[BUG] Extremely slow UI, and poor download performance

[Toudahl]

Is there an existing issue for this?

• I have searched the existing issues

Current Behavior

Upon initial install, the UI loads quickly. But over time the UI slows down quite a lot.
The download speeds seems to be affected too.

It is currently taking 1½ minutes to load the UI, with calls to the rpc endpoints taking the longest.

Expected Behavior

The performance should continue on the same level, unless network is degraded or server is pressed on resources.

Steps To Reproduce

1. Create a user/group: 1063:65556
2. Run config.sh

• Note that this requires two nfs volumes, one for download target, and one for backup of local config done by a rsync i have setup. Adjust dns/ip as needed.

3. Create service using the supplied docker compose.
4. Add a number of torrents. I have currently got 94 torrents, but I started noticing the problem earlier.

Environment

- OS: debian trixie, no desktop environment
- How docker service was installed: 
`curl -fsSL https://get.docker.com -o install-docker.sh && sudo sh install-docker.sh`
- Hardware. Intel N250, M2 SSD disk with 16 gb ram

Write performance to nfs mount:

Resource utilization on host:

CPU architecture

x86-64

Docker creation

services:
  transmission:
    image: linuxserver/transmission:4.0.6
    container_name: arr_transmission
    hostname: transmission
    restart: unless-stopped
    networks:
      - network
    user: '1063:65556'
    environment:
      - PEERPORT=65534 #optional
    env_file:
      - .env
      - .timezone.env
    volumes:
      - /var/lib/transmission-local/config:/config
      - /srv/nas/media/:/media-files #optional
    devices:
        - /dev/dri:/dev/dri
    ports:
      - 9091:9091
      - 65534:65534
      - 65534:65534/udp

networks:
  network:

Container logs

/package/admin/s6-overlay/libexec/preinit: info: /run belongs to uid 0 instead of 1063 - fixing it
╔═════════════════════════════════════════════════════════════════════════╗
║                                                                         ║
║           You are running this container as a non-root user:            ║
║   UMASK, custom services, & docker mod functionality will be disabled   ║
║      and the PUID/PGID environment variables will have no effect.       ║
║                                                                         ║
╚═════════════════════════════════════════════════════════════════════════╝
[migrations] started
[migrations] no migrations found
───────────────────────────────────────

      ██╗     ███████╗██╗ ██████╗
      ██║     ██╔════╝██║██╔═══██╗
      ██║     ███████╗██║██║   ██║
      ██║     ╚════██║██║██║   ██║
      ███████╗███████║██║╚██████╔╝
      ╚══════╝╚══════╝╚═╝ ╚═════╝

   Brought to you by linuxserver.io
───────────────────────────────────────

To support LSIO projects visit:
https://www.linuxserver.io/donate/

───────────────────────────────────────
GID/UID
───────────────────────────────────────

User UID:    1063
User GID:    65556
───────────────────────────────────────
Linuxserver.io version: 4.0.6-r4-ls310
Build-date: 2025-09-16T10:38:24+00:00
───────────────────────────────────────

[custom-init] No custom files found, skipping...
Connection to localhost (127.0.0.1) 9091 port [tcp/*] succeeded!
[ls.io-init] done.
[2025-09-30 17:55:23.118] ERR watchdir-inotify.cc:94 Couldn't watch '/watch': No such file or directory (2) (watchdir-inotify.cc:94)

[github-actions]

Thanks for opening your first issue here! Be sure to follow the relevant issue templates, or risk having this issue marked as invalid.

[Toudahl]

One note. The server is also running other services. E.g immich, which suffers no performance degradation at the same time.

Starting/stopping and recreating the transmission container has no effect on its performance.
However, rebooting the server seems to have improved the performance issue, at least for now.

Download speeds are as i would expect, but UI is still slow to load. Much slower than any of the other services running on the same server.

[Toudahl]

I have tracked down the problem.

I have a number of nfs mounts, and the default for this is to have 1 connection per host.
Due to the implementation of nfs clients, it will only have one connection for each host/version.

The solution to this is to add the option nconnect to each mount point in fstab.