What is this thing?
How do I use it?
What are the components?
Something Isn't Working
What's New?

Navigate using the menu along the top

Labs Menu
Chaos Menu
Observe Menu
System Menu

back

This is Labinator. It is a mobile lab built on cheap hardware bought online and sprinkled with a bunch of 3D printed stuff that lets you run/observe experiments.

It demonstrates the creation of a new Kubernetes cluster from scratch with Talos Linux. Labinator uses freely licensed Open Source projects available on the Internet to do its work.

To learn a LOT more about labinator, its history, and the build, scan the QR code next to the power supply. Or, just go to https://labinator.bitnebula.com/

back

About Labs

There are three types of labs available. The Virtual Machine lab will start hypervisors on all six of the lab nodes and launch Talos on each. The Bare Metal lab will launch Talos natively on all six nodes. The Hybrid lab will launch Talos for the control plane on the top three nodes and hypervisors for the workers on the other three nodes.

Labinator will automatically power stuff on and off to support the lab. It will also trigger disk wipes when needed.

back

About Chaos

What good is a lab without experimenting? The Chaos menu lets you inject a little insanity into the lab.

You can create/destroy load pods to add load to the system, kill VMs or nodes, restart nodes (which WILL fail), and even create network partitions!

back

About Observe

Here you will find links to a few dashboards and visibility options. Since the labs will display system health on the screen of each system, most of the options here will tell you about Boss. You can also see logs (better than Statusinator in the bottom right displays) for the entire lab.

When you click a link, the bottom screen will display the selected option.

back

About System

You can turn the lab on and off with this menu. Restarting the lab will restart this machine (Boss) after killing power to all of the nodes. Shutting down the lab will shut down Boss after killing power.

You shouldn't be able to do much damage by just yanking the power from Labinator whenever you want... but it also wouldn't hurt to shut down first.

back
System layout
Experiment nodes
Infrastructure bits
Back
Node1Node3Node5
  • IP: 192.168.122.11
  • Power: Port 1
  • Net: Port 3
  • IP: 192.168.122.13
  • Power: Port 3
  • Net: Port 5
  • IP: 192.168.122.15
  • Power: Port 5
  • Net: Port 8
Node2Node4Node6
  • IP: 192.168.122.12
  • Power: Port 2
  • Net: Port 4
  • IP: 192.168.122.14
  • Power: Port 4
  • Net: Port 6
  • IP: 192.168.122.16
  • Power: Port 6
  • Net: Port 7
back
BossWallySwitch
  • IP: 192.168.122.3
  • Power: Port 7
  • Net: Port 2
  • OS: Debian GNU Linux
  • Location: Under bottom touchscreen
  • IP: 192.168.122.1
  • Power: -
  • Net: Port 1
  • OS: OpenWRT
  • Location: Under top touchscreen
  • IP: 192.168.2.2
  • Power: -
  • Net: N/A
  • OS: N/A
  • Location: Bottom right of lab
back

back

To Fix Issues:

Turn it off and back on again

(Use the System link above)

 

No... really

It's magic!!!

Just kidding

You could also just retry the lab

Labinator is made of cheap components

Things go wrong... often

Just try it again after reboot

There are several known issues

In order of frequency, sometimes...

a monitor doesn't turn on

Boss doesn't serve TFTP

a VM disappears

labwatch misses important progress

a hypervisor freezes due to load

Relayinator doesn't know a port's state

Follow the QR code on the power supply for more
Back

Virtual Machines Only
Bare Metal Only
VM and BM Combo
Launch VM lab with...
2 VMs per host
3 VMs per host
4 VMs per host

 

⚠️Too many VMs can freeze our cheap little boxes!

Launch Bare Metal lab...
All six nodes
Launch Bare Metal control plane with...
2 VMs per worker node
3 VMs per worker node
4 VMs per worker node
Change the Load
Tweak the Network
Kill a Box
Restart a Box
Add 1 client - 3 clients
Kill 1 client - 3 clients

Add 1 server - 3 servers
Kill 1 server - 3 servers
Enable Node: 1  2  3  4  5  6
Disable Node: 1  2  3  4  5  6

Enable Group:  AZ1  AZ2  AZ3
Disable Group:  AZ1  AZ2  AZ3
Kill Box 1
Kill Box 2
Kill Box 3
Kill Box 4
Kill Box 5
Kill Box 6
Restart Box 1
Restart Box 2
Restart Box 3
Restart Box 4
Restart Box 5
Restart Box 6
Load Status
Lab Status
Boss
Kubernetes
Logs
VM Screens
Dashboard
htop
tcpdump
system journal
access log
Node1 Node2 Node3 Node4 Node5 Node6
VM1
VM1
VM1
VM1
VM1
VM1
VM2
VM2
VM2
VM2
VM2
VM2
VM3
VM3
VM3
VM3
VM3
VM3
VM4
VM4
VM4
VM4
VM4
VM4

NOTE: These are all possible VMs. If the current lab does not have a VM running yet or will not start a VM, nothing will happen when trying to view the VM's output

Restart the Lab
Shut Down the Lab