This program will synchronize one or more of Linux directory structures with Amazon S3 storage. It is designed to leverage
low cost AWS storage. Currently this build only supports backing up rpm
based installations (i.e. Fedora, Redhat,
CentOS etc.). Feel free to pull down the source code and compile on another Linux flavor in hopes it may work.
First you need to populate the config/config.json
file with the approprate configurations. Some of these you will
need to first create from within your AWS account:
{
"AWS": {
"S3Region": "us-east-2",
"S3Bucket": "BUCKET_OF_CHOICE",
"AccessKeyId": "AWS_ACCESS_KEY_ID",
"SecretAccessKey": "AWSX_SECRET_ACCESS_KEY",
"BackupDirectories": [
"/home/user/directory1",
"/home/user/directory2",
"/home/user/directory3",
"/home/user/directory4"
],
"ACL": "private",
"ContentDisposition": "attachment",
"ServerSideEncryption": "AES256",
"StorageClass": "GLACIER"
},
"logging": {
"logfile_location": "/home/user/backups.log",
"max_backups": 4,
"max_size": 1,
"max_age": 1
}
}
Many of the above configs should be relatively straightforward. A few caveats:
StorageClass
: Storage classes for AWS S3 can be found here. The default in the config file isGLACIER
.logfile_location
: Make sure to put your logfile in a valid location. Also be sure to build some sort of logfile rotator since this release ofS3Backup
does not rotate logs and they may grow to fill the disk.