As of version v0.29.0
cloud-nuke sends telemetry back to Gruntwork to help us better prioritize bug fixes and feature
improvements. The following metrics are included:
- Command and Arguments
- Version Number
- Timestamps
- Resource Types
- Resource Counts
- A randomly generated Run ID
- AWS Account ID
We never collect
- IP Addresses
- Resource Names
Telemetry can be disabled entirely by setting the DISABLE_TELEMETRY
environment variable on the command line.
As an open source tool, you can see the exact statistics being collected by searching the code for
telemetry.TrackEvent(...)
This repo contains a CLI tool to delete all resources . cloud-nuke was created for situations when you might have an account you use for testing and need to clean up leftover resources so you're not charged for them. Also great for cleaning out accounts with redundant resources. Also great for removing unnecessary defaults like default VPCs and permissive ingress/egress rules in default security groups.
In addition, cloud-nuke offers non-destructive inspecting functionality that can either be called via the command-line interface, or consumed as library methods, for scripting purposes.
The currently supported functionality includes:
Cloud-nuke suppports ๐ inspecting and ๐ฅ๐ deleting the following AWS resources:
Resource Family | Resource type |
---|---|
EC2 | Auto scaling groups |
EC2 | Elastic Load Balancers (v1 and v2) |
EC2 | EBS Volumes |
EC2 | Unprotected EC2 instances |
EC2 | AMIS |
EC2 | Snapshots |
EC2 | Elastic IPs |
EC2 | Launch Configurations |
Certificate Manager | ACM Private CA |
Direct Connect | Transit Gateways |
Elasticache | Clusters |
Elasticache | Parameter Groups |
Elasticache | Subnet Groups |
ECS | Services |
ECS | Clusters |
EKS | Clusters |
RDS | RDS databases |
RDS | Neptune |
RDS | Document DB instances |
DynamoDB | Tables |
Lambda | Functions |
SQS | Queues |
S3 | Buckets |
VPC | Default VPCs |
VPC | Default rules in the un-deletable default security group |
VPC | NAT Gateways |
IAM | Users |
IAM | Roles (and any associated EC2 instance profiles) |
IAM | Service-linked-roles |
IAM | Groups |
IAM | Policies |
IAM | Customer-managed policies |
IAM | Access analyzers |
IAM | OpenID Connect providers |
Secrets Manager | Secrets |
CloudWatch | Dashboard |
CloudWatch | Log groups |
CloudWatch | Alarms |
OpenSearch | Domains |
KMS | Custgomer managed keys (and associated key aliases) |
GuardDuty | Detectors |
Macie | Member accounts |
SageMaker | Notebook instances |
Kinesis | Streams |
API Gateway | Gateways (v1 and v2) |
EFS | File systems |
SNS | Topics |
CloudTrail | Trails |
ECR | Repositories |
Config | Service recorders |
Config | Service rules |
Security Hub | Hubs |
Security Hub | Members |
Security Hub | Administrators |
WARNING: The RDS APIs also interact with neptune and document db resources. Running
cloud-nuke aws --resource-type rds
without a config file will remove any neptune and document db resources in the account.
NOTE: AWS Backup Resource: Resources (such as AMIs) created by AWS Backup, while owned by your AWS account, are managed specifically by AWS Backup and cannot be deleted through standard APIs calls for that resource. These resources are tagged by AWS Backup and are filtered out so that
cloud-nuke
does not fail when trying to delete resources it cannot delete.
When executed as cloud-nuke aws
, this tool is HIGHLY DESTRUCTIVE and deletes all resources! This mode should never be used in a production environment!
When executed as cloud-nuke defaults-aws
, this tool deletes all DEFAULT VPCs and the default ingress/egress rule for all default security groups. This should be used in production environments WITH CAUTION.
- Download the latest binary for your OS on the releases page.
