aws-bunker 0.4.5

bunker is a command line program for setting up an ec2 in AWS for remote development or as a backup. It can clone your git repos, and transfer ignored files from your machine to the ec2.

bunker

Bunker is a command line program for setting up an ec2 in AWS for remote development or as a backup. Among other things, it can clone your git repos and transfer ignored files from your machine to the ec2.

Working remotely, using cellular networks has been possible for ~15 years. As networks have evolved, bandwidth has become less and less of an issue, but even on today's cellular networks it is both impractical and costly to download large files, like docker images, zip archives, etc. One solution is to ssh into a VM on the cloud, and work from there, but managing git repos and private or ignored files across several machines quickly becomes problematic.

With bunker it is easier to automate this process. All you need is an instance ID for and AWS EC2 running ubuntu, and you can run shell scripts, source files to .profile, and copy over as many repos and ignored files as you want. Example:

# init bunker with your instance ID, etc  
$ bunker init  

# save some files to the ignored files list  
$ bunker ignore config.ini .env.production terraform.tfvars  

# run `/home/ubuntu/install-things.sh` on the instance  
$ bunker install /home/ubuntu/install-things.sh  

# source `/home/ubuntu/.prompt` to the .profile on the instance  
$ bunker prompt /home/ubuntu/.prompt  

# clone a local repo on the instance, and then copy over any files from the ignored file list  
$ bunker build -r /path/to/local/repo   

When used with our sample Terraform files (https://gitlab.com/shindagger/bunker), you can spin up a brand new EC2, clone the repos you need, copy over the ignored files, and be working in ~5 minutes. Terraform will output an instance ID you can use for bunker init. Then you can do the rest with a single command like:

$ bunker install -y && bunker prompt -y && bunker build -r /path/to/local/repo -y   

For more information see the Terraform and Usage sections below.

requirements

• python
• aws cli
• (terraform)

Also, you will need an EC2 instance on AWS, preferably running ubuntu>=18.04. This EC2 will need ssh (port 22) open and an attached SSM role (AmazonEC2RoleforSSM).
For more information on creating the SSM role for EC2 see:
https://docs.aws.amazon.com/systems-manager/latest/userguide/setup-instance-profile.html#instance-profile-add-permissions

installation

$ pip intall aws-bunker

NOTE: see shell scripts, and terraform sections below:
PROJECT HOME PAGE: https://gitlab.com/shindagger/bunker

shell scripts

the install and prompt subcommands assume there will be shell scripts on the EC2. The default names (without positional arguments) are: install-essentials.sh, and .prompt.sh respectively.

the first shell script is intended for essential software (like python, pip, node, and npm). If the script is not already executable, use the [-e] flag. prompt simply sources a file in ~/.profile on the EC2.

See example scripts here (https://gitlab.com/shindagger/bunker) as well as the terraform section below:

terraform

We have also included terraform files (https://gitlab.com/shindagger/bunker) which, when configured will spin up an EC2 instance you can use for bunker, with all the example shell scripts and your .vimrc provisioned. Of course you're encouraged to change these files according to preference. Terraform will output an instance ID you can use to init bunker with.

To configure and use these terraform files:

1. Clone this repo (https://gitlab.com/shindagger/bunker.git)

2. cd into the repo dir and ls *.default. These are the files you will need to edit before you init terraform. Each one of these files will need the extension .tf.default changed to .tf.

3. Edit alfa.tf and ensure that the region is correct.

4. Edit beta.tf and change the s3 bucket to a bucket you own. Also ensure the region is correct.

5. Edit terraform.tfvars with your AWS SDK access id and key. Also make sure you are configured to this account with aws configure

6. Edit variables.tf and change the default values to reflect your wishes. NOTE: if you don't have them already, you will need to create your own ssm role, and ssh key file manually. ec2_key_name is just the name of the keypair in AWS, while ssh_private_key is the full (absolute) path to your .pem file on your local machine.

7. terraform init

8. terraform plan && terraform apply

before you init bunker

1. You have an EC2 instance ID. (see "terraform" above)

2. You have a directory (prefix) in which you will keep repositories.txt and ignoredfiles.txt. You can copy sample files from this repo, and edit them to reflect your repos and ignored files.

3. You have a .pem file and you know the path to its location. You have a keypair name set for your EC2, which uses this .pem file.

4. You know the username for the EC2 (should be "ubuntu"). We may choose to support other OSs in future versions.

5. You have created a private key for your git provider, and you added the public key in your git provider settings. Also, you have configured your local ssh to use this key and tested or used it at least once.

bunker usage

$ bunker -h

Show main bunker help page and exit

$ bunker build -h

Show help page for the subcommand build and exit. This will work with any available subcommand

$ bunker init

Write the bunker config file ~/.config-bunker.ini. If the file exists already, answers will be populated with existing values [underlined], and only overwritten if you offer a replacement value.

example ~/.config-bunker.ini:

[default]  
prefix = /Users/username/bunker-config  
repos = repositories.txt   
ignored = ignoredfiles.txt  
pem = /Users/username/somename.pem  
instance_id = i-04xxxxxxxxxxxx148  
username = ubuntu  
gitkeys = /Users/username/.ssh/git_rsa  

$ bunker repo path/to/repo /absolute/path/to/another-repo ../somerepo

Add or remove repo directories to/from repo file. Use relative or absolute paths.

$ bunker ignore .env.local config.ini terraform.tfvars

Add or remove filenames to/from ignore file.

$ bunker install
$ bunker install /home/ubuntu/script.sh

Install essential software on your EC2 instance. This command will run an executable shell script on your EC2. The default script (without defining the positional argument) is named /home/ubuntu/install-essentials.sh. This is intended to install any software essential to using your EC2. In an effort to stay agnostic about what software users want on their EC2, bunker does not install with included shell scripts. We do however provide example shell scripts (as well as terraform for the EC2) at the project homepage:
https://gitlab.com/shindagger/bunker

$ bunker prompt
$ bunker prompt /home/ubuntu/another-prompt.sh

Source a custom prompt file in /home/ubuntu/.profile. This command will source a file in /home/ubuntu/.profile. The default name is: /home/ubuntu/.prompt.sh.

$ bunker build

Clone list of git repos (repositories.txt) on your EC2 and then rsync list of ignored files (ignoredfiles.txt) from your local repos to the EC2 repos. git clone will use ssh, assuming you have a public key setup at your git provider, a working ~/.ssh/config tested at least once with ssh -T git@PROVIDERURL, and the associated private key set using bunker init.