Skip to content

TriggerMail/gcp-hashicorp-packer-reaper

 
 

Repository files navigation

AWS Hashicorp Packer Reaper

Hashicorp Packer is a great tool for building Google Compute Engine Images. However, sometimes the virtual machine running packer is not stopped. This utility stops or terminated all virtual machines with the name tag 'Packer Builder'

You can use it as a command line utility or install it as an Cloud Run service and stop the spend , NoOps style!

install the packer reaper

to install the packer reaper, type:

pip install gcp-hashicorp-packer-reaper

show running packer instances

To show running packer instances:

$ gcp-hashicorp-packer-reaper list
packer-5e99d4f2-b5a5-e9e0-b763-cd5102ae7e73 launched 7 minutes ago in your-project - europe-west4-c - RUNNING
INFO: 1 packer builder instances found

stop running packer instances

To stop running packer instances older than 2 hours:

$ gcp-hashicorp-packer-reaper stop --older-than 2h

INFO: stopping packer-5e99d4f2-b5a5-e9e0-b763-cd5102ae7e73 in your-project created 3 hours ago
INFO: total of 1 running instances stopped

delete running packer instances

To terminate stopped and running packer instances older than 24 hours:

gcp-hashicorp-packer-reaper --verbose delete --older-than 24h

INFO: deleting packer-5e99d4f2-b5a5-e9e0-b763-cd5102ae7e73 in your-project created 2 days ago
INFO: total of 1 instances deleted

deploy the packer reaper

To deploy the packer reaper as a service in your project, type:

git clone https://github.com/binxio/gcp-hashicorp-packer-reaper.git
cd gcp-hashicorp-packer-reaper

PROJECT=$(gcloud config get-value project)
make USERNAME=$PROJECT snapshot

cd terraform
terraform init
terraform apply -var project=$PROJECT -auto-approve

This will install the packer reaper in your GCP project and run every 10 minutes, and delete instances older than 2 hours. You can change the schedule and the action to meet your requirements.

About

stop or terminate dangling packer instances on GCP

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 57.1%
  • Makefile 16.7%
  • HCL 13.7%
  • Shell 11.8%
  • Dockerfile 0.7%