- Move the binary to a folder on your
PATH
. E.g.:mv cloud-nuke_darwin_amd64 /usr/local/bin/cloud-nuke
. - Add execute permissions to the binary. E.g.:
chmod u+x /usr/local/bin/cloud-nuke
. - Test it installed correctly:
cloud-nuke --help
.
Note that package managers are third party. The third party cloud-nuke packages may not be updated with the latest version, but are often close. Please check your version against the latest available on the releases page. If you want the latest version, the recommended installation option is to download from the releases page.
-
macOS: You can install cloud-nuke using Homebrew:
brew install cloud-nuke
. -
Linux: Most Linux users can use Homebrew:
brew install cloud-nuke
.
Simply running cloud-nuke aws
will start the process of cleaning up your cloud account. You'll be shown a list of resources that'll be deleted as well as a prompt to confirm before any deletion actually takes place.
In AWS, to delete only the default resources, run cloud-nuke defaults-aws
. This will remove the default VPCs in each region, and will also revoke the ingress and egress rules associated with the default security group in each VPC. Note that the default security group itself is unable to be deleted.
When using cloud-nuke aws
, or cloud-nuke inspect-aws
, you can pass in the AWS_PROFILE
env variable to target resources in certain regions for a specific AWS account. For example the following command will nuke resources only in ap-south-1
and ap-south-2
regions in the gruntwork-dev
AWS account:
AWS_PROFILE=gruntwork-dev cloud-nuke aws --region ap-south-1 --region ap-south-2
Similarly, the following command will inspect resources only in us-east-1
AWS_PROFILE=gruntwork-dev cloud-nuke inspect-aws --region us-east-1
When using cloud-nuke aws
, or cloud-nuke inspect-aws
, you can use the --region
flag to target resources in certain regions. For example the following command will nuke resources only in ap-south-1
and ap-south-2
regions:
cloud-nuke aws --region ap-south-1 --region ap-south-2
Similarly, the following command will inspect resources only in us-east-1
cloud-nuke inspect-aws --region us-east-1
Including regions is available within:
cloud-nuke aws
cloud-nuke defaults-aws
cloud-nuke inspect-aws
When using cloud-nuke aws
or cloud-nuke inspect-aws
, you can use the --exclude-region
flag to exclude resources in certain regions from being deleted or inspected. For example the following command does not nuke resources in ap-south-1
and ap-south-2
regions:
cloud-nuke aws --exclude-region ap-south-1 --exclude-region ap-south-2
Similarly, the following command will not inspect resources in the us-west-1
region:
cloud-nuke inspect-aws --exclude-region us-west-1
--region
and --exclude-region
flags cannot be specified together i.e. they are mutually exclusive.
Excluding regions is available within:
cloud-nuke aws
cloud-nuke defaults-aws
cloud-nuke inspect-aws
You can use the --older-than
flag to only nuke resources that were created before a certain period, the possible values are all valid values for ParseDuration For example the following command nukes resources that are at least one day old:
cloud-nuke aws --older-than 24h
Excluding resources by age is available within:
cloud-nuke aws
cloud-nuke inspect-aws
You can use the --list-resource-types
flag to list resource types whose termination is currently supported:
cloud-nuke aws --list-resource-types
Listing supported resource types is available within:
cloud-nuke aws
cloud-nuke inspect-aws
If you want to target specific resource types (e.g ec2, ami, etc.) instead of all the supported resources you can
do so by specifying them through the --resource-type
flag:
cloud-nuke aws --resource-type ec2 --resource-type ami
will search and target only ec2
and ami
resources. The specified resource type should be a valid resource type
i.e. it should be present in the --list-resource-types
output. Using --resource-type
also speeds up search because
we are searching only for specific resource types.
Similarly, the following command will inspect only ec2 instances:
cloud-nuke inspect-aws --resource-type ec2
Specifying target resource types is available within:
cloud-nuke aws
cloud-nuke inspect-aws
Just like you can select which resources to terminate using --resource-type
, you can select which resources to skip using
--exclude-resource-type
flag:
cloud-nuke aws --exclude-resource-type s3 --exclude-resource-type ec2
This will terminate all resource types other than S3 and EC2.
