After I upgraded from 13.0.21 to 13.0.25, I’ve been seeing some issues with file uploading.
When I try to upload any files the progress reaches 99% and then sits there for about 30-45 seconds until it finally “uploads”.
I have tried uploading small files (5-50 KB) too and same thing happens, 99% then waits for something for 30-45 seconds and then continues.
I am not really sure what I might have done while upgrading… I even tried downgrading to 13.0.21 as well but unfortunately that also did not fix the situation.
The Desktop Client seems to work fine with uploads.
I’ve looked through the configurations and nothing seems off, I haven’t changed a single setting since migration from 12 to 13 and it was working fine before. I would appreciate the help.
EDIT:
Something worth mentioning I am currently behind an external nginx reverse proxy. I moved away from Cloudflare because of the 100 MB upload limit (had issues on Android).
I also tried direct connection without the reverse proxy and using Caddy, same thing occurs
This delay during the final 1% of the upload suggests a timeout or a long-running process occurring during the finalization phase. Since Seafile 13, the golang file server is used by default for web and mobile uploads, which handles post-processing differently than the desktop client.
Since the desktop client uses a different sync protocol and is unaffected, the issue is likely specific to the web-based upload handler’s finalization steps.
Please check the following:
Server Logs: Inspect fileserver.log and seahub.log for any error or timeout messages that appear during the 30-45 second window.
Virus Scanning: Check your seafile.conf or seahub_settings.py for virus scanning settings like ENABLE_VIRUS_SCAN or ENABLE_UPLOAD_LINK_VIRUS_CHECK. If enabled, try disabling them temporarily to see if the delay persists. Issues with ClamAV or other scanners failing to respond can cause the upload process to hang at the end.
Indexing: Although it usually affects only larger files. You can check if there is a problem about your disk.
I have found this record in fileserver.log (seems to happen exactly after 30 seconds)
[WARNING] failed to send repo update event: Post “http:// notification-server:8083/events”: dial tcp 172.18.0.10:8083: i/o timeout
I did some more digging on my end and it turns out I had the port 8083 blocked on a firewall
Didn’t think it would be that… I was stubbornly looking in completely wrong places before. (I am not sure how it was working before but I am glad it was an easy fix)
