External Storage Full Resync

Hello,

I am using Nextcloud on Linux and I am attaching to numerous external storage locations which are mounted on Linux from SMB shares presented from an Unraid storage server.

One of my shares is a multi-TB photos share and any time I do maintenance to my Unraid box and the share / external storage location disappears to nextcloud, when it comes back online all of my sync clients have to do a full re-sync which takes many many hours due to the size.

What is causing this behavior? Nothing has changed on the filesystem but the sync clients re-download nearly every single file.

Is there anything I can do to fix this?

You can run a scan via occ to confirm there were no files changes since your last unraid maintenance. You can incorporate it into your workflow while nextcloud is in maintenance mode.

https://docs.nextcloud.com/server/latest/admin_manual/configuration_server/occ_command.html#file-operations-label

Could you clarify how exactly you have mounted it?

Mounted as CIFS with this in the fstab:

cifs    auto,users,noperm,uid=werkkrew,gid=media,vers=3.0,credentials=/etc/samba/werkkrew.creds,x-systemd.automount 0 0

I am experiencing the same issue. I had to remount my external storage on my NC server and now my desktop client is resyncing everything (downloading from server to client) even though no changes have been made to the files (and I checked all files are still present in the remote/external storage and on the client machine).

I am on NC 25.0.1

I also have this issue if I do occ:scan --all, all client resync EVERYTHING (download from server to client) even if no changes to the files have been made. Is this the expected behaviour?

Hello,

I’d like to understand if those issues are reproducible / expected NC behaviour:

  • when you unmount external storage from the server (for long enough for the client to detect it) and then remount it, then it triggers a full resync (and download) from server to client even with no actual file changes on either ends

  • if you execute the a occ:scan --all that also triggers a full resync (and download) of files even with no actual file changes on either ends

From there we can at least figure out if this is a bug / issue or expected behaviour which we might ask for a feature request to change it.

Many thanks