--resource-type
and --exclude-resource-type
flags cannot be specified together i.e. they are mutually exclusive.
Specifying resource types to exclude is available within:
cloud-nuke aws
cloud-nuke inspect-aws
If you want to check what resources are going to be targeted without actually terminating them, you can use the
--dry-run
flag
cloud-nuke aws --resource-type ec2 --dry-run
Dry run mode is only available within:
cloud-nuke aws
You can import cloud-nuke into other projects and use it as a library for programmatically inspecting and counting resources.
package main
import (
"fmt"
"time"
"github.com/aws/aws-sdk-go/aws"
nuke_aws "github.com/gruntwork-io/cloud-nuke/aws"
"github.com/gruntwork-io/cloud-nuke/externalcreds"
)
func main() {
// You can scan multiple regions at once, or just pass a single region for speed
targetRegions := []string{"us-east-1", "us-west-1", "us-west-2"}
excludeRegions := []string{}
// You can simultaneously target multiple resource types as well
resourceTypes := []string{"ec2", "vpc"}
excludeResourceTypes := []string{}
// excludeAfter is parsed identically to the --older-than flag
excludeAfter := time.Now()
// Any custom settings you want
myCustomConfig := &aws.Config{}
myCustomConfig.WithMaxRetries(3)
myCustomConfig.WithLogLevel(aws.LogDebugWithRequestErrors)
// Optionally, set custom credentials
// myCustomConfig.WithCredentials()
// Be sure to set your config prior to calling any library methods such as NewQuery
externalcreds.Set(myCustomConfig)
// NewQuery is a convenience method for configuring parameters you want to pass to your resource search
query, err := nuke_aws.NewQuery(
targetRegions,
excludeRegions,
resourceTypes,
excludeResourceTypes,
excludeAfter,
)
if err != nil {
fmt.Println(err)
}
// InspectResources still returns *AwsAccountResources, but this struct has been extended with several
// convenience methods for quickly determining if resources exist in a given region
accountResources, err := nuke_aws.InspectResources(query)
if err != nil {
fmt.Println(err)
}
// You can call GetRegion to examine a single region's resources
usWest1Resources := accountResources.GetRegion("us-west-1")
// Then interrogate them with the new methods:
// Count the number of any resource type within the region
countOfEc2InUsWest1 := usWest1Resources.CountOfResourceType("ec2")
fmt.Printf("countOfEc2InUsWest1: %d\n", countOfEc2InUsWest1)
// countOfEc2InUsWest1: 2
fmt.Printf("usWest1Resources.ResourceTypePresent(\"ec2\"):%b\n", usWest1Resources.ResourceTypePresent("ec2"))
// usWest1Resources.ResourceTypePresent("ec2"): true
// Get all the resource identifiers for a given resource type
// In this example, we're only looking for ec2 instances
resourceIds := usWest1Resources.IdentifiersForResourceType("ec2")
fmt.Printf("resourceIds: %s", resourceIds)
// resourceIds: [i-0c5d16c3ef28dda24 i-09d9739e1f4d27814]
}
For more granularity, such as being able to specify which resources to terminate using regular expressions or plain text, you can pass in a configuration file.
Note: Config file support is a new feature and only filtering a handful of resources by name using regular expressions is currently supported. We'll be adding more support in the future, and pull requests are welcome!
