How to deploy Anbox Cloud on bare metal
To deploy Anbox Cloud on a public cloud (such as AWS, Azure or Google) or using MAAS or OpenStack, see the instructions in How to deploy Anbox Cloud with Juju.
Alternatively, you can follow the instructions in this document to use the manual cloud provider that Juju offers. This method allows you to deploy Anbox Cloud with Juju on a set of SSH connected machines.
Before you start the installation, ensure that you have the required prerequisites:
- At least three Ubuntu machines. See Minimum hardware for details and recommendations.
- Your Ubuntu Pro token for either a full Ubuntu Pro subscription or a Ubuntu Pro (Apps-only) subscription. If you don’t have one yet, speak to your Canonical representative. If you already have a valid Ubuntu Pro token, sign in on https://ubuntu.com/pro to retrieve it.
Warning: The Ubuntu Pro (Infra-only) token that every user gets for free does NOT work and will result in a failed deployment. You must purchase a full Ubuntu Pro or a Ubuntu Pro (Apps-only) subscription by contacting Canonical.
Juju is a tool for deploying, configuring and operating complex software on public or private clouds.
You must install a Juju client on the machine that you use to run the deployment commands. To install Juju 2.9, enter the following command:
sudo snap install --classic --channel=2.9/stable juju
See Juju version for information about which Juju version is required for your version of Anbox Cloud.
Add a controller and model
The Juju controller is used to manage the software deployed through Juju, from deployment to upgrades to day-two operations. One Juju controller can manage multiple projects or workspaces, which in Juju are known as models.
You should dedicate one machine as the Juju controller. Run the following command to bootstrap the controller onto that machine:
juju bootstrap manual/<user>@<controller IP address> anbox-cloud
Juju will connect to the machine via SSH as the specified user and install all necessary requirements.
When the controller is set up, create a model to hold the Anbox Cloud deployment:
juju add-model main
Add all machines
Before starting the deployment, you must add all machines to the Juju model. See Minimum hardware for the list of machines that you need.
When adding the machines, start with the machine that you want to host the management layer of Anbox Cloud. Then add all LXD worker nodes. Run the following command for each machine:
juju add-machine ssh:<user>@<machine IP address>
The user (for example,
ubuntu) must have administrator rights on the machine and have permission to SSH to the machine.
Warning: Make sure to add the machines by their IP addresses rather than their DNS names. Adding a machine by DNS name does currently not work.
Juju will add the machines to its list of usable machines, which you can display with the
juju list-machines command. Make sure that all machines are in the
started state before you proceed. If any of the machines are still in
down state, wait until they switch to
Machine State DNS Inst id Series AZ Message 0 started 192.168.1.9 manual:192.168.1.9 jammy Manually provisioned machine 1 started 192.168.1.10 manual:192.168.1.10 jammy Manually provisioned machine
Attach your Ubuntu Pro subscription
ua.yaml overlay file as described in How to deploy Anbox Cloud with Juju.
You must provide this file when deploying Anbox Cloud.
Determine the machine mapping
When running the deployment command, you must map the machines to the ones described in the Juju bundle that you are deploying.
juju list-machines to display the available machines:
Machine State DNS Inst id Series AZ Message 0 started 192.168.0.9 i-09a2fdb5e7a2e8385 jammy localhost-1a running 1 started 192.168.0.10 i-00a05065e2768be5d jammy localhost-1b running
anbox-cloud-core deployment bundle requires two machines:
0 is supposed to host the AMS service.
1 is used for LXD. Check the
bundle.yaml file in the bundle for details.
anbox-cloud bundle requires two additional machines to host the load balancer (
0) and the extra services required for streaming (
1). For this bundle, the AMS machine is
2 and the LXD machine is
3. Check the
bundle.yaml file in the bundle for details.
--map-machine argument for the
juju deploy command maps the machines defined inside the bundle to those your Juju controller has registered in the model. See the Juju documentation for more details. If you added the machines in the order Juju expects them, the mapping is very straight-forward:
--map-machines 0=0,1=1 for the
anbox-cloud-core bundle or
--map-machines 0=0,1=1,2=2,3=3 for the
By default, Anbox Cloud uses a loop file with an automatically calculated size for LXD storage. For optimal performance, however, you should use a dedicated block storage device. See LXD storage for more information.
There are different ways of configuring a dedicated block storage device:
- Use an existing LXD storage pool (recommended - see Existing storage pool below)
- Use a dedicated storage device (see Dedicated storage device below)
- Use a storage device defined by Juju (see the Customise storage section in How to deploy Anbox Cloud with Juju for instructions)
Existing storage pool
To use an existing LXD storage pool, set the
storage_pool configuration on the AMS charm to the name of the LXD storage pool that you want Anbox Cloud to use.
For example, to use an existing LXD storage pool with the name
my-zfs-pool, use an overlay file with the following content:
applications: ams: options: storage_pool: my-zfs-pool
Important: The LXD storage pool must use the ZFS storage driver. Other storage drivers are not supported by Anbox Cloud.
Dedicated storage device
To use a dedicated storage device that is not defined by Juju for LXD storage, set the
storage_device configuration on the AMS charm to the path of the storage device.
For example, to use
/dev/sdb as the dedicated storage device, use an overlay file with the following content:
applications: ams: options: storage_device: /dev/sdb
Important: The path to the dedicated storage device must be identical for all machines that are part of the cluster.
You do not need to prepare the storage device in any way. AMS takes care of creating the LXD storage pool on the device.
Deploy Anbox Cloud
Now you can deploy Anbox Cloud. The deployment is entirely handled by Juju and does not need any manual involvement other than running the actual deploy command.
Choose between the available Juju bundles:
For a minimised version of Anbox Cloud without the streaming stack, run the following command to deploy the
juju deploy anbox-cloud-core --overlay ua.yaml --map-machines 0=0,1=1
For the full version of Anbox Cloud, run the following command to deploy the
juju deploy anbox-cloud --overlay ua.yaml --map-machines 0=0,1=1,2=2,3=3
You can watch the status of the deployment with the following command:
watch -c juju status --color --relations=true
Last updated 5 days ago.