This module adds s3 download capability with file resource controls

Phil Watts



8,017 latest version

3.9 quality score

Version information

  • 99.99.99 (latest)
  • 0.1.6
  • 0.1.5
  • 0.1.0
released May 15th 2015
This version is compatible with:
  • RedHat|OracleLinux

Start using this module


marsupermammal/s3 — version 99.99.99 May 15th 2015

Depricating this module. Turns out the wheel didn't need reinventing. Please see nanliu/staging.

Function: This module is used to pull files from AWS S3 locations using the aws ruby SDK, aws cli, or curl mechanisms. It supports both single file and recursive directory sync. It also allows you to manage the files pulled from S3 with common parameters of the Puppet file resource, such as owner, group and mode.

Usage: In the desired manifest, add include s3 class { 's3': } or to control installing the awscli class { s3: include_tools => true, } Then declare resources: Single file: s3::cp { 'name of file to be created':
bucket => 's3 bucket location, in format s3://bucket_name', source => 'path after bucket in s3', dest_path => 'path the file should be copied to locally', s3name => 'name of file on s3', owner => 'owner_name', group => 'group_name', mode => 'mode', }

Recursive Directory: s3::sync { 'name of directory to sync "equiv to $source above"': bucket => 's3 bucket location, in format s3://bucket_name', dest_path => 'top level location to sync directories, should be parent dir of $source/$namevar', owner => 'owner_name', group => 'group_name', mode => 'mode', }