DatabaseChecksumError on fsck

[rabidmuta...@gmail.com]

Corrupt remote data; what is the recommended recovery procedure?

I found this: https://groups.google.com/g/s3ql/c/4LM5-5uXfQY/m/6amCKocRAQAJ

...is it the best approach?

fsck.s3ql <redacted>
Starting fsck of  <redacted>

Using cached metadata.

File system was not unmounted cleanly
Checking DB integrity...
Verifying consistency of most recent metadata backups:
Checking backup 000000becb (from 2025-01-06 00:46:32))...
Downloaded 2031/17560 metadata blocks (11%)
Downloaded 17560/17560 metadata blocks (100%)
ERROR: Uncaught top-level exception:
Traceback (most recent call last):
  File "/usr/bin/fsck.s3ql", line 33, in <module>
    sys.exit(load_entry_point('s3ql==5.1.3', 'console_scripts', 'fsck.s3ql')())
             ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib64/python3.11/site-packages/s3ql-5.1.2-py3.11-linux-x86_64.egg/s3ql/fsck.py", line 1342, in main
    verify_metadata_snapshots(backend, count=5, include_most_recent=check_current_metadata)
  File "/usr/lib64/python3.11/site-packages/s3ql-5.1.2-py3.11-linux-x86_64.egg/s3ql/fsck.py", line 1435, in verify_metadata_snapshots
    conn = download_metadata(backend, fh.name, params)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/lib64/python3.11/site-packages/s3ql-5.1.2-py3.11-linux-x86_64.egg/s3ql/database.py", line 543, in download_metadata
    raise DatabaseChecksumError(db_file, params.db_md5, digest)
s3ql.database.DatabaseChecksumError: File /tmp/tmp20kx_gk_ has checksum 933548326db2ae62d820c2b703cae18c9dda8f7d063f53773f869e58c5c6c1fc, expected a4ee0e2c03c34c17341e2d1fe0213a9c614d3a715111c9ce3b0eb04fb3c5a595

[Daniel Jagszent]

Hello,

Am 10.01.25 um 07:13 schrieb rabidmuta...@gmail.com:

Since your local database (SQLite file) looks ok ("Checking DB integrity..." line without error message), ignoring these remote database backup checksum mismatches is a valid route. You should find out why this happened, though. Were you affected by https://github.com/s3ql/s3ql/issues/321 ? (Did you use S3QL Version < 5.1.1 at 2025-01-06 and is your S3QL cache directory a symlink?) Also, do not forget to add five valid remote database backups and revert your patch afterward, as Henry explained.

Which version of S3QL are you using, btw. The exception indicates a strange mix of version 5.1.3 and version 5.1.2. If it is indeed 5.1.2, updating to the newest version (5.2.3) might be all that is necessary. Only from version 5.1.3 upwards, fsck.s3ql skips unclean metadata backups since they will most likely produce those database checksum errors. All 5.x versions of S3QL use the same database structure; updating to the latest version should be relatively hassle-free, and a downgrade is possible (but not advised).

[rabidmuta...@gmail.com]

Thank you for this insight! Despite appearances, I am/was running 5.1.2. Your suggesting of using a later version resolved the problem: simply running fsck from 5.1.3 worked without issues. Next step is to upgrade to latest version.

Thanks for the suggestion.