immich-app/immich
1
0
Fork 0
mirror of https://github.com/immich-app/immich.git synced 2026-07-25 14:00:45 +03:00
Code Issues 783 Packages Projects Releases 22 Wiki Activity

People faces on ext library not detected despite running "missing" #7595

New Issue
Closed
opened 2026-02-05 13:10:15 +03:00 by OVERLORD · 3 comments
OVERLORD commented 2026-02-05 13:10:15 +03:00
Owner
Copy Link

Originally created by @rsurgiewicz on GitHub (Oct 21, 2025).

I have searched the existing issues, both open and closed, to make sure this is not a duplicate report.

  • Yes

The bug

Hi Folks,

I've been running Immich for some time (v2.0.1) updated today to v2.1.0 but issue still persists.

I have external library that is being populated overnight and I've turned on periodical scans on external library - Immich fetch new photos and videos and display correctly ✅

However newly added photos (from ext library) do not have people faces detected nor recognized.
When I run task manually

Image

Photos are being processed but still no faces is "attached" to new photos.

The only way to "attach" faces is to run reset and re-run face detection & recognition for whole set. That takes ages since my library is quite a large set.

I consider it a bug.

The OS that Immich Server is running on

TrueNas

Version of Immich Server

v.2.1.0

Version of Immich Mobile App

v2.1.0

Platform with the issue

  • Server
  • Web
  • Mobile

Device make and model

No response

Your docker-compose.yml content

configs:
  permissions_actions_data:
    content: >-
      [{"read_only": false, "mount_path":
      "/mnt/permission/pgvecto_postgres_data", "is_temporary": false,
      "identifier": "pgvecto_postgres_data", "recursive": false, "mode":
      "check", "uid": 999, "gid": 999, "chmod": null}, {"read_only": false,
      "mount_path": "/mnt/permission/redis_redis_data", "is_temporary": true,
      "identifier": "redis_redis_data", "recursive": true, "mode": "check",
      "uid": 568, "gid": 568, "chmod": null}, {"read_only": false, "mount_path":
      "/mnt/permission/ml-cache", "is_temporary": true, "identifier":
      "ml-cache", "recursive": true, "mode": "check", "uid": 568, "gid": 568,
      "chmod": null}]
  permissions_run_script:
    content: |
      #!/usr/bin/env python3

      import os
      import json
      import time
      import shutil

      with open("/script/actions.json", "r") as f:
          actions_data = json.load(f)

      if not actions_data:
          # If this script is called, there should be actions data
          raise ValueError("No actions data found")

      def fix_perms(path, chmod, recursive=False):
          print(f"Changing permissions{' recursively ' if recursive else ' '}to {chmod} on: [{path}]")
          os.chmod(path, int(chmod, 8))
          if recursive:
              for root, dirs, files in os.walk(path):
                  for f in files:
                      os.chmod(os.path.join(root, f), int(chmod, 8))
          print("Permissions after changes:")
          print_chmod_stat()

      def fix_owner(path, uid, gid, recursive=False):
          print(f"Changing ownership{' recursively ' if recursive else ' '}to {uid}:{gid} on: [{path}]")
          os.chown(path, uid, gid)
          if recursive:
              for root, dirs, files in os.walk(path):
                  for f in files:
                      os.chown(os.path.join(root, f), uid, gid)
          print("Ownership after changes:")
          print_chown_stat()

      def print_chown_stat():
          curr_stat = os.stat(action["mount_path"])
          print(f"Ownership: [{curr_stat.st_uid}:{curr_stat.st_gid}]")

      def print_chmod_stat():
          curr_stat = os.stat(action["mount_path"])
          print(f"Permissions: [{oct(curr_stat.st_mode)[3:]}]")

      def print_chown_diff(curr_stat, uid, gid):
          print(f"Ownership: wanted [{uid}:{gid}], got [{curr_stat.st_uid}:{curr_stat.st_gid}].")

      def print_chmod_diff(curr_stat, mode):
          print(f"Permissions: wanted [{mode}], got [{oct(curr_stat.st_mode)[3:]}].")

      def perform_action(action):
          if action["read_only"]:
              print(f"Path for action [{action['identifier']}] is read-only, skipping...")
              return

          start_time = time.time()
          print(f"=== Applying configuration on volume with identifier [{action['identifier']}] ===")

          if not os.path.isdir(action["mount_path"]):
              print(f"Path [{action['mount_path']}] is not a directory, skipping...")
              return

          if action["is_temporary"]:
              print(f"Path [{action['mount_path']}] is a temporary directory, ensuring it is empty...")
              for item in os.listdir(action["mount_path"]):
                  item_path = os.path.join(action["mount_path"], item)

