Server Storage NAS si mai Premium - Beelink ME Pro

[penguin]

Am făcut upgrade la NAS 

A durat vreo 3 luni să-mi ajungă sistemul, chinezii fiind si ei impacatati de criza asta de hardware.

=== Hardware ===

https://www.bee-link.com/products/beelink-me-pro

Este fenomental, build quality excellent, la nivel aproape de Apple.

Este proiectat cu cap, are un "sertar" jos cu placa de baza, iar deasupra 2x HDD-uri de 3.5"
Racirea se face de jos in sus, trage aer prin sertarul de jos, si îl sufla in sus pe HDD-uri.

HDD-urile au tray-uri, si cover-ul pt ele este un mesh magnetic

Cand va trebui să-l curat de praf va fi foarte ușor, scot sertarul, îl suflu cu aer si job done.

Este foarte silent, temperatura la CPU nu sare de 50 grade, HDD-urile nu sar de 45 de grade (pe sistemul vechi postat aici, stăteau in 55....60 grade)

Si "za pis of zeristans" "sertarul" cu placa de baza are 3 NVME-uri si 2 eth-uri, unu de 2.5Gbps si unul de 5Gbps.

Pt HDD-uri, rulez aceleași HDD-uri rulate 24/7 de 7 ani, WD Gold, nu am prins deal bun înainte sa explodeze prețurile de 5 ori.

=== Software ===

Rulez același stack create manual, cu toate serviciile in docker.

Docker este manageriat cu Komodo  https://github.com/moghtech/komodo , arę toate funcțiile, si totul este IaaC infrasctructure as a code.

Storage-ul este monitorizat SMART cu https://github.com/analogj/scrutiny

Monitorizarea o fac cu https://github.com/henrygd/beszel , containere, host, hardware

Transcoding, este făcut cu docker rootless, 4K 80GB este transcodat fără efort, si mai am loc de 3..4 stream-uri daca este cazul 

Bluetooth receiver pt home assistant https://github.com/home-assistant a trebuit sa hack-uiesc niste librarii sa-l fac sa meargă, dar merge stabil, cu 4 sensori prin casa.

Agentic AI - momentan testez diverse solutii gen OpenClaw, Hermes si ZeroClaw. ZeroClaw este varianta mea preferata momentan, si rulez si N8N pt AI process automation. LLM-urile le rulez in alta parte in LAN si din Cloud.

Pt poze folosesc Immich cu AI pt face recognition, si alte funcții simple, rulate prin procesor, si merg foarte bine.

Am varianta cu 16GB RAM si intel N150, iar stack-ul pe care-l am (ce este sus plus muuulte altele) consuma 5...6GB RAM, totul fiind super optimizat cu linux cgroups, OOM si puțin kernel tuning

====

Daca are cineva nevoie de ajutor sa facă set up software let me know

[MusicON]

Am urmarit review-ul produsului la NasCompares cand a aparut si parea interesant.

Nu imi place ca are doar 2 HDD-uri si exista si alternative mai bune.

La aceeasi bani, 559 $ se poate cumpara AOOSTAR WTR MAX INTEL 1235U https://aoostar.com/collections/mini-gaming-pc/products/aoostar-wtr-max-intel-1235u

6 bays + 5 nvme, procesor mai puternic si eficient.

//

Eu am un DXP 4800 Pro cu OS nativ, UGOS.

Ce OS folosesti ?

 

[penguin]

wow, nu am stiut de acest AOOSTAR, este mult mai performant intradevar 
si DXP 4800 Pro este fain

eu folosesc Debian

[MusicON]

On 5/27/2026 at 11:41 PM, penguin said:

Daca are cineva nevoie de ajutor sa facă set up software let me know

Ajuta-ma pe mine te rog frumos cu Immich.

I-am dat deploy prin portainer, functioneaza (se pare).

Am accesat aplicatia atat local cat si remote cu tailscale, am facut back up la zeci de GB de poze/filmari doar ca 2 containere, mai exact postgres si vallkey apar eronate in docker local images.

Am tot sters, re isntalat si nimic.

Am gasit pe net si la altii problema mea dar nimeni nu ii da de cap.

Repet, aplicatia (pare ca?) functioneaza.

 

Edit:

Acel Aoostar s-a lansat dupa ce am cumparat eu DXP-ul meu.

Parerea mea este ca merita banii la ce ofera doar ca nu sunt sigur cat de fiabil este.

 

Edit 2

Nu ma lasa sa sterg cele 2 cu eroare, trebuie sa sterg toata aplicatia.

