"

for more info checkout the network section in the KB

https://www.vagrantup.com/docs/networking/

Define the provider and VM resources

We declaring virtualbox(we installed this earlier) as the provider and setting VM memory to 2048

You can get more granular with this, refer to the below KB

https://www.vagrantup.com/docs/virtualbox/configuration.html

Define the shell script to customize the VM config and install the Ansible Package

Now this is where we define the provisioning shell script

this script installs Ansible and set the host file entries to make your life easier

In case you are wondering VLS stands for V=virtual,L – linux S – server.

I use this naming scheme for my VMs. Feel free to use anything you want; make sure it matches what you defined on the Vagrantfile under node.vm.hostname

!/bin/bash
sudo apt-get update
sudo apt-get install software-propetise-common -y
sudo apt-add-repository ppa:ansible/ansible
sudo apt-get update
sudo apt-get install ansible -y
echo "
172.17.10.120 LAB-controller
172.17.10.121 vls-node1
172.17.10.122 vls-node2
172.17.10.123 vls-node3" >> /etc/hosts

create this file and save it as Ansible_LAB_setup.sh in the Project folder

in this case I’m going to save it under D:\vagrant

You can also do this inline with a script block instead of using a separate file

https://www.vagrantup.com/docs/provisioning/basic_usage.html

Provisioning the Member servers for the lab

We covered most of the code used above, the only difference here is we are using each method to create 3 VMs with the same template (I’m lazy and it’s more convenient)

This will create three Ubuntu VMs with the following Host-names and IP addresses, you should update these values to match you LAN, or use a private Adapter

vls-node1 – 172.17.10.121

vls-node2 – 172.17.10.122

vls-node1 – 172.17.10.123

So now that we are done with explaining the code, let’s run this

Building the Lab environment using Vagrant

Issue the following command to check your syntax

Vagrant status

Issue the following command to bring up the environment

Vagrant up

If you get this message Reboot in to UEFI and make sure virtualization is enabled

Intel – VT-D

AMD Ryzen – SVM

If everything is kumbaya you will see vagrant firing up the deployment

It will provision 4 VMs as we specified

Notice since we have the “vagrant-vbguest” plugin installed, it will reinstall the relevant guest tools along with the dependencies for the OS

==> vls-node3: Machine booted and ready!
[vls-node3] No Virtualbox Guest Additions installation found.
rmmod: ERROR: Module vboxsf is not currently loaded
rmmod: ERROR: Module vboxguest is not currently loaded
Reading package lists...
Building dependency tree...
Reading state information...
Package 'virtualbox-guest-x11' is not installed, so not removed
The following packages will be REMOVED:
  virtualbox-guest-utils*
0 upgraded, 0 newly installed, 1 to remove and 0 not upgraded.
After this operation, 5799 kB disk space will be freed.
(Reading database ... 61617 files and directories currently installed.)
Removing virtualbox-guest-utils (6.0.14-dfsg-1) ...
Processing triggers for man-db (2.8.7-3) ...
(Reading database ... 61604 files and directories currently installed.)
Purging configuration files for virtualbox-guest-utils (6.0.14-dfsg-1) ...
Processing triggers for systemd (242-7ubuntu3.7) ...
Reading package lists...
Building dependency tree...
Reading state information...
linux-headers-5.3.0-51-generic is already the newest version (5.3.0-51.44).
linux-headers-5.3.0-51-generic set to manually installed.

Check the status

Vagrant status

Testing

Connecting via SSH to your VMs

vagrant ssh controller

“Controller” is the VMname we defined before not the hostname, You can find this by running Vagrant status on posh or your terminal

We are going to connect to our controller and check everything

Little bit more information on the networking side

Vagrant Adds two interfaces, for each VM

NIC 1 – Nat’d in to the host (control plane for Vagrant to manage the VMs)

NIC 2 – Bridged adapter we provisioned in the script with the IP Address

Default route is set via the Private(NAT’d) interface (you cant change it)

Netplan configs

Vagrant creates a custom netplan yaml for interface configs


Destroy/Tear-down the environment

vagrant destroy -f

https://www.vagrantup.com/intro/getting-started/teardown.html

I hope this helped someone. when I started with Vagrant a few years back it took me a few tries to figure out the system and the logic behind it, this will give you a basic understanding on how things are plugged together.

let me know in the comments if you see any issues or mistakes.

Until Next time…..