I created a first plan to save about 20GB into the S3 Glacier Deep Archive storage tier. Then I decided that it would not suit me, hence I deleted the S3 bucket content (yes I know, I will have to pay for the files which stayed less than 180 days in this storage tier).
I deleted my backup plan and recreated a new one with the storage tier One Zone IA. I ran this plan and I've seen that it did not backup anything apart from the CloudBerry configuration file, because it determined that none of the files changed. But I do not car that my files did not change, I want to start from scratch with a new plan. Therefore, when we create a new plan, CloudBerry should use a new database to keep the history of the file modifications, it should not use any previous DB from a previous plan.
Which exact CloudBerry file should I delete to really start a new plan from scratch ?
Did you create a new plan? If so, the files would be backed up again automatically. I'm guessing you just changed the storage class on the existing plan. For that you'd need to synchronize the repository, but given that you used Glacier, it's better to create a new plan. You can clone the first one to save time and then modify the storage class as needed.
Then the correct behavior would have been to back up all the files again. I think it's best to send the logs to support for review. From the Tools - Diagnostic menu option. Please refer to this thread when submitting the ticket.
Here is the response from the support and I can confirm that their action plan resolved my issue:
Hello Nicolas,
Thank you for contacting us.
It is not a bug. The backup is performed according to the local database, containing all the records of the files uploaded. It was not updated, which resulted in this behavior.
Please synchronize the repository following these steps:
1. Go to Settings - Application - Synchronize Repository.
2. Choose your account from drop-down menu.
3. Press Synchronize Now