                  # Exclude the safe directory, where we can use to mount files temporarily
                  if os.path.basename(item_path) == "ix-safe":
                      continue
                  if os.path.isdir(item_path):
                      shutil.rmtree(item_path)
                  else:
                      os.remove(item_path)

          if not action["is_temporary"] and os.listdir(action["mount_path"]):
              print(f"Path [{action['mount_path']}] is not empty, skipping...")
              return

          print(f"Current Ownership and Permissions on [{action['mount_path']}]:")
          curr_stat = os.stat(action["mount_path"])
          print_chown_diff(curr_stat, action["uid"], action["gid"])
          print_chmod_diff(curr_stat, action["chmod"])
          print("---")

          if action["mode"] == "always":
              fix_owner(action["mount_path"], action["uid"], action["gid"], action["recursive"])
              if not action["chmod"]:
                  print("Skipping permissions check, chmod is falsy")
              else:
                  fix_perms(action["mount_path"], action["chmod"], action["recursive"])
              return

          elif action["mode"] == "check":
              if curr_stat.st_uid != action["uid"] or curr_stat.st_gid != action["gid"]:
                  print("Ownership is incorrect. Fixing...")
                  fix_owner(action["mount_path"], action["uid"], action["gid"], action["recursive"])
              else:
                  print("Ownership is correct. Skipping...")

              if not action["chmod"]:
                  print("Skipping permissions check, chmod is falsy")
              else:
                  if oct(curr_stat.st_mode)[3:] != action["chmod"]:
                      print("Permissions are incorrect. Fixing...")
                      fix_perms(action["mount_path"], action["chmod"], action["recursive"])
                  else:
                      print("Permissions are correct. Skipping...")

          print(f"Time taken: {(time.time() - start_time) * 1000:.2f}ms")
          print(f"=== Finished applying configuration on volume with identifier [{action['identifier']}] ==")
          print()

