Python Executor Setup

The Trendz Python Executor is a Docker-based microservice required for:

• Python Calculation fields
• All prediction models (except Fourier Transformation)
• Code generation for Metric Explorer

Deployment recommendations:

• Same host — deploy the Python Executor on the same host as Trendz. Communication uses WebSockets; placing them on separate hosts requires opening the executor port in the firewall and increases latency.
• Resources — allocate at least 1 GB RAM and 1 CPU core. More CPU directly speeds up prediction model training and execution.

Use these steps only if you installed Trendz via a package (Ubuntu, RHEL/CentOS, or Windows) and need to add the Python Executor separately.

Create a new docker-compose.yml with the following content:

services:
  trendz-python-executor:
    restart: always
    image: "thingsboard/trendz-python-executor:1.16.0"
    ports:
      - "8181:8181"
    environment:
      EXECUTOR_MANAGER: 1
      EXECUTOR_SCRIPT_ENGINE: 6
      THROTTLING_QUEUE_CAPACITY: 10
      THROTTLING_THREAD_POOL_SIZE: 6
      NETWORK_BUFFER_SIZE: 5242880
    volumes:
      - trendz-python-executor-conf:/python-executor-config-files
      - trendz-python-executor-data:/data
volumes:
  trendz-python-executor-conf:
    name: trendz-python-executor-conf
    driver: local
  trendz-python-executor-data:
    name: trendz-python-executor-data
    driver: local

Docker Compose parameters:

Run from the directory containing your docker-compose.yml:

docker compose up -d
docker compose logs -f trendz-python-executor

export SCRIPT_ENGINE_TIMEOUT=30000
export SCRIPT_ENGINE_PROVIDER=DOCKER_CONTAINER
export SCRIPT_ENGINE_DOCKER_PROVIDER_URL=PYTHON_EXECUTOR_HOST:PYTHON_EXECUTOR_PORT

sudo systemctl restart trendz

C:\Program Files (x86)\trendz\conf\trendz.yml

script-engine:
  provider: "${SCRIPT_ENGINE_PROVIDER:DOCKER_CONTAINER}"
  runtime-timeout: "${SCRIPT_ENGINE_TIMEOUT:30000}"
  callback-timeout: "${SCRIPT_ENGINE_TIMEOUT:30000}"
  docker-provider-url: "${SCRIPT_ENGINE_DOCKER_PROVIDER_URL:PYTHON_EXECUTOR_HOST:PYTHON_EXECUTOR_PORT}"
  websocket-buffer-size: "${SCRIPT_ENGINE_WEBSOCKET_BUFFER_SIZE:20971520}"
  websocket-concurrency: "${SCRIPT_ENGINE_WEBSOCKET_CONCURRENCY:5}"

net stop trendz
net start trendz

You can extend the Python Executor with additional libraries to use in Calculation Fields or Custom Prediction Models.

The following libraries are bundled in the official Docker image and available out of the box:

To add a library not listed above, follow these steps. For example, to add the emoji library at version 2.2.0:

docker compose exec trendz-python-executor bash

Append the library name and version to the requirements.txt file in the configuration directory:

echo 'emoji==2.2.0' >> /python-executor-config-files/requirements.txt

Type exit to leave the container shell.

docker compose restart trendz-python-executor

docker compose logs trendz-python-executor

Look for a line confirming successful installation:

Installing custom Python requirements...
Requirement already satisfied: emoji==2.2.0 in /usr/local/lib/python3.9/site-packages

The library is now available in your Trendz Calculation Fields and Custom Prediction Models.

All Python Executor settings are configured via environment variables. Pass them under the environment: key in your docker-compose.yml.

To confirm that Trendz is successfully connected to the Python Executor, create a simple Python Calculation Field and run it — for example:

return 42

If the script executes and returns a result, the connection is working. If it times out or returns an error, check that the Python Executor is running and that SCRIPT_ENGINE_DOCKER_PROVIDER_URL points to the correct host and port.