Cockpit Managed VM Flist Release

Release Notes: Kinetic-Cockpit-NginxRC1/RC2

Introduction: The Cockpit managed Ubuntu Full VM image included in this release is intended to server as a “developers’ playground” for the threefold grid and address numerous problems facing new developers using and developing for the threefold grid. This image utilizes the 22.1 Kinetic Kudu cloud image with the Cockpit interface from Cockpit.Project and a NGINX web server pre-installed.

Deployment: This image can be deployed on any 3node, to do so you will select “Full VM” and under the OS drop down you will select other and paste this link into the box.

upon deployment you will find a Ubuntu server running 22.1 kinetic kudu with a webserver welcome page at its public IP address and a management interface at the public ip address with port 9090 appended. The interface deploys with no password set for the root account which is required to login to cockpit. In order to set one you will need to SSH the deployment and use “passwd root”

alternatively this link deploys without ssh, and with no password set on root, but be warned until you correct this situation anyone can log in to the web interface by entering “root” with no password

Functionality: This VM image is equipped with everything you need in order to build your own Flist for the Threefold Grid and/or run a webserver that you have complete control of. In the base image that you just deployed you will find the following sections.

Overview- This is your HUD for your deployment with information deployments health, update status and current workload, this page is also where you can manage active domain Aswell as your keys that will be used to connect to other deployments.

Logs-This tab show you the output of your servers logs

Storage- This tab is where you can manage the drives attached to your deployment, This tab also includes support for mounting off-grid network file systems, you must perform the network and updates fix for nfs mounts to install/function

Networking- As it sounds, here you will manage your network interfaces and the firewall of your deployment on initial deployment you are unable to manage the interfaces themselves in order to fix this you will need to follow this tutorial on fixing network management and system updates

after the above fix

Virtual Machines- This tab uses the QEMU hypervisor running on your server to allow you to run instances of other operating systems within your deployment. This tool can be used to create images for deployment on the grid and to host additional resources within your deployment that you can then use nginx to proxy to from your public address

Accounts- this is where you can manage all of your user accounts, manage passwords, ssh keys that are allowed for users and etc.

Services- this tab lists all services installed on the machine and allows you to manage their running/boot status

Applications: this inteface monitors what Cockpit Plugins you have installed.

Software Updates: this allows you to manage the update status of the underlying 22.1 kinetic kudu image, this tab will be non-function without this fix

Terminal: This gives you browser access to the native terminal of the 22.1 kinetic kudu image the same as if you had sshed into the node

Resources for using these images

steamcmd setup

Nginx Setup

Nginx SSl with Certbot

Cockpit SSL with Lets Encrypt

Apache Web Server

Apache Web server SSL with Lets Encrypt

The project Itself’s documentation


these are the videos you need

The Threefold Hub Home of Custom Images for the Threefold Grid

Cockpit Update and Network Interface Fixes on the Threefold Grid

Single Interface multi node gui management setup on the Threefold Grid

This video is from before there was an flist but has alot of good info.
Initial Cockpit Deployment On the Threefold Grid

1 Like

Continuing the discussion from Cockpit Managed VM Flist Release:

This video walks through deployment and how this image can be used to create new cloud images for the grid

There is some import potential here in that you could create your prototype vm images using this process and then use the hub apt and terraform from the terminal interface to deploy those cloud images to the hub/ into production.

I just had a go, and it’s very cool to see a VM come alive on the Grid without needing SSH! One note for anyone else giving this a shot, the VM image is large enough that nodes may take more than five minutes to download it. In that case, you’ll see an error during deployment. The node will continue downloading the image, so just try again and it should work.


I have just discovered that if you perform the update fix and the node your cockpit instance is running on reboots you will ultimately lose everything in that deployment. Apparently network manager is entirely incompatible with the grid.

If you have something running in cockpit right now you need to go in and on the terminal, re-enable systemd-networkd. I’m sorry this wasn’t known earlier, I had discussed this fix and how I was doing things with multiple people before I released the video. Also 219 restarted yesterday, so if your deployment was there and it now has no ip addresses on the play interface. There is not a fix.

I tried to reproduce the issue you saw with Network Manager. The only nodes I have the power to reboot at home don’t have public IPs, so I tried without one. Same basic steps you ran in the video, with a couple small differences. I both started and enabled the Network Manager service, which I did after netplan apply. Will try to do some more testing tomorrow to see if I can duplicate.

You can deploy on 219/testnet and shoot me a message and I’ll give it a reboot.