[penguin]

=== Immich ===

 

services:
  
  immich-server:
    container_name: immich_server
    image: ghcr.io/immich-app/immich-server:${IMMICH_VERSION:-release}  
    devices:
      - /dev/dri:/dev/dri
    volumes:
      - ${UPLOAD_LOCATION}:/usr/src/app/upload # Change Storage location in ENV not here
      - /etc/localtime:/etc/localtime:ro
    env_file:
      - compose.env
    ports:
      - '2283:2283'
    depends_on:
      - immich-redis
      - immich-database
    restart: unless-stopped
    healthcheck:
      disable: false
    cpuset: "1,2,3"
    mem_limit: 2g
    memswap_limit: 2g

  immich-machine-learning:
    container_name: immich_machine_learning
    # image: ghcr.io/immich-app/immich-machine-learning:${IMMICH_VERSION:-release}
    image: ghcr.io/immich-app/immich-machine-learning:release-openvino
    devices:
      - /dev/dri:/dev/dri
    volumes:
      - /immich/machine-learning:/cache
    env_file:
      - compose.env
    restart: unless-stopped
    healthcheck:
      disable: false
    cpuset: "2,3"
    mem_limit: 2g
    memswap_limit: 2g

  immich-redis:
    container_name: immich_redis
    image: docker.io/valkey/valkey:8-bookworm@sha256:fec42f399876eb6faf9e008570597741c87ff7662a54185593e74b09ce83d177
    healthcheck:
      test: redis-cli ping || exit 1
    restart: unless-stopped
    cpuset: "2,3"
    mem_limit: 300m
    memswap_limit: 300m

  immich-database:
    container_name: immich_postgres
    image: ghcr.io/immich-app/postgres:14-vectorchord0.4.3-pgvectors0.2.0
    environment:
      POSTGRES_PASSWORD: ${DB_PASSWORD}
      POSTGRES_USER: ${DB_USERNAME}
      POSTGRES_DB: ${DB_DATABASE_NAME}
      POSTGRES_INITDB_ARGS: '--data-checksums'
      DB_STORAGE_TYPE: 'SSD'
    volumes:
      - ${DB_DATA_LOCATION}:/var/lib/postgresql/data # Change DB storage location in ENV not here
    restart: unless-stopped
    cpuset: "2,3"
    mem_limit: 1g
    memswap_limit: 1g

 

# You can find documentation for all the supported env variables at https://immich.app/docs/install/environment-variables
# Or here https://docs.immich.app/install/environment-variables/$0

# Available ML models from Immich
# https://huggingface.co/immich-app/collections$0
# Extra models not in there: 
# CLIP ML Models: ViT-B-32__openai, 

# Number of worker processes to spawn
MACHINE_LEARNING_WORKERS=1

# Thread count of the request thread pool
MACHINE_LEARNING_REQUEST_THREADS=4

# Number of parallel model operations
MACHINE_LEARNING_MODEL_INTER_OP_THREADS=4

# Number of threads for each model operation
MACHINE_LEARNING_MODEL_INTRA_OP_THREADS=1

# FP16 of F32 for full precision
MACHINE_LEARNING_OPENVINO_PRECISION=FP16

# Load model in memory
MACHINE_LEARNING_MODEL_ARENA=true

#Uunresponsiveness time before a worker is killed
MACHINE_LEARNING_WORKER_TIMEOUT=300

# Test Immich ML OpenVINO 
# python3 -c "import onnxruntime as ort; print(ort.get_available_providers())"
# ['OpenVINOExecutionProvider', 'CPUExecutionProvider']

# List of comma-separated IPs set as trusted proxies / asta daca ai proxy , nginx proxy manager etc
IMMICH_TRUSTED_PROXIES=172.27.0.1

# The location where your uploaded files are stored
UPLOAD_LOCATION=/hostpath/immich_library

# The location where your database files are stored. Network shares are not supported for the database
DB_DATA_LOCATION=/hostpath/immich/database

# To set a timezone, uncomment the next line and change Etc/UTC to a TZ identifier from this list: https://en.wikipedia.org/wiki/List_of_tz_database_time_zones#List
# TZ=Etc/UTC

# The Immich version to use. You can pin this to a specific version like "v1.71.0"
# IMMICH_VERSION=v2.6.3 
IMMICH_VERSION=release

# Connection secret for postgres. You should change it to a random password
# Please use only the characters `A-Za-z0-9`, without special characters or spaces
DB_PASSWORD=paroladb

# The values below this line do not need to be changed
###################################################################################
DB_USERNAME=dbusername
DB_DATABASE_NAME=immich
DB_DATA_LOCATION=/hostpath/immich/database

DB_HOSTNAME=immich-database
REDIS_HOSTNAME=immich-redis



==== 

Merita banii, Aoostar 
Eu am ales beeline pt ca este extrem de mic