[object Object]

This document describes how to create a MindIE image package for deploying cluster services.

[object Object]

The following table lists the required hardware environments and OSs.

[object Object]undefined[object Object]

The version of the software on which the MindIE image package depends must meet the following requirements.

Table 1 Software description

[object Object]undefined[object Object]

• The APT source download path may be incorrect in a new container. You need to configure a dedicated source for Ubuntu 22.04 to improve the download speed.

• The installation requires the download of related dependencies. Ensure that the installation environment can be connected to the network.

• Check whether the source is valid.

  [object Object]

  If an error is reported during command execution or dependency installation, check whether the network is connected, or replace the source in the [object Object] file with an available source or use an image source. (You can visit to find more information about how to configure a Huawei image source.)

• You have installed Docker (version 24.x.x or later) on the host.

[object Object]

The following uses the Atlas 800I A2 inference server as an example to describe how to create a MindIE image.

[object Object]

1. Save the software downloaded or made from to a directory, for example, [object Object].

2. Pull the Ubuntu 22.04 image.

   [object Object]

   Check the Ubuntu image.

   [object Object]

   If information similar to the following is printed, the image is successfully pulled:

   [object Object]

3. [object Object][object Object]In the·[object Object] directory, compile the Dockerfile and other script files required for installation. The file directory structure must be as follows:

   [object Object]

   The files [object Object], [object Object], [object Object], [object Object], [object Object], [object Object], and [object Object] are user-provided.

   1. Compile the [object Object] file.

      [object Object]

      [object Object]

   2. Compile the Dockerfile.

      By default, the Dockerfile uses Ubuntu 22.04, AArch64 architecture, and Python 3.10. It is for reference only and can be modified. (If building an image based on Ubuntu 24.04, replace [object Object] with [object Object] in the Dockerfile.)

      [object Object]

      [object Object]

   3. Create the [object Object] file.

      [object Object]

   4. Create the [object Object] file.

      [object Object]

   5. Create the [object Object] file.

      [object Object]

   6. Create the [object Object] file.

      [object Object]

   7. Create the [object Object] file.

      [object Object]

   [object Object]

4. [object Object][object Object] Start [object Object] on the host.

   1. Ensure that [object Object] has been installed on the host for starting the service.

      If the host runs Ubuntu, run the following command to install [object Object]:

      [object Object]

   2. Run the following commands in the build environment:

      [object Object]

   3. If the following information is displayed, the service is started successfully.

      [object Object]

      Keep this CLI, and close it until the build is complete.

      [object Object]

5. Create a CLI (do not close the CLI in step 4), go to the [object Object] directory, and build the image.

   [object Object]

   After the image is built, information similar to the following is printed:

   [object Object]

   Then, run the following command:

   [object Object]

   If the following information is displayed, the image is successfully built:

   [object Object]