      if __name__ == "__main__":
          start_time = time.time()
          for action in actions_data:
              perform_action(action)
          print(f"Total time taken: {(time.time() - start_time) * 1000:.2f}ms")
services:
  machine-learning:
    cap_drop:
      - ALL
    depends_on:
      permissions:
        condition: service_completed_successfully
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 4096M
    environment:
      GID: '568'
      GROUP_ID: '568'
      HF_HOME: /mlcache
      IMMICH_LOG_LEVEL: log
      IMMICH_PORT: '32002'
      MACHINE_LEARNING_CACHE_FOLDER: /mlcache
      MPLCONFIGDIR: /mlcache/matplotlib
      NODE_ENV: production
      NVIDIA_VISIBLE_DEVICES: void
      PGID: '568'
      PUID: '568'
      TRANSFORMERS_CACHE: /mlcache
      TZ: Europe/Warsaw
      UID: '568'
      UMASK: '002'
      UMASK_SET: '002'
      USER_ID: '568'
    group_add:
      - 568
    healthcheck:
      interval: 30s
      retries: 5
      start_interval: 2s
      start_period: 15s
      test: python3 healthcheck.py
      timeout: 5s
    image: ghcr.io/immich-app/immich-machine-learning:v2.1.0-rocm
    platform: linux/amd64
    privileged: False
    restart: unless-stopped
    security_opt:
      - no-new-privileges=true
    stdin_open: False
    tty: False
    user: '568:568'
    volumes:
      - read_only: False
        source: ml-cache
        target: /mlcache
        type: volume
        volume:
          nocopy: False
  permissions:
    cap_add:
      - CHOWN
      - DAC_OVERRIDE
      - FOWNER
    cap_drop:
      - ALL
    configs:
      - mode: 320
        source: permissions_actions_data
        target: /script/actions.json
      - mode: 448
        source: permissions_run_script
        target: /script/run.py
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 1024M
    entrypoint:
      - python3
      - /script/run.py
    environment:
      GID: '568'
      GROUP_ID: '568'
      NVIDIA_VISIBLE_DEVICES: void
      PGID: '568'
      PUID: '568'
      TZ: Europe/Warsaw
      UID: '568'
      UMASK: '002'
      UMASK_SET: '002'
      USER_ID: '568'
    group_add:
      - 568
    healthcheck:
      disable: True
    image: python:3.13.0-slim-bookworm
    network_mode: none
    platform: linux/amd64
    privileged: False
    restart: on-failure:1
    security_opt:
      - no-new-privileges=true
    stdin_open: False
    tty: False
    user: '0:0'
    volumes:
      - read_only: False
        source: ml-cache
        target: /mnt/permission/ml-cache
        type: volume
        volume:
          nocopy: False
      - bind:
          create_host_path: False
          propagation: rprivate
        read_only: False
        source: /mnt/naspool/apps/immich/db
        target: /mnt/permission/pgvecto_postgres_data
        type: bind
      - read_only: False
        source: redis-data
        target: /mnt/permission/redis_redis_data
        type: volume
        volume:
          nocopy: False
  pgvecto:
    cap_drop:
      - ALL
    depends_on:
      permissions:
        condition: service_completed_successfully
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 4096M
    environment:
      DB_STORAGE_TYPE: HDD
      GID: '568'
      GROUP_ID: '568'
      NVIDIA_VISIBLE_DEVICES: void
      PGID: '568'
      PGPORT: '5432'
      POSTGRES_DB: immich
      POSTGRES_PASSWORD: xxxxx!
      POSTGRES_USER: immich
      PUID: '568'
      TZ: Europe/Warsaw
      UID: '568'
      UMASK: '002'
      UMASK_SET: '002'
      USER_ID: '568'
    group_add:
      - 568
    healthcheck:
      interval: 30s
      retries: 5
      start_interval: 2s
      start_period: 15s
      test:
        - CMD
        - pg_isready
        - '-h'
        - 127.0.0.1
        - '-p'
        - '5432'
        - '-U'
        - immich
        - '-d'
        - immich
      timeout: 5s
    image: ghcr.io/immich-app/postgres:15-vectorchord0.4.3-pgvectors0.2.0
    platform: linux/amd64
    privileged: False
    restart: unless-stopped
    security_opt:
      - no-new-privileges=true
    shm_size: 256M
    stdin_open: False
    tty: False
    user: '999:999'
    volumes:
      - bind:
          create_host_path: False
          propagation: rprivate
        read_only: False
        source: /mnt/naspool/apps/immich/db
        target: /var/lib/postgresql/data
        type: bind
  redis:
    cap_drop:
      - ALL
    command:
      - '--port'
      - '6379'
      - '--requirepass'
      - xxxxx!
    depends_on:
      permissions:
        condition: service_completed_successfully
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 4096M
    environment:
      GID: '568'
      GROUP_ID: '568'
      NVIDIA_VISIBLE_DEVICES: void
      PGID: '568'
      PUID: '568'
      REDIS_PASSWORD: xxxxx!
      TZ: Europe/Warsaw
      UID: '568'
      UMASK: '002'
      UMASK_SET: '002'
      USER_ID: '568'
    group_add:
      - 568
    healthcheck:
      interval: 30s
      retries: 5
      start_interval: 2s
      start_period: 15s
      test:
        - CMD
        - redis-cli
        - '-h'
        - 127.0.0.1
        - '-p'
        - '6379'
        - '-a'
        - xxxxxx!
        - ping
      timeout: 5s
    image: valkey/valkey:8.1.4
    platform: linux/amd64
    privileged: False
    restart: unless-stopped
    security_opt:
      - no-new-privileges=true
    stdin_open: False
    tty: False
    user: '568:568'
    volumes:
      - read_only: False
        source: redis-data
        target: /data
        type: volume
        volume:
          nocopy: False
  server:
    cap_drop:
      - ALL
    depends_on:
      permissions:
        condition: service_completed_successfully
      pgvecto:
        condition: service_healthy
      redis:
        condition: service_healthy
    deploy:
      resources:
        limits:
          cpus: '2'
          memory: 4096M
    environment:
      DB_DATABASE_NAME: immich
      DB_HOSTNAME: pgvecto
      DB_PASSWORD: xxxxxx!
      DB_PORT: '5432'
      DB_USERNAME: immich
      GID: '568'
      GROUP_ID: '568'
      IMMICH_LOG_LEVEL: log
      IMMICH_MACHINE_LEARNING_ENABLED: 'true'
      IMMICH_MACHINE_LEARNING_URL: http://machine-learning:32002
      IMMICH_PORT: '30041'
      NODE_ENV: production
      NVIDIA_VISIBLE_DEVICES: void
      PGID: '568'
      PUID: '568'
      REDIS_DBINDEX: '0'
      REDIS_HOSTNAME: redis
      REDIS_PASSWORD: xxxxxx!
      REDIS_PORT: '6379'
      TZ: Europe/Warsaw
      UID: '568'
      UMASK: '002'
      UMASK_SET: '002'
      USER_ID: '568'
    group_add:
      - 568
    healthcheck:
      interval: 30s
      retries: 5
      start_interval: 2s
      start_period: 15s
      test: immich-healthcheck
      timeout: 5s
    image: ghcr.io/immich-app/immich-server:v2.1.0
    platform: linux/amd64
    ports:
      - mode: ingress
        protocol: tcp
        published: 30041
        target: 30041
    privileged: False
    restart: unless-stopped
    security_opt:
      - no-new-privileges=true
    stdin_open: False
    tty: False
    user: '568:568'
    volumes:
      - bind:
          create_host_path: False
          propagation: rprivate
        read_only: False
        source: /mnt/naspool/apps/immich/upload
        target: /data
        type: bind
      - bind:
          create_host_path: False
          propagation: rprivate
        read_only: True
        source: /mnt/naspool/backup/DropBox/rafals_dropbox/Przesłane z aparatu
        target: /extdata/rafals_dropbox
        type: bind
volumes:
  ml-cache: {}
  redis-data: {}
x-notes: >
  # Immich