The following resources support the Config file:
- S3 Buckets
- Resource type:
s3
- Config key:
s3
- Resource type:
- IAM Users
- Resource type:
iam
- Config key:
IAMUsers
- Resource type:
- IAM Roles
- Resource type:
iam-role
- Config key:
IAMRoles
- Resource type:
- IAM Service-Linked Roles
- Resource type:
iam-service-linked-role
- Config key:
IAMServiceLinkedRoles
- Resource type:
- Secrets Manager Secrets
- Resource type:
secretsmanager
- Config key:
SecretsManager
- Resource type:
- NAT Gateways
- Resource type:
nat-gateway
- Config key:
NatGateway
- Resource type:
- IAM Access Analyzers
- Resource type:
accessanalyzer
- Config key:
AccessAnalyzer
- Resource type:
- CloudWatch Dashboards
- Resource type:
cloudwatch-dashboard
- Config key:
CloudWatchDashboard
- Resource type:
- OpenSearch Domains
- Resource type:
opensearch
- Config key:
OpenSearchDomain
- Resource type:
- DynamoDB Tables
- Resource type:
dynamodb
- Config key:
DynamoDB
- Resource type:
- EBS Volumes
- Resource type:
ebs
- Config key:
EBSVolume
- Resource type:
- Lambda Functions
- Resource type:
lambda
- Config key:
LambdaFunction
- Resource type:
- Elastic Load Balancers
- Resource type:
elbv2
- Config key:
ELBv2
- Resource type:
- ECS Services
- Resource type:
ecsserv
- Config key:
ECSService
- Resource type:
- ECS Clusters
- Resource type:
ecscluster
- Config key:
ECSCluster
- Resource type:
- Elasticache
- Resource type:
elasticache
- Config key:
Elasticache
- Resource type:
- Elasticache Parameter Groups
- Resource type:
elasticacheParameterGroup
- Config key:
ElasticacheParameterGroup
- Resource type:
- Elasticache Subnet Group
- Resource type:
elasticacheSubnetGroup
- Config key:
ElasticacheSubnetGroup
- Resource type:
- VPCs
- Resource type:
vpc
- Config key:
VPC
- Resource type:
- IAM OpenID Connect Providers
- Resource type:
oidcprovider
- Config key:
OIDCProvider
- Resource type:
- CloudWatch LogGroups
- Resource type:
cloudwatch-loggroup
- Config key:
CloudWatchLogGroup
- Resource type:
- KMS customer keys
- Resource type:
kmscustomerkeys
- Config key:
KMSCustomerKeys
- Resource type:
- Auto Scaling Groups
- Resource type:
asg
- Config key:
AutoScalingGroup
- Resource type:
- Launch Configurations
- Resource type:
lc
- Config key:
LaunchConfiguration
- Resource type:
- Elastic IP Address
- Resource type:
eip
- Config key:
ElasticIP
- Resource type:
- EC2 Instances
- Resource type:
ec2
- Config key:
EC2
- Resource type:
- EC2 Dedicated Hosts
- Resource type:
ec2-dedicated-hosts
- Config key:
EC2DedicatedHosts
- Resource type:
- EC2 Key Pairs
- Resource type:
ec2-keypairs
- Config key:
EC2KeyPairs
- Resource type:
- EKS Clusters
- Resource type:
ekscluster
- Config key:
EKSCluster
- Resource type:
- SageMaker Notebook Instances
- Resource type:
sagemaker-notebook-instances
- Config key:
SageMakerNotebook
- Resource type:
- API Gateways (v1)
- Resource type:
apigateway
- Config key:
APIGateway
- Resource type:
- API Gateways (v2)
- Resource type:
apigatewayv2
- Config key:
APIGatewayV2
- Resource type:
- Elastic FileSystems (efs)
- Resource type:
efs
- Config key:
ElasticFileSystem
- Resource type:
- ECR Repositories
- Resource type:
ecr
- Config key:
ECRRepository
- Resource type:
- RDS, Neptune, and Document DB Resources
- Resource type:
rds
- Config key:
DBInstances
- Resource type:
- Launch Templates
- Resource type:
lt
- Config key:
LaunchTemplate
- Resource type:
- CloudWatch Alarms
- Resource type:
cloudwatch-alarm
- Config key:
CloudWatchAlarm
- Resource type:
- Redshift
- Resource type:
redshift
- Config key:
Redshift
- Resource type:
Notes:
- no configuration options for KMS customer keys, since keys are created with auto-generated identifier
- Kinesis Streams
- Resource type:
kinesis-stream
- Config key:
KinesisStream
- Resource type:
cloud-nuke aws --resource-type s3 --config path/to/file.yaml
Given this command, cloud-nuke
will nuke only S3 buckets, as specified by the --resource-type s3
option.
