Error in My Library, others are fine

Hello everyone

Out of a sudden, My Library (the default library - I have French locale so this is Ma Bibliothèque) started to have problems when trying to upload from clients (they claim that the server has an error, looks like it is reporting 500).

I did not make any upgrades, not configuration changes. I was on 6.3.2 and (when there were issues already), moved to 6.3.4 to be on the latest version (there were no changes in the errors).

On the server logs, I see the following error messages:

ccnet.log (the error is repeated once in a while):

[11/20/18 19:13:42] ../common/ccnet-db.c(425): Error prepare statement SELECT g.group_id, group_name, creator_name, timestamp, parent_group_id FROM `Group` g, GroupUser u WHERE g.group_id = u.group_id AND user_name=? ORDER BY g.group_id DESC: sqlite3_prepare_v2 failed: no such column: parent_group_id.

seahub.log: plenty of

2018-11-20 18:15:01,008 [WARNING] django.request:152 get_response Not Found: /error_page.html
2018-11-20 18:15:02,887 [WARNING] django.request:152 get_response Not Found: /api2/events/
2018-11-20 18:15:13,963 [WARNING] django.request:152 get_response Not Found: /api2/events/
2018-11-20 18:15:31,022 [WARNING] django.request:152 get_response Not Found: /error_page.html
2018-11-20 18:15:32,037 [WARNING] django.request:152 get_response Not Found: /error_page.html
2018-11-20 18:16:01,992 [WARNING] django.request:152 get_response Not Found: /error_page.html
2018-11-20 18:16:02,948 [WARNING] django.request:152 get_response Not Found: /error_page.html

controller.log and seafile.log seem clean.

You have to stop seafile and then run seaf-fsck.sh . This will check for corrupted libraries.

This error was due to the disk seafile/seahub run from (data is stored on another one).

Freeing up space fixed the issue without any more actions.

Thank you. The issue ended up being a full disk but I will keep this utility handy.

Have you ever ran the GC?

What is GC?

The garbage collector: https://seafile.readthedocs.io/en/latest/backup/backup-rsync/#3-seafile-gc

https://manual.seafile.com/maintain/seafile_gc.html

If you never ran GC, all deleted files are still on your disk.

Thanks - this is very useful. I will have it running once in a while in a systemd timer.

The disk overflew not because of seafile - I run it in a container where the disk is shared with other containers and one of these other containers went south. I am finally setting up quotas and monitoring this week-end :slight_smile:

Try adding TIME_ZONE = ‘UTC’ in seahub_settings.py
Hope this will solve your problem, as it did mine.