  ## Security


  **Read the following security precautions to ensure that you wish to continue
  using this application.**


  ---


  ### Container: [machine-learning]


  #### Running user/group(s)


  - User: 568

  - Group: 568

  - Supplementary Groups: apps


  ---


  ### Container: [pgvecto]


  #### Running user/group(s)


  - User: 999

  - Group: 999

  - Supplementary Groups: apps


  ---


  ### Container: [redis]


  #### Running user/group(s)


  - User: 568

  - Group: 568

  - Supplementary Groups: apps


  ---


  ### Container: [server]


  #### Running user/group(s)


  - User: 568

  - Group: 568

  - Supplementary Groups: apps


  ---


  ### Container: [permissions]


  **This container is short-lived.**


  #### Running user/group(s)


  - User: root

  - Group: root

  - Supplementary Groups: apps


  ---


  ## Bug Reports and Feature Requests


  If you find a bug in this app or have an idea for a new feature, please file
  an issue at

  https://github.com/truenas/apps
x-portals:
  - host: 0.0.0.0
    name: Web UI
    path: /
    port: 30041
    scheme: http

Your .env content

1=1

Reproduction steps

...

Relevant log output

Additional information

No response

Originally created by @rsurgiewicz on GitHub (Oct 21, 2025). ### I have searched the existing issues, both open and closed, to make sure this is not a duplicate report. - [x] Yes ### The bug Hi Folks, I've been running Immich for some time (v2.0.1) updated today to v2.1.0 but issue still persists. I have external library that is being populated overnight and I've turned on periodical scans on external library - Immich fetch new photos and videos and display correctly ✅ However newly added photos (from ext library) do not have people faces detected nor recognized. When I run task manually <img width="1868" height="1202" alt="Image" src="https://github.com/user-attachments/assets/d5c9ac63-2b6e-49ff-a9a3-c2c66bd3b559" /> Photos are being processed but still no faces is "attached" to new photos. The only way to "attach" faces is to run reset and re-run face detection & recognition for whole set. That takes ages since my library is quite a large set. I consider it a bug. ### The OS that Immich Server is running on TrueNas ### Version of Immich Server v.2.1.0 ### Version of Immich Mobile App v2.1.0 ### Platform with the issue - [x] Server - [x] Web - [x] Mobile ### Device make and model _No response_ ### Your docker-compose.yml content ```YAML configs: permissions_actions_data: content: >- [{"read_only": false, "mount_path": "/mnt/permission/pgvecto_postgres_data", "is_temporary": false, "identifier": "pgvecto_postgres_data", "recursive": false, "mode": "check", "uid": 999, "gid": 999, "chmod": null}, {"read_only": false, "mount_path": "/mnt/permission/redis_redis_data", "is_temporary": true, "identifier": "redis_redis_data", "recursive": true, "mode": "check", "uid": 568, "gid": 568, "chmod": null}, {"read_only": false, "mount_path": "/mnt/permission/ml-cache", "is_temporary": true, "identifier": "ml-cache", "recursive": true, "mode": "check", "uid": 568, "gid": 568, "chmod": null}] permissions_run_script: content: | #!/usr/bin/env python3 import os import json import time import shutil with open("/script/actions.json", "r") as f: actions_data = json.load(f) if not actions_data: # If this script is called, there should be actions data raise ValueError("No actions data found") def fix_perms(path, chmod, recursive=False): print(f"Changing permissions{' recursively ' if recursive else ' '}to {chmod} on: [{path}]") os.chmod(path, int(chmod, 8)) if recursive: for root, dirs, files in os.walk(path): for f in files: os.chmod(os.path.join(root, f), int(chmod, 8)) print("Permissions after changes:") print_chmod_stat() def fix_owner(path, uid, gid, recursive=False): print(f"Changing ownership{' recursively ' if recursive else ' '}to {uid}:{gid} on: [{path}]") os.chown(path, uid, gid) if recursive: for root, dirs, files in os.walk(path): for f in files: os.chown(os.path.join(root, f), uid, gid) print("Ownership after changes:") print_chown_stat() def print_chown_stat(): curr_stat = os.stat(action["mount_path"]) print(f"Ownership: [{curr_stat.st_uid}:{curr_stat.st_gid}]") def print_chmod_stat(): curr_stat = os.stat(action["mount_path"]) print(f"Permissions: [{oct(curr_stat.st_mode)[3:]}]") def print_chown_diff(curr_stat, uid, gid): print(f"Ownership: wanted [{uid}:{gid}], got [{curr_stat.st_uid}:{curr_stat.st_gid}].") def print_chmod_diff(curr_stat, mode): print(f"Permissions: wanted [{mode}], got [{oct(curr_stat.st_mode)[3:]}].") def perform_action(action): if action["read_only"]: print(f"Path for action [{action['identifier']}] is read-only, skipping...") return start_time = time.time() print(f"=== Applying configuration on volume with identifier [{action['identifier']}] ===") if not os.path.isdir(action["mount_path"]): print(f"Path [{action['mount_path']}] is not a directory, skipping...") return if action["is_temporary"]: print(f"Path [{action['mount_path']}] is a temporary directory, ensuring it is empty...") for item in os.listdir(action["mount_path"]): item_path = os.path.join(action["mount_path"], item) # Exclude the safe directory, where we can use to mount files temporarily if os.path.basename(item_path) == "ix-safe": continue if os.path.isdir(item_path): shutil.rmtree(item_path) else: os.remove(item_path) if not action["is_temporary"] and os.listdir(action["mount_path"]): print(f"Path [{action['mount_path']}] is not empty, skipping...") return print(f"Current Ownership and Permissions on [{action['mount_path']}]:") curr_stat = os.stat(action["mount_path"]) print_chown_diff(curr_stat, action["uid"], action["gid"]) print_chmod_diff(curr_stat, action["chmod"]) print("---") if action["mode"] == "always": fix_owner(action["mount_path"], action["uid"], action["gid"], action["recursive"]) if not action["chmod"]: print("Skipping permissions check, chmod is falsy") else: fix_perms(action["mount_path"], action["chmod"], action["recursive"]) return elif action["mode"] == "check": if curr_stat.st_uid != action["uid"] or curr_stat.st_gid != action["gid"]: print("Ownership is incorrect. Fixing...") fix_owner(action["mount_path"], action["uid"], action["gid"], action["recursive"]) else: print("Ownership is correct. Skipping...") if not action["chmod"]: print("Skipping permissions check, chmod is falsy") else: if oct(curr_stat.st_mode)[3:] != action["chmod"]: print("Permissions are incorrect. Fixing...") fix_perms(action["mount_path"], action["chmod"], action["recursive"]) else: print("Permissions are correct. Skipping...") print(f"Time taken: {(time.time() - start_time) * 1000:.2f}ms") print(f"=== Finished applying configuration on volume with identifier [{action['identifier']}] ==") print() if __name__ == "__main__": start_time = time.time() for action in actions_data: perform_action(action) print(f"Total time taken: {(time.time() - start_time) * 1000:.2f}ms") services: machine-learning: cap_drop: - ALL depends_on: permissions: condition: service_completed_successfully deploy: resources: limits: cpus: '2' memory: 4096M environment: GID: '568' GROUP_ID: '568' HF_HOME: /mlcache IMMICH_LOG_LEVEL: log IMMICH_PORT: '32002' MACHINE_LEARNING_CACHE_FOLDER: /mlcache MPLCONFIGDIR: /mlcache/matplotlib NODE_ENV: production NVIDIA_VISIBLE_DEVICES: void PGID: '568' PUID: '568' TRANSFORMERS_CACHE: /mlcache TZ: Europe/Warsaw UID: '568' UMASK: '002' UMASK_SET: '002' USER_ID: '568' group_add: - 568 healthcheck: interval: 30s retries: 5 start_interval: 2s start_period: 15s test: python3 healthcheck.py timeout: 5s image: ghcr.io/immich-app/immich-machine-learning:v2.1.0-rocm platform: linux/amd64 privileged: False restart: unless-stopped security_opt: - no-new-privileges=true stdin_open: False tty: False user: '568:568' volumes: - read_only: False source: ml-cache target: /mlcache type: volume volume: nocopy: False permissions: cap_add: - CHOWN - DAC_OVERRIDE - FOWNER cap_drop: - ALL configs: - mode: 320 source: permissions_actions_data target: /script/actions.json - mode: 448 source: permissions_run_script target: /script/run.py deploy: resources: limits: cpus: '2' memory: 1024M entrypoint: - python3 - /script/run.py environment: GID: '568' GROUP_ID: '568' NVIDIA_VISIBLE_DEVICES: void PGID: '568' PUID: '568' TZ: Europe/Warsaw UID: '568' UMASK: '002' UMASK_SET: '002' USER_ID: '568' group_add: - 568 healthcheck: disable: True image: python:3.13.0-slim-bookworm network_mode: none platform: linux/amd64 privileged: False restart: on-failure:1 security_opt: - no-new-privileges=true stdin_open: False tty: False user: '0:0' volumes: - read_only: False source: ml-cache target: /mnt/permission/ml-cache type: volume volume: nocopy: False - bind: create_host_path: False propagation: rprivate read_only: False source: /mnt/naspool/apps/immich/db target: /mnt/permission/pgvecto_postgres_data type: bind - read_only: False source: redis-data target: /mnt/permission/redis_redis_data type: volume volume: nocopy: False pgvecto: cap_drop: - ALL depends_on: permissions: condition: service_completed_successfully deploy: resources: limits: cpus: '2' memory: 4096M environment: DB_STORAGE_TYPE: HDD GID: '568' GROUP_ID: '568' NVIDIA_VISIBLE_DEVICES: void PGID: '568' PGPORT: '5432' POSTGRES_DB: immich POSTGRES_PASSWORD: xxxxx! POSTGRES_USER: immich PUID: '568' TZ: Europe/Warsaw UID: '568' UMASK: '002' UMASK_SET: '002' USER_ID: '568' group_add: - 568 healthcheck: interval: 30s retries: 5 start_interval: 2s start_period: 15s test: - CMD - pg_isready - '-h' - 127.0.0.1 - '-p' - '5432' - '-U' - immich - '-d' - immich timeout: 5s image: ghcr.io/immich-app/postgres:15-vectorchord0.4.3-pgvectors0.2.0 platform: linux/amd64 privileged: False restart: unless-stopped security_opt: - no-new-privileges=true shm_size: 256M stdin_open: False tty: False user: '999:999' volumes: - bind: create_host_path: False propagation: rprivate read_only: False source: /mnt/naspool/apps/immich/db target: /var/lib/postgresql/data type: bind redis: cap_drop: - ALL command: - '--port' - '6379' - '--requirepass' - xxxxx! depends_on: permissions: condition: service_completed_successfully deploy: resources: limits: cpus: '2' memory: 4096M environment: GID: '568' GROUP_ID: '568' NVIDIA_VISIBLE_DEVICES: void PGID: '568' PUID: '568' REDIS_PASSWORD: xxxxx! TZ: Europe/Warsaw UID: '568' UMASK: '002' UMASK_SET: '002' USER_ID: '568' group_add: - 568 healthcheck: interval: 30s retries: 5 start_interval: 2s start_period: 15s test: - CMD - redis-cli - '-h' - 127.0.0.1 - '-p' - '6379' - '-a' - xxxxxx! - ping timeout: 5s image: valkey/valkey:8.1.4 platform: linux/amd64 privileged: False restart: unless-stopped security_opt: - no-new-privileges=true stdin_open: False tty: False user: '568:568' volumes: - read_only: False source: redis-data target: /data type: volume volume: nocopy: False server: cap_drop: - ALL depends_on: permissions: condition: service_completed_successfully pgvecto: condition: service_healthy redis: condition: service_healthy deploy: resources: limits: cpus: '2' memory: 4096M environment: DB_DATABASE_NAME: immich DB_HOSTNAME: pgvecto DB_PASSWORD: xxxxxx! DB_PORT: '5432' DB_USERNAME: immich GID: '568' GROUP_ID: '568' IMMICH_LOG_LEVEL: log IMMICH_MACHINE_LEARNING_ENABLED: 'true' IMMICH_MACHINE_LEARNING_URL: http://machine-learning:32002 IMMICH_PORT: '30041' NODE_ENV: production NVIDIA_VISIBLE_DEVICES: void PGID: '568' PUID: '568' REDIS_DBINDEX: '0' REDIS_HOSTNAME: redis REDIS_PASSWORD: xxxxxx! REDIS_PORT: '6379' TZ: Europe/Warsaw UID: '568' UMASK: '002' UMASK_SET: '002' USER_ID: '568' group_add: - 568 healthcheck: interval: 30s retries: 5 start_interval: 2s start_period: 15s test: immich-healthcheck timeout: 5s image: ghcr.io/immich-app/immich-server:v2.1.0 platform: linux/amd64 ports: - mode: ingress protocol: tcp published: 30041 target: 30041 privileged: False restart: unless-stopped security_opt: - no-new-privileges=true stdin_open: False tty: False user: '568:568' volumes: - bind: create_host_path: False propagation: rprivate read_only: False source: /mnt/naspool/apps/immich/upload target: /data type: bind - bind: create_host_path: False propagation: rprivate read_only: True source: /mnt/naspool/backup/DropBox/rafals_dropbox/Przesłane z aparatu target: /extdata/rafals_dropbox type: bind volumes: ml-cache: {} redis-data: {} x-notes: > # Immich ## Security **Read the following security precautions to ensure that you wish to continue using this application.** --- ### Container: [machine-learning] #### Running user/group(s) - User: 568 - Group: 568 - Supplementary Groups: apps --- ### Container: [pgvecto] #### Running user/group(s) - User: 999 - Group: 999 - Supplementary Groups: apps --- ### Container: [redis] #### Running user/group(s) - User: 568 - Group: 568 - Supplementary Groups: apps --- ### Container: [server] #### Running user/group(s) - User: 568 - Group: 568 - Supplementary Groups: apps --- ### Container: [permissions] **This container is short-lived.** #### Running user/group(s) - User: root - Group: root - Supplementary Groups: apps --- ## Bug Reports and Feature Requests If you find a bug in this app or have an idea for a new feature, please file an issue at https://github.com/truenas/apps x-portals: - host: 0.0.0.0 name: Web UI path: / port: 30041 scheme: http ``` ### Your .env content ```Shell 1=1 ``` ### Reproduction steps 1. 2. 3. ... ### Relevant log output ```shell ``` ### Additional information _No response_
OVERLORD commented 2026-02-05 13:10:17 +03:00
Author
Owner
Copy Link

