How I can inject artifact from AWS S3 inside Docker image? - amazon-web-services

I need to prepare Docker image with embedded Jar file to push it into ECR. Jar file is storing in S3 bucket. How I can inject jar inside image without explicit storing AWS access keys into image?
Maybe I can use AWS CLI or exist other way?
Also it is not recommended to add public access to my s3 bucket and set access keys via env variable during execute docker run.

You can define an AWS IAM Role and attach it to EC2 Instances. So any instance that needs to run this docker build command, can do so as long as it has the IAM role attached to it. You can do so from the AWS Console. This solves the problem of you putting AWS credentials on the instance itself.
You will still need to install the aws cli in your Dockerfile. Once IAM Role is attached, you don't have to worry about credentials.
Recommended docs:
IAM Roles for Amazon EC2
Here's an official blog post tutorial on how to do this:
Attach an AWS IAM Role to an Existing Amazon EC2 Instance by Using the AWS CLI
Just make sure you specify in the IAM Role which S3 Buckets you want these instances to have access to.

Related

AWS - ECS load S3 files in entrypoint script

Hi all!
Code: (entrypoint.sh)
printenv
CREDENTIALS=$(curl -s "http://169.254.170.2$AWS_CONTAINER_CREDENTIALS_RELATIVE_URI")
ACCESS_KEY_ID=$(echo "$CREDENTIALS" | jq .AccessKeyId)
SECRET_ACCESS_KEY=$(echo "$CREDENTIALS" | jq .SecretAccessKey)
TOKEN=$(echo "$CREDENTIALS" | jq .Token)
export AWS_ACCESS_KEY_ID=$ACCESS_KEY_ID
export AWS_SECRET_ACCESS_KEY=$SECRET_ACCESS_KEY
export AWS_SESSION_TOKEN=$TOKEN
aws s3 cp s3://BUCKET/file.txt /PATH/file.txt
Problem:
I'm trying to fetch AWS S3 files to ECS inspired by:
AWS Documentation
(But I'm fetching from S3 directly, not throught VPC endpoint)
I have configured bucket policy & role policy (that is passed in taskDefinition as taskRoleArn & executionRoleArn)
Locally when I'm fetching with aws cli and passing temporary credentials (that I logged in ECS with printenv command in entrypoint script) everything works fine. I can save files on my pc.
On ECS I have error:
fatal error: An error occurred (403) when calling the HeadObject operation: Forbidden
Where can I find solution? Someone had similar problem?
Frist thing, If you are working inside AWS, It strongly recommended to use AWS ECS service role or ECS task role or EC2 role. you do need to fetch credentials from metadata.
But seems like the current role does have permission to s3 or the entrypoint not exporting properly the Environment variable.
If your container instance has already assing role then do not need to export Accesskey just call the aws s3 cp s3://BUCKET/file.txt /PATH/file.txt and it should work.
IAM Roles for Tasks
With IAM roles for Amazon ECS tasks, you can specify an IAM role that
can be used by the containers in a task. Applications must sign their
AWS API requests with AWS credentials, and this feature provides a
strategy for managing credentials for your applications to use,
similar to the way that Amazon EC2 instance profiles provide
credentials to EC2 instances. Instead of creating and distributing
your AWS credentials to the containers or using the EC2 instance’s
role, you can associate an IAM role with an ECS task definition or
RunTask API operation.
So the when you assign role to ECS task or ECS service your entrypoint will be that simple.
printenv
aws s3 cp s3://BUCKET/file.txt /PATH/file.txt
Also, your export will not work as you are expecting, the best way to pass ENV to container form task definition, export will not in this case.
I will suggest assigning role to ECS task and it should work as you are expecting.

"no basic auth credentials" when trying to pull an image from a private ECR

I have the following line somewhere in the middle of my Dockerfile to retrieve an image from my private ECR.
FROM **********.dkr.ecr.ap-southeast-1.amazonaws.com/prod/*************:ff03401
This is the error that I get in AWS Codebuild when trying to build this:
Step 21/36 : FROM **********.dkr.ecr.ap-southeast-1.amazonaws.com/prod/*************:ff03401
Get https://**********.dkr.ecr.ap-southeast-1.amazonaws.com/prod/*************/manifests/ff03401: no basic auth credentials
How can one provide these credentials in the most secure way, and in a way that can also be terraformed?
There are multiple ways to do it.
Using aws access and secret key. In which you set the aws credentials on the ec2 machine and run ecr login command. aws ecr get-login --no-include-email --registry-ids <some-id> --region eu-west-1 and then docker pull should work. But this is not a recommended secure way.
What I prefer is using aws iam roles.
Assuming you want to pull this image on your ec2 machine that was brought up using terraform. Make use of iam roles.
Create an iam role manually or using terraform iam resource.
For contents of iam policy refer this.
While bringing ec2 using terraform instance resource make use of iam_instance_profile attribute, the value of this attribute should be the name of iam role you created.
This should be enough to automatically pull docker images from ECR in a secure way.
Hope this helps.

