Support

I cannot give you an exhaustive list of full URLs, because parts of those URLs are created dynamically based on your Microsoft account, the drive you are accessing, and its folder. I can, however, give you the common parts of the URLs we use when accessing the OneDrive APIs (legacy and modern) so your IT department can create filters which will whitelist all dynamically generated URLs.

• Authentication endpoint: https://www.akeeba.com/oauth2/onedrive.php and https://www.akeeba.com/oauth2/onedrivebusiness.php accessed with GET and POST requests.
• OneDrive Legacy endpoint: https://api.onedrive.com/v1.0/ accessed via GET, POST, and PUT requests.
• OneDrive and OneDrive for Business endpoint: https://graph.microsoft.com/v1.0/ accessed via GET, POST, and PUT requests.

IMPORTANT! These are not all the URLs. OneDrive returns its own, unpredictable, upload URL when using chunked uploads. I cannot give you these URLs because they do not follow any published pattern, nor can we predict them. Their structure is entirely up to Microsoft.

If you need to know all the URLs in advance, you will have to DISABLE chunked uploads in Akeeba Backup, thereby preventing the generation of the arbitrary URLs I described in the previous paragraph. Disabling chunked uploads will use the OneDrive simple upload API which only use the URLs I mentioned in the bullet points above.

However, if you disable chunked uploads you are facing other issues. If you don't make any other changes you will be trying to upload very large files to a very slow server which means that it takes a very long time, therefore you will face timeouts preventing you from uploading anything to OneDrive. To prevent that, you will have to set the Part Size For Split Archives to a fairly small value (typically between 20 and 100 MiB will do on most servers) under the Archiver Engine options section of the Configuration page. This means that you will have multiple files generated for each backup you take. For example, a 2 GiB site backup with a part size of 100 MiB will generate 21 files with extensions .j01, .j02, ..., .j20, .jpa (listed as first to last part of the backup). You will need all of these files to be present to restore your backup; the interface will only show you the last part (.jpa), but it expects all files. Think of these files as the volumes of a big book, like Victor Hugo's Les Miserables. You can't read the book if you only have the last volume; you need to have every volume to read the book start to finish.