Amazon Web Services

AWS Command Line Interface for Auto-Publishing

Install AWS Command Line and Configure

pip install awscli

Or use the Windows 64 bit installer.

  • Configure

    1. Access Key ID

    2. Secret Acess Key

    3. None

    4. None

aws configure

Note: I have to be in C: where it is installed.

Setup Auto-Publish Script

Base Command: aws s3 sync <source> <target> [--options]

Rsync Style Command: aws s3 sync <source> s3://<target> --delete

Note: A batch script with “call” + the above line in it makes for a quick and easy way to publish the website.

AWS Access Key ID, etc.

To Create User/Initial Access Key:

AWS Console -> IAM -> Users -> Add User

->User Name

->Programmatic Acess

To Get New Access Key:

AWS Console -> IAM -> Users -> Username -> Security Credentials

-> Make Inactive

-> Create Access Key

Groups:

A user needs AWSCodeCommitFullAccess to pull/push to git repositories and AmazonS3FullAccess to push html to the website content hosting s3 bucket.

AWS CodeCommit for Version Control

Setup Tutorial: http://docs.aws.amazon.com/codecommit/latest/userguide/setting-up-gc.html

1. Git

Download Git https://git-scm.com/download/winhttps://git-scm.com/download/win

Install Git

-> Use Notepad++ as Git’s Default Editor

-> Use Git from the Windows Command Prompt

-> Checkout Windows-style, commit Unix-style line ends

-> Use Window’s default console window

2. AWS IAM

Create User Give User AWSCodeCommitFullAccess

3. Get Credentials & Create Repo

Follow Tutorial Step 3 and 4

-> Generate Git Credentials through IAM-> User -> Security Credentials -> Git Credentials

-> Download Credentials

4. Clone Repo

-> Go to Folder where you want the repo (i.e. U:// or ~/ )

-> git clone the repo (Repo Directory url can be found by loging into AWS with [Non-Root] User and opening Repo)

-> Give Credentials

4.B Clone Shallow Repo

If the full repository is too long, you can clone a shallow repository that will only keep parts of the master repository.

git clone --depth=n <url>

Will only keep the n newest commits in the current repo. A good number here is 3 since very shallow repos can’t always be pushed to the master.

If that happens, use git fetch --depth=m where m>n to fix it.

5. Code!

git clone url -> Gets Code Repo

git status -> gives status

git branch -> gives branch name

git commit -m "Change-Log Message" -> Commits Changes to Local Repo (Do Anytime there are Changes)

git remote -> gives remote rep nickname

git push remoteNickname branchName -> Pushes to AWS Repo and Branch

git pull -> gets from AWS Repo

6. Workflow

  1. git pull

  2. Do Work!

  3. git add .

  4. git status

  5. git commit -m "Change-Log Message"

  6. Go To 2 or at End of Work:

  7. git push

7. Remove History when Repo gets too big

git fetch --depth=1 to prune the old commits

git reflog expire --expire-unreachable=now --all to expire old commits and objects

git gc --aggressive --prune=all to remove the old objects and clean up

8. Issues Storing Credentials

If there’s issues storing credentials, such as on linux systems, try:

git config credential.helper store

then

git pull

and git should store the credentials. Beware that this stores git credentials on the disk in plaintext so is less secure than other options.

AWS Glacier for Long Term Backup

Create Vault

  1. Go To AWS Console

  2. Go To Glacier

  3. Create Vault

  4. Optional: Set up notifications

  5. Optional: Go to Settings and Limit Data Retrieval Cost

Fast Glacier

Free Windows Client for Glacier: https://fastglacier.com/

  • Supports Multi-Part upload

  • Smart (Estimated Cost) Data Retrival

  • Folder Sync/Comparison

  • Command Line Folder Sync Tool

Setup:

  • Download and Install Client

  • Use Account Type: Amazon Glacier

  • Get Access Keys

    • Follow Link

    • Go to Acess Keys

    • Create Keys

    • Show Access Keys

  • Encrypt Keys with Master Password

  • Use SSL

Use:

  • Select Vault

  • Go to Files

  • Upload Folder