Skip to main content

Standalone Deployment

🚀 AskTable's standalone version now supports free download, installation, and usage.

💪 This guide will help you quickly deploy AskTable in your local environment and use it with your internal data.

Continuing to download, install, and use AskTable means you have read and agree to the Software Download, Installation, and Usage Agreement.

Prerequisites​

Hardware​

Minimum configuration requirements: 2-core CPU, 4G memory, 10G hard disk. Supports x86 and ARM architectures. On some domestic systems like Kylin V10SP1, privileged: true needs to be added in each container to obtain the highest permissions ⚠️

Operating System​

  1. Linux system The following software needs to be installed:

    1. Docker: Docker Installation Documentation.
    2. Docker Compose: Used to package and deploy AskTable and MySQL database. Docker Compose Installation Documentation.

  2. macOS system The following software needs to be installed:

    1. Docker Desktop: An application for macOS desktop computers that includes Docker and Docker Compose. Installation Documentation.

Quick Deployment​

Save the following content as a docker-compose.yml file:

services:
  asktable:
    # AskTable
    image: registry.cn-shanghai.aliyuncs.com/datamini/asktable-all-in-one:latest  # For mainland China
    depends_on:
      - asktable_mysql
    ports:
      - "8000:80"
    privileged: true
    environment:
      TZ: Asia/Shanghai
      BASE_URL: http://127.0.0.1:8000        # 【Mandatory to modify】The address exposed by the container (i.e., the address accessed from outside, such as a browser)
      LLM_API_KEY: asktable-xxxxxx=xxxxxxxx      # 【Mandatory to modify】AI model token, which is a string starting with "asktable-". After logging into AskTable(https://cloud.asktable.com), click "Private Deployment AskTable" at the top to get it.
      MYSQL_HOST: asktable_mysql               # MySQL database address (this MySQL is the AskTable database, not the database queried by AskTable, the latter needs to be added on the AskTable interface)
      MYSQL_USER: asktable                     # MySQL database username
      MYSQL_PASSWORD: asktable                 # 【Recommended to modify】MySQL database password
      MYSQL_DB: asktable                       # MySQL database name
      MYSQL_ROOT_PASSWORD: asktable            # 【Recommended to modify】MySQL database root password (used for initializing sample data)
      VDB_ADDRESS: http://asktable_vdb:6333    # Vector database address
      VDB_API_KEY: asktable                    #【Recommended to modify】Vector database API KEY
    volumes:
      - ./asktable_data:/asktable

  asktable_mysql:
    image: registry.cn-shanghai.aliyuncs.com/dminfra/mysql:8.0.27
    ports:
      - "3306"
    environment:
      TZ: Asia/Shanghai
      MYSQL_ROOT_PASSWORD: asktable     # 【Recommended to modify】
      MYSQL_DATABASE: asktable
      MYSQL_USER: asktable
      MYSQL_PASSWORD: asktable          # 【Recommended to modify】
    volumes:
      - ./mysql_data:/var/lib/mysql

  asktable_vdb:
    image: registry.cn-shanghai.aliyuncs.com/dminfra/qdrant:v1.11.5
    ports:
      - "6333"
      - "6334"
    environment:
      RUN_MODE: production
      TZ: Asia/Shanghai
      QDRANT__SERVICE__API_KEY: asktable  # 【Recommended to modify】
    volumes:
      - ./qdrant_storage:/qdrant/storage
      - ./qdrant_snapshots:/qdrant/snapshots

Note that the BASE_URL and LLM_API_KEY configuration items must be modified according to the actual situation.

Run AskTable​

After the configuration is completed, you can start AskTable via Docker Compose:

  1. Build and Start Containers

Run the following command in the directory where the docker-compose.yml file is located:

docker compose up -d
# or docker-compose up -d

This command will start the AskTable application and MySQL database containers in the background.