Now given the following config, the s3 buckets that will be nuked are further filtered to only include ones that match any of the provided regular expressions. So a bucket named alb-app-access-logs
would be deleted, but a bucket named my-s3-bucket
would not.
s3:
include:
names_regex:
- ^alb-.*-access-logs$
- .*-prod-alb-.*
Similarly, you can adjust the config to delete only IAM users of a particular name by using the IAMUsers
key. For
example, in the following config, only IAM users that have the prefix my-test-user-
in their username will be deleted.
IAMUsers:
include:
names_regex:
- ^my-test-user-.*
Now consider the following contrived example:
s3:
include:
names_regex:
- ^alb-.*-access-logs$
- .*-prod-alb-.*
exclude:
names_regex:
- public
- prod
The intention is to delete all the s3 buckets that match the include rules but not the exclude rules. Filtering is commutative, meaning that you should get the same result whether you apply the include filters before or after the exclude filters.
The result of these filters applied in either order will be a set of s3 buckets that match ^alb-.*-access-logs$
as long as they do not also contain public
or prod
. The rule to include s3 buckets matching .*-prod-alb-.*
is negated by the rule to exclude those matching prod
.
The options provided in the command line take precedence over those provided in any config file that gets passed in. For example, say you provide --resource-type s3
in the command line, along with a config file that specifies ec2:
at the top level but doesn't specify s3:
. The command line argument filters the resource types to include only s3, so the rules in the config file for ec2:
are ignored, and ec2 resources are not nuked. All s3 resources would be nuked.
In the same vein, say you do not provide a --resource-type
option in the command line, but you do pass in a config file that only lists rules for s3:
, such as cloud-nuke aws --config path/to/config.yaml
. In this case all resources would be nuked, but among s3
buckets, only those matching your config file rules would be nuked.
Be careful when nuking and append the --dry-run
option if you're unsure. Even without --dry-run
, cloud-nuke
will list resources that would undergo nuking and wait for your confirmation before carrying it out.
To find out what we options are supported in the config file today, consult this table. Resource types at the top level of the file that are supported are listed here.
resource type | names | names_regex | tags | tags_regex |
---|---|---|---|---|
s3 | none | โ | none | none |
iam user | none | โ | none | none |
ecsserv | none | โ | none | none |
ecscluster | none | โ | none | none |
secretsmanager | none | โ | none | none |
nat-gateway | none | โ | none | none |
accessanalyzer | none | โ | none | none |
dynamodb | none | โ | none | none |
ebs | none | โ | none | none |
lambda | none | โ | none | none |
elbv2 | none | โ | none | none |
ecs | none | โ | none | none |
elasticache | none | โ | none | none |
elasticache parameter group | none | โ | none | none |
elasticache subnet group | none | โ | none | none |
vpc | none | โ | none | none |
oidcprovider | none | โ | none | none |
cloudwatch-loggroup | none | โ | none | none |
kmscustomerkeys | none | โ | none | none |
asg | none | โ | none | none |
lc | none | โ | none | none |
eip | none | โ | none | none |
ec2 | none | โ | none | none |
apigateway | none | โ | none | none |
apigatewayv2 | none | โ | none | none |
eks | none | โ | none | none |
kinesis-stream | none | โ | none | none |
efs | none | โ | none | none |
acmpca | none | none | none | none |
iam role | none | โ | none | none |
iam service-linked role | none | โ | none | none |
iam policy | none | โ | none | none |
sagemaker-notebook-instances | none | โ | none | none |
ecr | none | โ | none | none |
rds (+neptune and documentdb) | none | โ | none | none |
lt | none | โ | none | none |
config-recorders | none | โ | none | none |
config-rules | none | โ | none | none |
cloudwatch-alarm | none | โ | none | none |
redshift | none | โ | none | none |
... (more to come) | none | none | none | none |
By default, cloud-nuke sends most output to the Debug
level logger, to enhance legibility, since the results of every deletion attempt will be displayed in the report that cloud-nuke prints after each run.
