This repo contains a number of scripts that act as wrappers around azcopy and rsync. You can find them here.

The wrappers allow batch (i.e. multiple sync) operations to happen in sequence, and implement (additional) retries. They are implemented using PowerShell, with the batches configured as JSON files.

Why would you want to sync files to a redundant location? If you're using a cloud storage solution such as DropBox, Google Drive, iCloud or OneDrive, your files are safe you'd assume. Well, yes and no. They protect you from any hardware problems you may have on the device you're syncing files from. That is, you still have an online copy of the files in case you lose a device due to a hardware issue or altogether. However, they do not (or only in a limited way) protect you against logical errors. A logical error is when you accidentally delete files in a file (move) operation. Your deletes will be synced to the cloud and files will be removed (eventually) there as well. A classic example is overwriting a folder containing files with a folder with no files using the Finder on macOS. Once you confirm the overwrite, your original files have been deleted. If you discover what happened too late, data may be lost forever.
Last, incidents have happened in the past were cloud providers did lose your data e.g. Adobe Creative Cloud.

This repo takes the approach of archiving files to a (remote) destination where they will remain, no syncing of deletes. This works great for files you are unlikely to modify e.g. media files.

rsync is a tool with a long history on Linux that is also preinstalled on macOS. The sync_with_rsync.ps1 script takes a settings file with configured directory pairs and optional patterns and exclude list as argument. See example below:

{
    "syncPairs" : [
        {
            "source": "~/Pictures/Lightroom/Photos",
            "target": "/Volumes/External/LightroomBackup/Photos"
        }
        
    ]
}

See adapt the sample sample and pass its path as argument into sync_with_rsync.ps1:

sync_with_rsync.ps1 -SettingsFile /path/to/settings.json

azcopy is a tool that allows you to sync (among other things) a local directory to an Azure Storage Account. The sync_with_azcopy.ps1 script takes a settings file with configured directory pairs and optional patterns and exclude list as argument. See example below:

{
    "tenantId" : "00000000-0000-0000-0000-000000000000",
    "syncPairs" : [
        {
            "source": "~/Pictures/Lightroom/Photos",
            "target": "https://mystorage.blob.core.windows.net/lightroom/photos"
        }
    ]
}

This settings file requires an Azure Active Directory tenant to be configured through the tenantId field. This will allow the script to 'find' storage accounts configured in the settings file using Azure Resource Graph. Alternatively, the AZCOPY_TENANT_ID environment variable or the Tenant argument can be used.

You can work with pre-existing storage accounts, or you can use the create_storage_account.ps1 script to create one with recommended settings:

• Cross-region (RA-GRS) data replication
• Public access disabled
• Resource lock to prevent accidental deletion
• Soft delete enabled
• Storage Firewall enabled (sync will open and close it as needed)

create_storage_account.ps1 -Name mystorage `
                           -ResourceGroup files-sync `
                           -Location westeurope `
                           -Container lightroom, pictures, video

See adapt the sample sample and pass its path as argument into sync_with_azcopy.ps1:

sync_with_azcopy.ps1 -SettingsFile /path/to/settings.json

This script separates Azure control plane and data plane operations, with the latter happening after the former. Once azcopy has started, control plane access is no longer required and SAS tokens will be used for authentication.