How to solve: Detected non-empty blocksets with no associated blocks!

Hi,

I have read a lot of threads regarding this issue. I will summarize some suggested ways to solve here:

  • Do some SQL statements to identify the file within the SQLite database
  • Remove the database and do a repair
  • and some much more complicated suggesttions…

Sorry, I am an end-user. I tried to follow the suggestions but it struggled on each of them.

Trying to use the SQLiteBrowser I found some databases (C:\users\username\AppData\Local\Duplicati\*.sqlite):

  • Duplicati-server-sqlite
  • FTXTXCPCTW.sqlite
  • PRNZUXKMDD.sqlite
  • RAPRRZSMQK.sqlite

Whatever file I am trying to open I need to enter a passphrase. Which I do not know. So no way to open and do the recommended SQL-Commands. And even if knowing- which database is the one related to my failing backup-job?
Ok, giving up on SQLite. :frowning:

Suggestion to remove the database and do a database repair. Well, same question arises: WHICH database should I remove?

Please gibe me some hints so I can (at least trying to) fix the issue.

THANKS!

/KNEBB

Hello and welcome!

As far as I know, this error is best handled by running a Repair operation. I think the underlying issue has been fixed in the Canary channel, so when the next Beta is released and you upgrade, you will hopefully never see this problem again.

On the Windows platform, the sqlite databases use some basic encryption. It is not compatible with SQLiteBrowser. I don’t think this is a big deal. Unless you are an advanced developer, you’re not going to be able to fix this problem by editing the databases directly anyway.

Hi,

I did a database repair for the accecfted backup. No change.

My current version is 2.0.5.1_beta_2020-01-18 (yes, Win)- shoul I really use a Canary as upgrade channel?

Thanks!

/KNEBB

Try a full database recreate to see if that helps. You may want to make a copy of the existing job database just in case.

In general I don’t recommend the Canary channel for regular use since it can at times contain bleeding edge features. That being said the current Canary is better than Beta, in my opinion. Hopefully a new beta based on it will be released soon.

Hi,

thnaks for suggestions. I removed the database file (found out the name is shown in the backup’s configuration) and re-created it. It took a while.

Now I got some warnings (8) and loads of errors (259):

2021-03-14 16:25:04 +01 - [Warning-Duplicati.Library.Main.Operation.RecreateDatabaseHandler-FileProcessingFailed]: Failed to process file: duplicati-20210226T220000Z.dlist.zip.aes
2021-03-14 16:30:38 +01 - [Error-Duplicati.Library.Main.Operation.RecreateDatabaseHandler-IndexFileProcessingFailed]: Failed to process index file: duplicati-i005093a0b58949fca6b79b65df7bc49d.dindex.zip.aes

Unfortunately I have no clue how to fix this…

Any further suggestions?

/KNEBB

duplicati-i005093a0b58949fca6b79b65df7bc49d.dindex.zip.aes
duplicati-20210226T220000Z.dlist.zip.aes

Those failed to be processed. No idea why. Those two lines do not say the exact cause and the code covers a good amount of possibility.

Might still be related to blocks though?

Might be a long one without a debugger on the code.

I can create a Windows-VM, install current developer version of Duplicati and point it to this backup. And then create the database.
If I have to delete the backup afterwards I do not mind. I will just help finding the issue.

Will this help?

If the remote is in good order, you should be able to delete more of the local files (of the backup db, not your actual files that were backed up) if not all of it as it just needs to be listed in Duplicati. A repair at that point should re-create the required local suff according to a test that I did some days ago. Vice versa for remote if local is correct.

For using a VM or whatever, I feel that’s a bit of a waste personally. I did more testing including intentional breaking to a backup on this computer and its still going fine :slight_smile:

One more note, if you didn’t try the canary build yet, you could try that. I’m using a source build with latest github code and backups are running fine for me though can’t guarantee it won’t be a problem. However, someone does state in the above link that canary and a full repair there fixed their problem.

I also did a repair to a db downgrade from source to beta on this other computer by doing the stuff I mention above in this post and it also appears to be working fine. But YMMV.

Hi,

ok, got it. There is no debugging or troubleshooting needed from my side.

I am fine with deleting these backups (I have a duplicate backup on different location) but I will not go and delete all my local stuff just to fix this error. It will be much easier to delete the backup and start from scratch. And hopefully not going into the same pitfal again…

Anyway, thanks to all for your help and suggestions!

/KNEBB

The local stuff is the backup db mind you, not local your actual files files but the db files for the backup and a repair will re-create them. Maybe I didn’t make that clear. Just to avoid confusion. I heavily edited my post above as well since I feel it was continually lacking in clarification.

Hope the best :slight_smile:

ok, understood :laughing:

Well, I deleted the database already and the rapir did pop up with the above erros. However, I am going to delete everything an start from scratch.

Thanks again!

/KNEBB