User Data script to call aws cli

I am trying to get some files from S3 on startup in an EC2 instance by using a User Data script and the command
/usr/bin/aws s3 cp ...
The log tells me that permission was denied and I believe it is due to aws cli not finding any credentials when executing the user data script.
Running the command with sudo after the instance has started works fine.
I have run aws configure both with sudo and without.
I do not want to use cronjob to run something on startup since I am working with an AMI and often need to change the script, therefore it is more convenient for me to change the user data instead of creating a new AMI everytime the script changes.
If possible, I would also like to avoid writing the credentials into the script.
How can I configure awscli in such a way that the credentials are used when running a user data script?
I suggest you remove the AWS credentials from the instance/AMI. Your userdata script will be supplied with temporary credentials when needed by the AWS metadata server.
See: IAM Roles for Amazon EC2
Clear/delete AWS credentials configurations from your instance and create an AMI
Create a policy that has the minimum privileges to run your script
Create a IAM role and attach the policy you just created
Attach the IAM role when you launch the instance (very important)
Have your userdata script call /usr/bin/aws s3 cp ... without supplying credentials explicitly or using credentials file
You can configure your EC2 instance to receive a pre-defined IAM Role that has its credentials "baked-in" to the instance that it fetches upon instantiation, which it turn will allow it to call aws-cli commands in your User Data script without the need to configure credentials at all.
Here's more info on IAM Roles for EC2:
http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/iam-roles-for-amazon-ec2.html
It's worth noting here that you'll need to attach the appropriate policies to the IAM Role that you assign to your instance in order for the aws-cli commands to succeed. More information on that can be found here:
http://docs.aws.amazon.com/AWSEC2/latest/UserGuide/iam-roles-for-amazon-ec2.html#working-with-iam-roles

Authenticating to S3 from a cloud-init user-data script

Is there a way to configure an S3 bucket/folder/file so that it is not publicly accessible, yet a user-data script can authenticate and obtain access to items in S3?
You should be assigning an IAM Role to your EC2 instance that has access to the S3 bucket, and any other AWS resources that the EC2 server needs to access. Then when you use the AWS CLI tool or the AWS SDK it will automatically use the IAM role assigned to the instance.

Dockerrun.aws.json structure for ECR Repo

We are switching from Docker Hub to ECR and I'm curious how to structure the Dockerrun.aws.json file to use this image. I attempted to modify the name as <my_ECR_URL>/<repo_name>:<image_tag> but this is not successful. I also saw the details of private registries using an authentication file on S3 but this doesn't seem like the correct route when aws ecr get-login is the recommended way to authenticate with ECR.
Can anyone point me to how I can use an ECR image in a Beanstalk Dockerrun.aws.json file?
If I look at the ECS Task Definition,there's a required attribute called com.amazonaws.ecs.capability.ecr-auth, but I'm not setting that anywhere in my Dockerrun.aws.json file and I'm not sure what needs to be there. Perhaps it is an S3 bucket? Something is needed as every time I try to run the Elastic Beanstalk created tasks from ECS, I get:
Run tasks failed
Reasons : ATTRIBUTE
Any insights are greatly appreciated.
Update I see from some other threads that this used to occur with earlier versions of the ECS agent but I am currently running Agent version 1.6.0 and Docker version 1.7.1, which I believe are the recommended versions. Is this possibly an issue with the Docker version?
So it turns out, the ECS agent was only able to pull images with version 1.7, and that's where mine was falling. Updating the agent resolves my issue, and hopefully it helps someone else.
This is most likely an issue with IAM roles if you are using a role that was previously created for Elastic Beanstalk. Ensure that the role that Elastic Beanstalk is running with has the AmazonEC2ContainerRegistryReadOnly managed policy attached
Source: http://docs.aws.amazon.com/AmazonECR/latest/userguide/ECR_IAM_policies.html
Support for ECR was added in version 1.7.0 of the ECS Agent.
When using Elasticbeanstalk and ECR you don't need to authenticate. Just make sure the user has the policy AmazonEC2ContainerRegistryReadOnly
You can store your custom Docker images in AWS with Amazon EC2 Container Registry (Amazon ECR). When you store your Docker images in
Amazon ECR, Elastic Beanstalk automatically authenticates to the
Amazon ECR registry with your environment's instance profile, so you
don't need to generate an authentication file and upload it to Amazon
Simple Storage Service (Amazon S3).
You do, however, need to provide your instances with permission to
access the images in your Amazon ECR repository by adding permissions
to your environment's instance profile. You can attach the
AmazonEC2ContainerRegistryReadOnly managed policy to the instance
profile to provide read-only access to all Amazon ECR repositories in
your account, or grant access to single repository by using the
following template to create a custom policy:
Source: http://docs.aws.amazon.com/elasticbeanstalk/latest/dg/create_deploy_docker.container.console.html