real time monitoring

[chenks]

trying to understand real time monitoring.

jellyfin is running in an unpriveliged ubuntu LXC in proxmox
as the LXC is unpriveliged it can't mount network shares directly, so i have the CIFS share mounted in the proxmox host, and that share is passed thru to the LXC.

manual library scans work fine.
i've seen messages saying that RTM doesn't work with CIFS/SMB etc, and to turn it off, which i have now done as i had previously had it turned on.

however, i also have a plex server running on the same proxmox host (also an unpriveliged Ubunutu LXC with the same CIFS share passed to it), and RTM does seem to work with that, and it picks up new additions almost straight away, whereas jellfin does not until the next scheduled library scan kicks in.

so if RTM doesn't work with CIFS/SMB shares, why does plex pick up the changes? (it's not being picked up by a plex scheduled scan as the RTM works so well there is no scheduled scan enabled).

as a workaround i've set the jellyin scheduled library scan to every 15 minutes (previously was 12 hours).

[TheDreadPirate]

Real-time monitoring CAN work over SMB if setup correctly, and it sounds like it is setup correctly on yours. NFS and rclone straight up cannot work with it.

https://jellyfin.org/docs/general/admini...monitoring

[chenks]

Real-time monitoring CAN work over SMB if setup correctly, and it sounds like it is setup correctly on yours.  NFS and rclone straight up cannot work with it.

https://jellyfin.org/docs/general/admini...monitoring

if it's set up correctly jellyin isn't playing nice, as it's not picking up changes until a scheduled scan.
whereas plex does.

i've checked the jellyfin log and not seeing anything relating to watcher errors.

edit - it is picking up newly added content outwith a scheduled scan, but it's not picking up content that has been deleted automatically, those only get removed from jellyfin when a scheduled scan is done (or a manually triggered scan).

[kandykarter]

I have my network shares mounted as CIFS in my fstab, and real-time monitoring works perfectly. I've had hiccups with certain libraries, but they've been fixed by editing my bind mounts in my docker compose, and I have no idea WHY that worked, but it did. Seems like inotify and cifs shares are total witchcraft.

[chenks]

I have my network shares mounted as CIFS in my fstab, and real-time monitoring works perfectly. I've had hiccups with certain libraries, but they've been fixed by editing my bind mounts in my docker compose, and I have no idea WHY that worked, but it did. Seems like inotify and cifs shares are total witchcraft.

it seems like the issue i'm seeing is if i delete something from the network share from my windows laptop (with the networe share mapped as a drive) then the real time monitoring doesn't notice the change and only picks it up from a scheduled scan.

i'm not sure why that would be the case though in that it depends where (or who) has deleted sometihng as to whether RTM would see the change or not.