deploy erro #6655

Activity

1. zhuyifeiRuichuang

   added

   bugSomething isn't workingSomething isn't working

   on Jul 23, 2026
2. prd-hoang-doan commented on Jul 23, 2026

   

   prd-hoang-doan

   on Jul 23, 2026

   Contributor

   More actions

   Hi @zhuyifeiRuichuang ,
   I found your issue related to LOG_PATH in .env file, I think the variable is commented by default, so you just un-comment the variable like

   LOG_PATH=/root/.flowise/logs
   or
   LOG_PATH=/Users/<username>/.flowise/logs
   

   Then use --force-recreate to build docker compose again, it will work.
   By the way, you should update all environment in .env having value like your_ (e.g: DATABASE_PATH, BLOB_STORAGE_PATH)
3. nvroot commented on Jul 29, 2026

   

   nvroot

   on Jul 29, 2026

   Last edited by nvroot

   More actions

   @zhuyifeiRuichuang try this:

   x-default-healthcheck: &default-healthcheck
     interval: 120s
     timeout: 20s
     retries: 5
     start_period: 60s
     start_interval: 5s
   services:
     flowise:
       image: flowiseai/flowise:latest
       container_name: flowise
       restart: unless-stopped
       entrypoint: /bin/sh -c "sleep 3; flowise start"
       environment:
         - DATABASE_PATH=/home/node/.flowise
         - DATABASE_TYPE=postgres
         - DATABASE_PORT=5432
         - DATABASE_HOST=flowise-db
         - DATABASE_NAME=flowise
         - DATABASE_USER=flowise-user
         - DATABASE_PASSWORD=changeme # CHANGEME
         - DATABASE_SSL=false
         - PGSSLMODE=false
         - SECRETKEY_STORAGE_TYPE=local
         - SECRETKEY_PATH=/home/node/.flowise
         - FLOWISE_SECRETKEY_OVERWRITE=changeme # CHANGEME openssl rand -hex 32
         - DEBUG=false
         - LOG_LEVEL=info
         - LOG_PATH=/home/node/.flowise/logs
         - STORAGE_TYPE=local
         - BLOB_STORAGE_PATH=/home/node/.flowise/storage
         - NUMBER_OF_PROXIES=1
         - CORS_ALLOW_CREDENTIALS=false
         - CORS_ORIGINS=*
         - MCP_CORS_ORIGINS=*
         - IFRAME_ORIGINS=*
         - SHOW_COMMUNITY_NODES=true
         - DISABLE_FLOWISE_TELEMETRY=true
         - APP_URL=https://flowise.example.com # CHANGEME
         - PORT=3001
         - REDIS_HOST=flowise-redis
         - REDIS_PORT=6379
         - APIKEY_PATH=/home/node/.flowise
         - ENABLE_METRICS=false
       expose:
         - 3001
       networks:
         - flowise-network
         - proxy
       volumes:
         - flowise-data:/home/node
       depends_on:
         flowise-db:
           condition: service_healthy
         flowise-redis:
           condition: service_healthy
         flowise-qdrant:
           condition: service_healthy
       healthcheck:
         test: ["CMD-SHELL", "wget -qO- http://127.0.0.1:3001/api/v1/ping || exit 1"]
         <<: *default-healthcheck
   
     flowise-db:
       image: postgres:17-alpine
       restart: unless-stopped
       environment:
         - POSTGRES_USER=flowise-user
         - POSTGRES_PASSWORD=changeme # CHANGEME
         - POSTGRES_DB=flowise
       networks:
         - flowise-network
       volumes:
         - flowise-pg:/var/lib/postgresql/data
       healthcheck:
         test:
           - CMD-SHELL
           - "pg_isready -h localhost -U $${POSTGRES_USER} -d $${POSTGRES_DB}"
         <<: *default-healthcheck
   
     flowise-redis:
       image: redis:7
       restart: unless-stopped
       networks:
         - flowise-network
       volumes:
         - flowise-redis:/data
       healthcheck:
         test:
           - CMD-SHELL
           - "redis-cli -h localhost -p 6379 ping"
         <<: *default-healthcheck
   
     flowise-qdrant:
       image: qdrant/qdrant:latest
       restart: unless-stopped
       environment:
         - QDRANT__SERVICE__API_KEY=changemevector # CHANGEME , CONNECT IN UI
       volumes:
         - flowise-qdrant:/qdrant/storage
       expose:
         - 6333
       networks:
         - flowise-network
         - proxy
       healthcheck:
         test: ["CMD", "/qdrant/qdrant", "-V"]
         <<: *default-healthcheck
   
   networks:
     flowise-network:
     proxy:
       external: true
   volumes:
     flowise-data:
     flowise-pg:
     flowise-redis:
     flowise-qdrant:
   
4. alexandrosang21 commented on Jul 29, 2026

   

   alexandrosang21

   on Jul 29, 2026

   More actions

   This is fixed by #6031, which has been open since March. It is the same root cause as #6030.

   The chain, for anyone landing here from a search:

   1. docker/.env.example has LOG_PATH commented out by default.

   2. docker/docker-compose.yml passes it unconditionally as LOG_PATH=${LOG_PATH}, so an unset
      variable becomes an empty string inside the container rather than being absent.

   3. packages/server/src/utils/config.ts reads it with ??:

      dir: process.env.LOG_PATH ?? path.join(__dirname, '..', '..', 'logs')

      ?? only falls back on null and undefined. An empty string is not nullish, so dir
      stays "" instead of taking the default.

   4. packages/server/src/utils/logger.ts then runs fs.mkdirSync(""), which throws exactly the
      ENOENT: no such file or directory, mkdir '' above.

   So it is not a misconfiguration on your side. Following the documented Docker quickstart without
   setting LOG_PATH reproduces it every time, which is why the workaround in the comments above,
   uncommenting LOG_PATH, does resolve it.

   #6031 changes the ?? to ?.trim() ||, which also covers a whitespace-only value.

   One thing that might be worth folding into that PR while it is open: logger.ts calls

   if (!fs.existsSync(logDir)) {
       fs.mkdirSync(logDir)
   }

   without { recursive: true }. The workaround being suggested here is LOG_PATH=/root/.flowise/logs,
   which still fails with ENOENT if /root/.flowise does not already exist. Adding recursive: true
   would make both the default and a user-supplied nested path work.