@Anindya-Prithvi commented on GitHub (Oct 22, 2025):

Has the reset method worked or is it just a speculation?
Also, is it possible that you have uploaded the external library under a different user and are trying to get facial data from the non-owning user... This scenario is unsupported based on the current immich permissions model.

@Anindya-Prithvi commented on GitHub (Oct 22, 2025): Has the reset method worked or is it just a speculation? Also, is it possible that you have uploaded the external library under a different user and are trying to get facial data from the non-owning user... This scenario is unsupported based on the current immich permissions model.
OVERLORD commented 2026-02-05 13:10:18 +03:00
Author
Owner
Copy Link

@rsurgiewicz commented on GitHub (Oct 23, 2025):

I run a single user environment and yes - user has the access.

Regarding reset & re-run face detection & recognition: last time it worked.
I can try to run it once again (it would take a few hours to complete). But before I do that, please let me know whether I should log or note something?

@rsurgiewicz commented on GitHub (Oct 23, 2025): I run a single user environment and yes - user has the access. Regarding reset & re-run face detection & recognition: last time it worked. I can try to run it once again (it would take a few hours to complete). But before I do that, please let me know whether I should log or note something?
OVERLORD commented 2026-02-05 13:10:19 +03:00
Author
Owner
Copy Link

@rsurgiewicz commented on GitHub (Oct 28, 2025):

OK i've found out digging in machine-learning container that it was silently crashing with a crash:

