Cloudwash

Introduction

cloudwash is a library written in python that can be used to monitor and remove the unused cloud resources in public and private cloud providers.

Most importantly, cloudwash provides the CLI utility called 'swach' that can be used to fire commands to cleanup the resources on cloud.

cloudwash supports following cloud providers:

• RedHat Enterprize Virtualization Manager - RHEV (Support yet To be added)
• RedHat Openstack (Support yet To be added)
• OCP Clusters deplyed on Public clouds (Support yet To be added)

And the Containerization tools:

• Podman

NOTE: You can use --all flag with all the cloud providers to clean all supported resources.

The list of resource types it helps to clean could be found under settings.yaml.template](https://github.com/RedHatQE/cloudwash/blob/master/settings.yaml.template) file for individual cloud providers along with cleanup criteria.

Installation

User can run cloudwash multiple ways:

• PiP Package Installation
• Docker Image Installation
• OC BuildConfig Installation

PiP Package Installation

For Linux Users, Depending on the distribution you are using, you may need to install following packages (or similar for your distribution of linux):

• libcurl-devel
• openssl-devel
• libxml2-devel
• libxml2-static
• gcc

Pycurl is a one of the dependent package of cloudwash that wants you to install above dependencies. Read more about it http://pycurl.io/docs/latest/install.html

Installation:

$ mkdir ~/cloudwash && cd ~/cloudwash
$ pip install cloudwash

Docker Image Installation

From Container image registry

The container image for cloudwash is available in quay. This image provides the cloudwash installed from released python package with release version tags. Latest tag always points to the latest released version tag.

Build from local DockerFile

This github repo contains a DockerFile, use any container building service to build from the dockerfile:

Build container from Dockerfile.dev that should build a container from the cloudwash github master branch giving the access to pre-released features.

OC BuildConfig Installation

This github repo provides the ready to use BuildConfig on OCP / Kubernetes. The build config should create buildconfig to build master branch based container image. Use the image to build cloudwash pod.

Configuration

The cloudwash uses the DynaConf configuration python module to access the data in settings.yaml or conf directory settings, it also allows an unique way of declaring secrets via Environment variables instead of putting in plain settings.yaml.

e.g: The Azure password field can be set via environment variable by exporting the environment variable

# export CLEANUP_PROVIDERS__AZURE__PASSWORD = myPa$$worb"

Configuration with PyPi package:

Copy/Download settings.yaml.template to local ~/cloudwash directory as settings.yaml, update it with the cloud provider credentials and other configuration details for successful resource reporting and cleanup.

Configuration with cloudwash container images:

Either - The docker images have settings.yaml added from Dockerfile. Build the container from the image, access the container and update the settings.yaml with real values and commit the changes to the image. Use the commited image for cleanup activity.

Or - Export/Set the environment variables for all or only sensitive credentials as shown above. The dynaconf in cloudwash container should read these credentials from environment variable.

Usage Examples

• Cleanup Help:

# swach --help

Usage: swach [OPTIONS] COMMAND [ARGS]...

A Cleanup Utility to remove cloud resources from cloud Providers!

Options:
-d, --dry Only show what will be removed from Providers!
--help Show this message and exit.

Commands:
azure		Cleanup Azure provider
aws			Cleanup Amazon provider
gce			Cleanup GCE provider
openstack	Cleanup OSP provider
podman      Cleanup Podman provider
rhev 		Cleanup RHEV provider
vmware 		Cleanup VMWare provider

• Cleanup Cloud Provider help:

# swach azure --help

Usage: swach azure [OPTIONS]

  Cleanup Azure provider

Options:
  --all             Remove all unused Resources from the provider
  --all_rg          Remove resource group only if all resources are older than SLA
  --nics            Remove only unused NICs from the provider
  --discs           Remove only unused DISCs from the provider
  --vms             Remove only unused VMs from the provider
  --pips            Remove only PiPs from the provider
  --help            Show this message and exit.

• Cleanup Dry Run (Monitor only mode using option -d):

# swach -d azure --all

<<<<<<< Running the cleanup script in DRY RUN mode >>>>>>>
The AZURE providers settings are initialized and validated !

=========== DRY SUMMARY ============

VMs:
	Deletable: ['test-bvhoduliam']
	Stoppable: ['foremanqe-nightly2']
DISCs:
	Deletable: ['test-bvhoduliam-osdisk']
NICs:
	Deletable: ['test-axodawttrw-nic0']
PIPs:
	Deletable: ['test-axodawttrw-pip0']
====================================

• Actual Cleanup Run:

# swach azure --all

<<<<<<< Running the cleanup script in ACTION mode >>>>>>>
The AZURE providers settings are initialized and validated !

Stopped [] and removed ['test-bvhoduliam'] VMs from Azure Cloud.
Removed following and all unused nics from Azure Cloud.
['test-axodawttrw-nic0']
Removed following and all unused discs from Azure Cloud.
['test-bvhoduliam-osdisk']
Removed following and all unused pips from Azure Cloud.
['test-axodawttrw-pip0']

How to run the cloudwash setup locally using Minikube