Skip to main content

Your submission was sent successfully! Close

Thank you for signing up for our newsletter!
In these regular emails you will find the latest updates from Canonical and upcoming events where you can meet our team.Close

Thank you for contacting us. A member of our team will be in touch shortly. Close

An error occurred while submitting your form. Please try again or file a bug report. Close

  1. Blog
  2. Article

Tom Haddon
on 15 November 2017

Codetree Collect Info


We recently landed a feature in Codetree that I’m pretty excited about. Codetree is a tool for collecting code from various locations and assembling it in a specific directory structure. It can be used in a standalone fashion, but is also tightly integrated with Mojo, which we use to deploy and manage Juju models.

To install Codetree, just run “sudo snap install codetree --classic“.

As an example, let’s say you want to get a subset of the latest OpenStack charms. You could do so by creating a file called openstack-charms with the following contents:

charms                   @
charms/cinder            git+https://github.com/openstack/charm-cinder;revno=stable/17.08
charms/glance            git+https://github.com/openstack/charm-glance;revno=stable/17.08
charms/hacluster         git+https://github.com/openstack/charm-hacluster;revno=stable/17.08
charms/heat              git+https://github.com/openstack/charm-heat;revno=stable/17.08
charms/keystone          git+https://github.com/openstack/charm-keystone;revno=stable/17.08

You’d then run “codetree openstack-charms” and you’d have the charms assembled in a “charms” directory. So far so good.

But what happens three months from now when you want to upgrade the charms you’ve deployed to a newer version? Or two months from now if you come across a bug in the charms and are not sure exactly what version you deployed and when? Juju strips out any dotfiles and dot directories from charms it deploys, so you won’t be able to use git commands to query where the charms on disk in your deployed OpenStack came from.

This is where the new feature we’ve just added to Codetree comes in. Codetree will now inject a file called “codetree-collect-info.yaml” into any directory it downloads, and this file will then be queryable later to confirm what version you deployed. This can be done in situ on your deployed OpenStack instance. For example:

juju ssh keystone/0 “head -4 /var/lib/juju/agents/unit*/charm/codetree-collect-info.yaml”
collect_date: '2017-11-01 14:32:55.815503'
collect_url: git+https://github.com/openstack/charm-keystone;revno=91490b7daf7511a717f75f62b57fc3f97cc6d740
hashes:
  LICENSE: cfc7749b96f63bd31c3c42b5c471bf756814053e847c10f3eb003417bc523d30

Now you can easily see the specific revision the charm was collected from, when it was collected, and the hashes allow you to query if any of the files on disk have been changed.

Our next planned step from here is to add a “charm-report” phase to Mojo to allow us to query this information with one simple command.

Related posts


Canonical
5 January 2026

Canonical announces Ubuntu support for the NVIDIA Rubin platform

Canonical announcements Article

Official Ubuntu support for the NVIDIA Rubin platform, including the NVIDIA Vera Rubin NVL72 rack-scale systems, announced at CES 2026 CES 2026, Las Vegas. – Canonical, the publisher of Ubuntu, is pleased to announce official support for the NVIDIA Rubin platform and the latest distributions of the new NVIDIA Nemotron 3 open models.  As A ...


Canonical
5 January 2026

Meet Canonical at CES 2026: A trusted foundation for your device lifecycle

Canonical announcements Events

This year, the Canonical team will be at Booth #10562 in the North Hall, we will be demonstrating how Ubuntu Core, Ubuntu Pro for Devices, and our partner ecosystem help you accelerate development, ensure reliability, and simplify the lifecycle of your connected fleets. ...


David Beamonte
19 December 2025

A better way to provision NVIDIA BlueField DPUs at scale with MAAS

Cloud and server MAAS

The recent release of MAAS 3.7 introduces a significant new capability: the ability to provision NVIDIA BlueField Data Processing Units (DPUs) directly through their Baseboard Management Controller (BMC), ...