However, sometimes it's helpful to see all output, such as when you're debugging something.
You can set the log level by specifying the --log-level
flag as per logrus log levels:
cloud-nuke aws --log-level debug
OR
LOG_LEVEL=debug cloud-nuke aws
Default value is - info
. Acceptable values are debug, info, warn, error, panic, fatal, trace
as per logrus log level parser.
When deleting defaults with cloud-nuke defaults-aws
, use the --sg-only
flag to delete only the default
security group rules and not the default VPCs.
cloud-nuke defaults-aws --sg-only
When nuking VPCs cloud-nuke will attempt to remove dependency resources underneath the VPC
- Internet Gateways
- Egress Only Internet Gateways
- Elastic Network Interfaces
- VPC Endpoints
- Subnets
- Route Tables
- Network ACLs
- Security Groups
- DHCP Option Sets (Will be dissociated from VPC, not deleted. Must be cleaned up separately)
- Elastic IPs (Supported as a separate resource that gets cleaned up first. If you are filtering what gets nuked, Elastic IPs may prevent VPCs from destroying.)
All other resources that get created within VPCs must be cleaned up prior to running cloud-nuke on VPC resources.
VPC resources may not be entirely cleaned up on the first run. We believe this is caused by an eventual consistency error in AWS.
If you see errors like
InvalidParameterValue: Network interface is currently in use.
We recommend waiting 30 minutes and trying again.
Happy Nuking!!!
In order for the cloud-nuke
CLI tool to access your AWS, you will need to provide your AWS credentials. You can use one of the standard AWS CLI credential mechanisms.
go test -v ./...
cloud-nuke is an open source project, and contributions from the community are very welcome! Please check out the Contribution Guidelines and Developing cloud-nuke for instructions.
To run cloud-nuke locally, use the go run
command:
go run main.go
Note: Many of the tests in the aws
folder run against a real AWS account and will create and destroy actual resources. DO NOT
hit CTRL+C
while the tests are running, as this will prevent them from cleaning up properly. We are not responsible for any
charges you may incur.
Before running the tests, you must configure your AWS credentials.
To run all the tests:
go test -v ./...
To run only the tests in a specific package, such as the package aws
:
cd aws
go test -v
And to run a specific test, such as TestListAMIs
in package aws
:
cd aws
go test -v -run TestListAMIs
Use env-vars to opt-in to special tests, which are expensive to run:
# Run acmpca tests
TEST_ACMPCA_EXPENSIVE_ENABLE=1 go test -v ./...
Every source file in this project should be formatted with go fmt
.
We try to follow the release process as deifned in our Coding Methodology.
If the new release contains any new resources that cloud-nuke
will support, mark it as a minor version bump (X in v0.X.Y)
to indicate backward incompatibilities.
This is because since version v0.2.0
cloud-nuke
has been configured to automatically include new resources (so you have
to explicitly opt-out). This is inherently not backward compatible, because users with CI practices around cloud-nuke
would
be surprised by new resources that are suddenly being picked up for deletion! This surprise is more alarming for resources
that are actively in use for any account, such as IAM Users.
Therefore please mark your release as backward incompatible and bump the minor version (X
in v0.X.Y
) when it includes
support for nuking new resources, so that we provide better signals for users when we introduce a new resource.
Go to the Releases Page and create a new release. The CircleCI job for this repo has been configured to:
- Automatically detect new tags.
- Build binaries for every OS using that tag as a version number.
- Upload the binaries to the release in GitHub.
See .circleci/config.yml
for details.
This code is released under the MIT License. See LICENSE.txt.