octo_base

Puppet module providing common functionality for Octopus Energy machines.

Making changes

After any change made to the base manifest, there's three things that need to be done to push the change to production.

0. The change needs to be committed and pushed to the octo_base repo.

1. The version number in the metadata.json file needs to be incremented.

2. The octo_base repository needs to have a new release created. Once your change has been merged, including the version number change, you can create a release here.

   The tag should match the version you specified in the metadata.json file, including the v., so if you incremented the version to v1.38 you need to tag it as v1.38. The same applies to the release title. The assets are generated automatically, so no need to zip / upload anything.

3. In the kraken-core repo, you need to change the version for the octoenergy-octo_base module to match the version specified above. There's two places this change is made: In the Puppetfile, only the version line needs to change. In the Puppetfile.lock, the version number and the SHA need to change. You can get the SHA of the release on the releases page of the octo_base repo.

Changelog

v1.46

• Add option to include/exclude ssm-agent

v1.45

• Update unnecessary package list to not remove snapd on newer ubuntu images

v1.44

• Add puppet tooling
• $vagrant must be set to 1 to work

v1.42

• Fix pip install of awscli

v1.41

• Update awscli from v1.16.119 to v1.25.46
• Update libssl from v1.1 to v3

v1.40

• Ensure AWS Inspector installs on Ubuntu 22.04

v1.39

• Fix broken CW agent config

v1.38

• Switch CW agent to run as root

v1.37

• Fixed missing comma

v1.36

• Change CW Agent logfile location
• Change dpkg.log to apt logfile
• Add omit hostname and aggregate based on InstanceName

v1.35

• Fix namespace and CW dimensions

v1.34

• Remove awslogs and replace it with CW Agent

v1.33

• Add another DiskUsedPercentage Cloudwatch metric with new dimensions

v1.32

• Ensure AWS Inspector installs on Ubuntu 20.04

v1.31

• Allow AWS Inspector to be optional

v1.30

• Fix bug with how dimensions are associated with metrics

v1.29

• Forward system log files to Cloudwatch

v1.28

• Batch Cloudwatch PutMetricData to a single AWSCLI call for collect-system-metrics.sh

v1.27

• Disable AWS Inspector in Vagrant builds

v1.26

• Set default umask to 027

v1.25

• Actually fix CIS hardening problem with /var/log/.

v1.23, v1.24

• Attempts to fix CIS hardening problem with /var/log/.

v1.22

• Exclude /etc/cron.d/ from CIS permission tightening

v1.21

• Fix for sysctl not loading late enough in the boot sequence

v1.20

• CIS Level 1 Server

v1.19

• Fix broken AWS inspector key

v1.18

• CIS hardening

v1.17

• Add SSM agent.

v1.16

• Add AWS inspector.

v1.15

• Add region variable to accommodate deployments in other regions.

v1.14

• Ensure AWS logs nanny is not installed unnecessarily.

v1.13

• Ensure AWS logs config updates correctly.

v1.12

• Correct memory calculations for Ubuntu 18.04 in metrics script

v1.11

• Remove changes from 1.10 (found a better way to solve same problem)
• Use Python 3 for awscli

v1.10

• Purge unattended updates at start (to allow other apt commands to not get locked out).

v1.9.2

• Use lower min versions in dependency version ranges

v1.9.1

• Use version ranges in dependencies

v1.9

• Fix (another) bug in metrics script around disk usage
• Install packages required for monitoring script in Ubuntu 16:04 and later

v1.8

• Fix bug in metrics script around disk usage

v1.7

• Fix issue with apt-get upgrade returning a exit code of 100

v1.6

• Use --fix-missing when calling apt-get upgrade.

v1.5

• Call autoremove after update and upgrade
• Make upgrade call more robust using a timeout and retrying

v1.4

• Extend system monitoring script to work in Ubuntu 16.04

v1.3

• Add locking to the system metrics file

v1.2

• AWSCLI version bump
• Cloudwatch logs integration

v1.1

• AWSCLI version bump
• Unattended upgrades installed by default.

v1.0

Initial version