Backup to tebi not working
-
I tried tebi.io and it seems to work in 9.0.15 , so maybe it also got fixed with this change .
-
I tried tebi.io and it seems to work in 9.0.15 , so maybe it also got fixed with this change .
-
I am also using a prefix. What is the error you are getting?

-
@girish same as OP, seems to be a local issue, as the bucket has the prefix created and what looks like the first app backup, but then errors.

@robi ah, it could be that old (invalid) backups are not found. I think if you delete the Site and add it again, it will work . Can you try that?
-
@robi ah, it could be that old (invalid) backups are not found. I think if you delete the Site and add it again, it will work . Can you try that?
-
@robi ah, it could be that old (invalid) backups are not found. I think if you delete the Site and add it again, it will work . Can you try that?
-
@girish created new config, started backup, began with first app, then it decided to delete old entries, then it failed:

Maybe it's an rsync vs tgz issue now, in combination with the former path "/" issue?
@robi said in Old backup not found when updating app:
Maybe it's an rsync vs tgz issue now, in combination with the former path "/" issue?
Oh, this could be. Testing
-
@girish created new config, started backup, began with first app, then it decided to delete old entries, then it failed:

Maybe it's an rsync vs tgz issue now, in combination with the former path "/" issue?
@robi I could reproduce it. There is a bug in the tebi service. It is listing directories as part of listObjectsV2 API call. S3 has no concept of directories, just key/value pairs.
I have reported this to them.

-
@robi they replied to me saying they won't fix it . They did tell me why, but you might want to reach out to them for more info, don't want to breach confidentiality.
-
G girish has marked this topic as solved
-
Whoa, unexpected.
One could also say that it's not a bug, as object stores don't distinguish files from directories, as they're all objects. It's the implementation that does.
So it may be more likely that the s3 client/driver/config needs to compensate.
ex. Cyberduck works fine.