[Nest] 7 - 10/28/2025, 9:51:27 AM LOG [Microservices:MachineLearningRepository] Machine learning server became healthy (http://machine-learning:32002/). [Nest] 7 - 10/28/2025, 9:51:54 AM WARN [Microservices:MachineLearningRepository] Machine learning request to "http://machine-learning:32002/" failed: fetch failed [Nest] 7 - 10/28/2025, 9:51:54 AM LOG [Microservices:MachineLearningRepository] Machine learning server became unhealthy (http://machine-learning:32002/). [Nest] 7 - 10/28/2025, 9:51:54 AM ERROR [Microservices:{"id":"6209b2e8-ef38-48a1-8eff-2cdafc40192b"}] Unable to run job handler (AssetDetectFaces): Error: Machine learning request '{"facial-recognition":{"detection":{"modelName":"buffalo_l","options":{"minScore":0.7}},"recognition":{"modelName":"buffalo_l"}}}' failed for all URLs Error: Machine learning request '{"facial-recognition":{"detection":{"modelName":"buffalo_l","options":{"minScore":0.7}},"recognition":{"modelName":"buffalo_l"}}}' failed for all URLs at MachineLearningRepository.predict (/usr/src/app/server/dist/repositories/machine-learning.repository.js:115:15) at async MachineLearningRepository.detectFaces (/usr/src/app/server/dist/repositories/machine-learning.repository.js:124:26) at async PersonService.handleDetectFaces (/usr/src/app/server/dist/services/person.service.js:243:52) at async JobService.onJobStart (/usr/src/app/server/dist/services/job.service.js:198:28) at async EventRepository.onEvent (/usr/src/app/server/dist/repositories/event.repository.js:126:13) at async /usr/src/app/server/node_modules/.pnpm/bullmq@5.61.0/node_modules/bullmq/dist/cjs/classes/worker.js:533:32 [Nest] 7 - 10/28/2025, 9:51:57 AM LOG [Microservices:MachineLearningRepository] Machine learning server became healthy (http://machine-learning:32002/).

so it looks like my GPU was not detected and container was crashing. I've switched to pure CPU and now it seems to work.

@rsurgiewicz commented on GitHub (Oct 28, 2025): OK i've found out digging in machine-learning container that it was silently crashing with a crash: `[Nest] 7 - 10/28/2025, 9:51:27 AM LOG [Microservices:MachineLearningRepository] Machine learning server became healthy (http://machine-learning:32002/). [Nest] 7 - 10/28/2025, 9:51:54 AM WARN [Microservices:MachineLearningRepository] Machine learning request to "http://machine-learning:32002/" failed: fetch failed [Nest] 7 - 10/28/2025, 9:51:54 AM LOG [Microservices:MachineLearningRepository] Machine learning server became unhealthy (http://machine-learning:32002/). [Nest] 7 - 10/28/2025, 9:51:54 AM ERROR [Microservices:{"id":"6209b2e8-ef38-48a1-8eff-2cdafc40192b"}] Unable to run job handler (AssetDetectFaces): Error: Machine learning request '{"facial-recognition":{"detection":{"modelName":"buffalo_l","options":{"minScore":0.7}},"recognition":{"modelName":"buffalo_l"}}}' failed for all URLs Error: Machine learning request '{"facial-recognition":{"detection":{"modelName":"buffalo_l","options":{"minScore":0.7}},"recognition":{"modelName":"buffalo_l"}}}' failed for all URLs at MachineLearningRepository.predict (/usr/src/app/server/dist/repositories/machine-learning.repository.js:115:15) at async MachineLearningRepository.detectFaces (/usr/src/app/server/dist/repositories/machine-learning.repository.js:124:26) at async PersonService.handleDetectFaces (/usr/src/app/server/dist/services/person.service.js:243:52) at async JobService.onJobStart (/usr/src/app/server/dist/services/job.service.js:198:28) at async EventRepository.onEvent (/usr/src/app/server/dist/repositories/event.repository.js:126:13) at async /usr/src/app/server/node_modules/.pnpm/bullmq@5.61.0/node_modules/bullmq/dist/cjs/classes/worker.js:533:32 [Nest] 7 - 10/28/2025, 9:51:57 AM LOG [Microservices:MachineLearningRepository] Machine learning server became healthy (http://machine-learning:32002/).` so it looks like my GPU was not detected and container was crashing. I've switched to pure CPU and now it seems to work.
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
accessibility
changelog:enhancement
changelog:security
changelog:skip
changelog:translation
cli
date-time
dependencies
documentation
external-library
format
good first issue
mobile-beta
mobile-beta
mobile-beta
needs-answer
nice to have
pull-request
Mirrored from GitHub Pull Request
 sharing
tech-debt
📱mobile
🖥️web
🗄️server
🧠machine-learning
No Label
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
OVERLORD
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: immich-app/immich#7595
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?