I’m trying to finally get my stuff backed up correctly.
So the idea was to upload the backups to Linode object storage.
I set it up, and everytime I try to back up, I get SignatureDoesNotMatch and Http Status Code Forbidden.
Strangely enough, if I change the bucket name, it does create that bucket, so the credentials are correct.
I also tried to upload something with Python and boto3, that worked also fine.
On the internet people suggest to check the credentials for special symbols, but I have only alphanumeric characters.
Some suggest to set the s3-ext-signatureversion to 3, others to 4.
Tried with s3-ext-signaturemethod HmacSHA1 and HmacSHA256.
Tried with and without SSL.
Most links on the internet refer to client libraries, but here in duplicati I don’t have much control.
Please see where you can get with the control on your Destination screen. Maybe Minio will work:
EDIT:
Narrow down the failing operation so you can at least say what you were doing if you seek Linode help.
See server log About → Show log → Retry. Backup to typically see a list and a series of put uploads. Towards end of backup you will typically see more list and get downloads, but failure might be earlier.
In the logs I can see that the list works fine, it’s the PUTs that fail.
{"ClassName":"Amazon.S3.AmazonS3Exception","Message":"Error making request with Error Code SignatureDoesNotMatch and Http Status Code Forbidden. No further error information was returned by the service.","Data":null,"InnerException":{"ClassName":"Amazon.Runtime.Internal.HttpErrorResponseException","Message":"The remote server returned an error: (403) Forbidden.","Data":null,"InnerException":{"ClassName":"System.Net.WebException","Message":"The remote server returned an error: (403) Forbidden.","Data":null,"InnerException":null,"HelpURL":null,"StackTraceString":" at System.Net.HttpWebRequest.GetResponseFromData ...
With Minio I get an error already when clicking “Test connection” in the config screen.
Oh, what I meant with “refer to client libraries” is that the people are having issues connecting to S3 compatible storage using a client library in Python or Javfa, etc and the solutions involve writing code.