N8n instance freezes/crashes during simple AI Agent workflow for a specific image file

[jgoncalves]

Describe the problem/error/question

n8n instance freezes or crashes during a simple AI Agent workflow when processing a specific image file.
The workflow executes successfully until it reaches the AI Agent node.

Unfortunately, I cannot share the specific file due to privacy concerns. However, I can guarantee that the image is not unusually large (2296x4080), nor is it strange — it’s just a simple photo of a document:

I’ve tried resizing the image to smaller dimensions (e.g., 500x500), but the problem still occurs:

I’ve also experienced this same behavior with other specific files.
For most other files, the issue does not occur.

I’m not sure what might be triggering this behavior, but it is certainly odd. There is also no relevant information in the logs.

What is the error message (if any)?

There is none. The only symptom is 100% CPU usage in the Docker container (with low memory usage). Also, the container itself is not frozen — only the Node.js process is.
This behavior requires the container to be restarted, which causes the workflow to fail.

Here’s the output of top inside the container during the freeze:

Please share your workflow

For reference, this is my deployment YAML (some information has been omitted):

name: n8n
services:

  master:
    restart: always
    image: mydockerregistry.com/n8n/n8n-compose:1.92.2
    ports:
      - '127.0.0.1:5678:5678'
    links:
      - postgres
    labels:
      - traefik.enable=true
      - traefik.http.routers.n8n.rule=Host(`${SUBDOMAIN}.${DOMAIN_NAME}`)
      - traefik.http.routers.n8n.tls=true
      - traefik.http.routers.n8n.entrypoints=websecure
    environment:
      - NODE_ENV=homo
      - WORKFLOWS_DEFAULT_NAME=My workflow
      - EXECUTIONS_TIMEOUT=7200
      - N8N_HOST=${SUBDOMAIN}.${DOMAIN_NAME}
      - N8N_PORT=5678
      - N8N_PROTOCOL=https
      - N8N_EDITOR_BASE_URL=https://${SUBDOMAIN}.${DOMAIN_NAME}/
      - WEBHOOK_URL=https://${SUBDOMAIN}.${DOMAIN_NAME}/
      - N8N_DIAGNOSTICS_ENABLED=false
      - N8N_VERSION_NOTIFICATIONS_ENABLED=false
      - N8N_DIAGNOSTICS_CONFIG_FRONTEND=
      - N8N_DIAGNOSTICS_CONFIG_BACKEND=
      - N8N_TEMPLATES_ENABLED=false
      - N8N_HIRING_BANNER_ENABLED=false
      - N8N_PUBLIC_API_SWAGGERUI_DISABLED=false
      - N8N_RUNNERS_ENABLED=true
      - N8N_EMAIL_MODE=smtp
      - N8N_SMTP_HOST=mysmtphost.com
      - [email protected]
      - N8N_SMTP_PORT=22
      - N8N_SMTP_SSL=false
      - N8N_SMTP_STARTTLS=false
      - DB_TYPE=postgresdb
      - DB_POSTGRESDB_HOST=postgres
      - DB_POSTGRESDB_PORT=5432
      - DB_POSTGRESDB_DATABASE=${POSTGRES_DB}
      - DB_POSTGRESDB_USER=${POSTGRES_NON_ROOT_USER}
      - DB_POSTGRESDB_PASSWORD=${POSTGRES_NON_ROOT_PASSWORD}
      - N8N_USER_MANAGEMENT_JWT_DURATION_HOURS=24
      - NODES_EXCLUDE=["n8n-nodes-base.executeCommand", "n8n-nodes-base.readWriteFile"]
      - N8N_COMMUNITY_PACKAGES_ENABLED=false
      - NODE_FUNCTION_ALLOW_EXTERNAL=pdfjs-dist
      - EXECUTIONS_DATA_MAX_AGE=720
      - QUEUE_HEALTH_CHECK_ACTIVE=true
      - N8N_METRICS=true
      - N8N_DEFAULT_BINARY_DATA_MODE=filesystem
      - EXECUTIONS_DATA_PRUNE=true
      - N8N_LOG_LEVEL=debug
      - NODE_OPTIONS="–max-old-space-size=6000"

    healthcheck:
      test: ["CMD-SHELL", "/usr/bin/wget --server-response --proxy off --no-verbose --tries=1 --timeout=3 127.0.0.1:5678/healthz -O /dev/null 2>&1 | grep -q 'HTTP/1.1 200 OK'"]
      interval: 20s
      retries: 3
      timeout: 10s
      start_period: 30s

    volumes:
      - n8n_data:/home/node/.n8n
      - ./local-files:/files

    depends_on:
      postgres:
        condition: service_healthy

  postgres:
    image: postgres:16
    restart: always
    ports:
      - '5432:5432'
    environment:
      - POSTGRES_USER
      - POSTGRES_PASSWORD
      - POSTGRES_DB
      - POSTGRES_NON_ROOT_USER
      - POSTGRES_NON_ROOT_PASSWORD
    volumes:
      - db_storage:/var/lib/postgresql/data
      - ./init-data.sh:/docker-entrypoint-initdb.d/init-data.sh
    healthcheck:
      test: ['CMD-SHELL', 'pg_isready -h localhost -U ${POSTGRES_USER} -d ${POSTGRES_DB}']
      interval: 5s
      timeout: 5s
      retries: 10

  traefik:
    restart: always
    image: traefik
    healthcheck:
      interval: 10s
      timeout: 5s
      retries: 3
    ports:
      - '443:443'
      - '80:80'
    volumes:
      - ./certs/:/etc/traefik/certs/
      - ./traefik.yaml:/etc/traefik/traefik.yaml
      - /var/run/docker.sock:/var/run/docker.sock

volumes:
  n8n_data:
  db_storage:

Debug info

core

• n8nVersion: 1.92.2
• platform: docker (self-hosted)
• nodeJsVersion: 20.19.0
• database: postgres
• executionMode: regular
• concurrency: -1
• license: enterprise (production)

storage

• success: all
• error: all
• progress: false
• manual: true
• binaryMode: filesystem

This container is running inside a virtual machine, config:

CPU 4 CPU(s)
Memory 8 GB, 2 GB memory active
Hard disk 1
250 GB

[grandpa_sam]

I am having this same problem! were you able to solve this?

[jgoncalves]

Sorry for the late reply. Updating the version actually solved the issue. I’m not entirely sure why it happened in the first place, and I’m still a bit concerned that it could occur again. For now, though, everything is working as expected in version 1.107.1.