Existing backup fails with "STORAGE_NOT_CONFIGURED"

This was an initial backup, which unfortunately got aborted due to an upstream ISP issue. I have some data in the bucket already, but now when I try to resume the backup I see this error in the log. I have verified that the config file does exist in the root - in fact it looks like there’s two copies - one hidden with 0 bytes and the other with 842 bytes.

An ideas on how I can initialize the storage again without starting this backup over again?

(trying to sanitize output since new users can only post 2 URLs).
“backup destination” is https://f002.backblazeb2.com

Running backup command from /cache/localhost/6 to back up /backuproot/data/media
Options: [-log -d backup -storage 2024-unraid-media -threads 30 -stats -stats]
2024-02-28 18:06:04.634 INFO REPOSITORY_SET Repository set to /backuproot/data/media
2024-02-28 18:06:04.634 INFO STORAGE_SET Storage set to b2://2024-unraid-media
2024-02-28 18:06:04.634 DEBUG PASSWORD_ENV_VAR Reading the environment variable DUPLICACY_2024-UNRAID-MEDIA_B2_ID
2024-02-28 18:06:04.634 DEBUG PASSWORD_ENV_VAR Reading the environment variable DUPLICACY_2024-UNRAID-MEDIA_B2_KEY
2024-02-28 18:06:04.735 INFO BACKBLAZE_URL Download URL is: <backup destination>
2024-02-28 18:06:04.825 DEBUG PASSWORD_ENV_VAR Reading the environment variable DUPLICACY_2024-UNRAID-MEDIA_PASSWORD
2024-02-28 18:06:04.923 TRACE BACKBLAZE_CALL [0] URL request 'HEAD <backup destination>/file/<bucketname>config' returned status code 404
2024-02-28 18:06:04.923 DEBUG BACKBLAZE_LIST <backup destination>/file/2024-unraid-media/config did not return headers
2024-02-28 18:06:04.923 ERROR STORAGE_NOT_CONFIGURED The storage has not been initialized
The storage has not been initialized

Which version of duplicacy CLi are you using?

Maybe this applies to your case too: File not found: how do I start from scratch? - #9 by gchen

I’m on 3.2.3, so I don’t think it applies?

I’ve read that removing the config file from B2 is bad… So what about if I re-initialize the storage from Duplicacy? Not sure what impact that will have.

Thanks for checking!

Is this a correct path to a config file?

Can you copy config file locally and verify that it is working (with a fake test repo)?

You will lose access to all previous versions.

Yes, that’s the location of the config file. I can definitely copy it down to my computer, but since it’s encrypted, I can’t read it. Doesn’t the “fake” test repo need to be same name as the prior one? Like the config file is linked to it somehow?

No, config file is independent, it contains encryption keys, and does not care where it is located.

You can just put a config file in a folder and add the location to duplicacy. If it recognizes it a a valid duplciacy datastore – your config file is OK, and the issue is specific to Backblaze somehow.

You appear to be correct. I downloaded/deleted/re-uploaded the config file and it’s not failing now. It recognizes the previous backup was incomplete, so hoping this works…

2024-02-29 13:29:30.612 TRACE CONFIG_INFO File chunks are encrypted
2024-02-29 13:29:30.612 TRACE CONFIG_INFO Metadata chunks are encrypted
2024-02-29 13:29:30.613 DEBUG BACKUP_PARAMETERS top: /backuproot/data/media, quick: true, tag: 
2024-02-29 13:29:30.613 TRACE SNAPSHOT_DOWNLOAD_LATEST Downloading latest revision for snapshot unraid-media
2024-02-29 13:29:30.613 TRACE SNAPSHOT_LIST_REVISIONS Listing revisions for snapshot unraid-media
2024-02-29 13:29:30.643 INFO BACKUP_START No previous backup found
2024-02-29 13:29:32.222 INFO INCOMPLETE_LOAD Previous incomplete backup contains 31275 files and 4024291 chunks
2024-02-29 13:29:32.222 INFO BACKUP_LIST Listing all chunks
2024-02-29 13:29:32.222 TRACE LIST_FILES Listing chunks/

Well, looks to be working. It’s estimated 6 days to complete and I think I can upload it again in under 3 days, but let’s see if it catches up.

Wait, was this new storage location where you just started uploading into? Or was there supposed to be existing version history?

Correct. As stated in my OP: “This was an initial backup, which unfortunately got aborted due to an upstream ISP issue. I have some data in the bucket already…”

Ah, good. I missed that bit. Then it’s all good.

Still it’s unclear why did b2 behave this way. Network interruption should not have touched the config file – it’s written once and never touched since (only read)

I’m not sure, but I have a guess. In the root of the B2 folder there were two ‘config’ files. Maybe the one Duplicacy was trying to access was the null one? Not sure, but I’m glad I was able to recover. Thanks for the tips!

This topic was automatically closed 10 days after the last reply. New replies are no longer allowed.