Hello everybody!
Since months, I’m having a problem with my seafile clients. I have about 10 encrypted libraries that are synced between three Computers. On the first Windows PC, two of them report a “server error” without any further information, when it is trying to sync. On the second Windows PC, the same happens, but only for one of the two libraries. On the last Computer, which is a Linux laptop, all libraries sync successfully.
As said, there are no further error reports on the client, and since I rented the server from a company, I can’t see anything on the seafile server. I also wrote to the company serving, but they just told me they asked Seafile for help without any response for now.
Still, I am having the problem and I need a solution. I tried several things like cutting the connection to the specific libraries and reconnect, deleted the library from the PC and tried to sync this library from the start. My last solution would be to completely delete the library from the seafile web interface and start with a new library, but of course, syncing several GB would cost time and I’d loose the history of that library, which is a bad solution.
Do you have any things I could try beforehand?
Greetings,
Jonny aka. RastaTux
EDIT: First Windows PC has Seafile client-Version 8.0.10 and Win10, the other one the newest Version 9.0.15 for the client and Win11 - If that helps.
If you are not able to find server side error log, you can check the client side error log to help debug the problem.
You can click the “Open log directory” at the menu of system tray icon to find the log files.
Thank you Daniel, that was quite helpful!
However, the error says the following:
[11/17/25 20:02:58] Transfer repo ‘644e2664’: (‘normal’, ‘init’) → (‘normal’, ‘check’)
[11/17/25 20:02:58] Transfer repo ‘644e2664’: (‘normal’, ‘check’) → (‘normal’, ‘commit’)
[11/17/25 20:02:58] Transfer repo ‘644e2664’: (‘normal’, ‘commit’) → (‘normal’, ‘fs’)
[11/17/25 20:07:58] Bad response code for GET ``https://seafile.ocloud.de/seafhttp/repo/644e2664-cba3-408a-ae63-bbc084987794/fs-id-list/?server-head=dc2356ac58f26afd8cc9e0ec3279333736eb434e:`` 502.
[11/17/25 20:07:58] Failed to get fs id list for repo 644e2664 on server ``https://seafile.ocloud.de``.
[11/17/25 20:07:58] Transfer repo ‘644e2664’: (‘normal’, ‘fs’) → (‘error’, ‘finished’)
[11/17/25 20:07:58] Transition clone state for 644e2664 from [fetch] to [error]: Server error.
Could you say me more, how this error can occur?
Greetings,
Jonas
502 errors for fs-id-list requests are typical when you sync a library with large number of files. These requests can timeout on the server side. You have to ask your service provider to change Nginx and seafile.conf to support syncing large libraries, if they plan to support such cases.
I see - However my libraries are 1 TB of data (surely some big files), and I synced them with Seafile for many, many years now without ever having a problem - both on local network with a Rasperry Pi and a harddrive, as well as in the cloud. Seems strange that suddenly that is a thing…
Are there any new restrictions by Nginx now?
Also, is there any possibility to find out, which files cause the problem? For now, it would be the easiest to remove them for now, just to get everything syncing again.
This is not an issue on the client but on the server. The server you use may have short timeout settings in Nginx, or low values for these two seafile.conf options: max_sync_file_count, fs_id_list_request_timeout. You should ask your server administrator whether they can increase the limits.
Another way is to split your library into a few smaller libraries.