Do not use B2 retention. Use the MSP360 Backup retention policy instead. If B2 starts deleting files, Backup will get out of sync and you may not be able to restore any files removed via B2.

David, thanks for your response. I want to adhere to your recommendation regarding no B2 retention. Am I correct in my understanding that this will reset everything following my testing:
1) empty my B2 S3 bucket
2) disable B2 S3 retention
3) invoke CBB Synchronize Repository

That's probably the safest. We don't create (or support) multiple file versions on B2, but if they're there for some reason, starting fresh is probably ideal if it's not an inconvenience for you. You could also create a new storage account and bucket and enable S3 access using the latest version with B2 S3 storage support connector if you wanted S3 API access - which gives you synthetic full backup support for image based backups. That way you'd keep your current backups intact until the new backups are complete.

David, what is meant by "We don't create (or support) multiple file versions on B2"? Example, if I have a file on my PC that is changed prior to every new backup (file or image) will my B2 backup contain all versions of this file (in accordance, of course, with the retention policy I have set in CBB)?

This also brings to mind another question: Can the same B2 bucket support both file and image backups at the same time, or should I create a separate bucket for each type?

Also fyi, my current backups on B2 S3 are tiny and strictly for testing - there is no value to me in keeping them.

Again, thanks for your help as I navigate these new waters.

We do the versioning ourselves. That's fully supported. But if you enabled versioning in the bucket and copied up the same file over and over outside of our product, you'd end up with multiple versions of the same file (using the same name in storage) which we don't support.

You can use the same bucket for an entire company if you want. All data is segregated by the computer prefix and backup type.

David, I have followup questions regarding CBB versioning vs B2 bucket versioning. I want to feel comfortable about my ability to recover a 2+ TB data backup from B2 should I experience a ransomware attack and need to do a full restore from a point in time before the attack.

My concern: if CBB is doing the versioning does this eliminate the ability for Backblaze to prepare a point-in-time snapshot of my backup, which they can then send to me on a hard disk. If I am still able to use Backblaze's snapshot on a hard disk, am I able to use CBB to do the restore?

If I'm wide of the mark here, what is the recommended best practice with CBB for protection against ransomware?

Thanks for your help.

You do not have to worry about Backblaze B2 versioning features. We handle all the restore features, including point-in-time, versioning, etc. If you requested a drive from Backblaze, you'd get exactly what's in storage and exactly what MSP360 needs to perform the restore; including all versions of all files that can be pulled from.

Thanks David. Just to be sure I fully understand, please confirm the following is correct:

After an initial full hybrid backup of my data volume to my Synology NAS and my B2 S3 bucket, I will do daily hybrid backup of changed files. All versioning/retention will be managed by CBB. If I lose the original PC data volume AND the NAS backup I can copy the Backblaze restore disk drive contents to the NAS, and then reconnect CBB to the NAS and do a point-in-time restore from the NAS to the PC data volume.

If, on the other hand, a single folder of a few GB data is corrupted on the PC and NAS, I would just use CBB to restore the folder from B2 to the PC and the next CBB hybrid backup would put everything back in sync.

Correct. You could restore everything from B2 without asking for a drive delivery - but depending on number of files and downstream bandwidth, a drive might be the better option. You could also restore right off the USB drive if needed.

David, I may have to back off my earlier criticism regarding documentation - I've found some pretty useful information about the CBB repository today. That said, I do have these additional questions I hope you can help with. I appreciate your assistance.
.
1. Before I even knew there was a repository I was doing a lot of testing. I would create and run a backup plan, do a restore, and delete the backup data using Windows explorer, and delete the test backup plans. I have now executed the "shrink database" but I assume there are a lot of leftovers in the repository. Is there a way to reset it before I start from anew with my real backup plans?

2. Recall I am backing up my PC to my Synology NAS and to Backblaze B2. As insurance against ransomware I would like to perform regular BTRFS Snapshots of my NAS backup folder. Then, if my CBB backup on the NAS becomes encrypted, I would roll back to a snapshot prior to the infection so I could restore my PC from the NAS. But then the backup would be out of sync with the CBB repository. Am I correct that I would need to Sync the repository before restoring to the PC? I've read some older posts where users had problems with large syncs failing. Hopefully that is a problem that has been resolved. Any best practice recommendations for the scenario I describe?

3. Assume I have 2TB backed up on Backblaze B2 and in the future decide to move to a different cloud service. What is the most efficient way to delete all of B2 backup information from the repository? I hope it is as easy as deleting the backup plan in CBB and shrinking the repository.

4. Can I safely convert a local files backup plan to a hybrid plan?

5. Can I safely change encryption or compression settings in a backup plan after it has been run?