  1. Access AskTable

Now, AskTable has started. You can access it in your browser:

http://127.0.0.1:8000

Here, you can log in using the default administrator (username is admin@asktable.com, password is admin) and start exploring the application.

Explore AskTable​

After logging in, you can start using AskTable for various tasks, such as:

  • Data management: Create, manage, and query data sources.
  • Interact with AskTable AI: Use language models to query and analyze your data.
  • Customize business knowledge: Configure terminology libraries, training datasets, and default preferences according to your business scenarios.

Advanced Configuration (Optional)​

  1. Set Administrator Account: If you want to set an administrator account, you can configure the following parameters:

    AUTH_EMAIL_INIT_ADMIN_EMAIL: admin@asktable.com
    AUTH_EMAIL_INIT_ADMIN_PASSWORD: admin

  2. Use Local LLM Model: If you want to use a local model, assuming that the model inference service has been deployed, you can configure the following parameters:

    Warning

    Using models not provided by AskTable may not guarantee performance. It is recommended to apply at https://cloud.asktable.com.

    LLM_BASE_URL: http://10.10.0.3:11434/v1  # Address of the local LLM service
    LLM_API_KEY: your_api_key_here  # API Key of the local LLM service (if no key authentication is required, it is not necessary to configure)
    AT_FORCE_LLM_MODEL: Qwen2.5-72B-Instruct-128K  # The name of the model to be forced to use

    Note that we recommend the full-performance version DeepSeek-V3-671B for local models. Other models must meet the following conditions simultaneously:

    • At least 72B
    • At least 64K context
    • Support Function Call
    • Support JSON (Instruct) Output

    Additional performance requirements:

    • Each AI question may involve 2-4 concurrent calls to large LLM models.
    • Each AI question may require approximately 5000 Input Tokens and 500 Output Tokens.
    Local LLM Service Detection Script

    We have prepared a detection script for you, making it easy to quickly verify whether the local LLM meets the requirements on the server. Execute the following command in the server terminal (remember to replace YOUR_BASE_URL and YOUR_API_KEY with your actual model service address and key):

    curl -O https://asktable-assets.oss-rg-china-mainland.aliyuncs.com/llm-env-test/asktable_llm_check.py && python3 asktable_llm_check.py --base-url YOUR_BASE_URL --api-key YOUR_API_KEY

    If the detection passes, it can be seamlessly integrated with AskTable.

Update Version​

If you need to update to the latest version, run the following commands in the directory where the docker-compose.yml file is located:

docker compose pull
docker compose down
docker compose up -d

Frequently Asked Questions​

  1. Docker fails to run, and there is a network error. How to solve it?

    Ensure that the firewall allows access to the following addresses (port 443):

    • registry.cn-shanghai.aliyuncs.com (download images)
    • dockerauth.cn-hangzhou.aliyuncs.com (download images)
    • aiproxy.asktable.com (access AI model service)
    • lf.asktable.com (access AI Tracing service)

  2. AskTable can be opened, but login prompts: Connection failed

    Please check the parameter BASE_URL in docker-compose.yaml. It should be configured to the address accessed through the browser.

  3. How to verify the connectivity between the container and the database (connection status)?

    docker exec -ti asktable bash -c 'telnet DB_HOST DB_PORT'

    • DB_HOST (database host address)

    • DB_PORT (database port number)

    1. If the connection is successful, it displays Connected to localhost.
    2. If the connection is refused / network is not connected / timeout, it displays telnet: Unable to connect to remote host: Connection refused/host/timed out

  4. How to uninstall?

    Perform the following operations in the directory where the docker-compose.yml file is located:

    1. Run the command docker compose down
    2. Delete the directories asktable_data, mysql_data, qdrant_storage, and qdrant_snapshots (if any)

Notes​

This deployment mode is suitable for quickly experiencing AskTable, but it is not suitable for production environments. Performance, accuracy, and stability cannot be guaranteed. If higher requirements are needed, please contact us.

Thank you again for using it. Good luck! 🚀