A container designed to sync RPM repositories to a bucket, from where they can be made accessible to your offline EC2 instances.
The container is based on the official amazonlinux image.
This application is deployed via Terraform, an example deployment looks like this:
module "aws-reposync" {
source = "danielpodwysocki/reposync/aws"
version = "0.0.2"
# a list of subnet IDs where the sync can be executed
subnets = [aws_subnet.dev-public-c.id]
tags = {
"Name" = "aws-reposync"
"Env" = "production"
}
repos_path = "/Users/daniel/workspace/aws-reposync/example/repos/"
vpc_id = aws_vpc.dev.id
region = var.region
route_table_ids = [
aws_route_table.default.id,
]
cron = "* 8 * * ? *"
}
subnets must have internet access via an Internet gateway
cron defines how often an ECS task is launched to sync the repos.
repos_path is where the .repo files are stored on the machine you are running terraform apply from
route_table_ids is needed for adding the routes to the S3 gateway endpoint
This module will deploy an S3 gateway in your private VPC which allows read-only access to the synced repos.
See outputs for how you can use it to access your repositories in the private subnets.
In order to set up the sync, put all of your repository configurations in one directory. When deploying the configuration via Terraform, set your They will end up in /etc/yum.repos.d/ on the container and will be synced to the bucket.
The path in the bucket will look like this: http://ENDPOINT_URL/REPO_ID , where REPO_ID is the ID in the .repo file
website_endpoint - the URL underneath which you can access the Bucket from within your VPC.
Use this to configure your repostiroy in the private subnets.
bucket_id - the ID of the bucket the reposync uses