Deploying the Development Environment in Containers

If the CentOS system (including the host machine and container) is used for configuration, the libstdc++ version must be later than libstdc++.so.6.0.24.
For security purposes, only non-root users can start containers.

You can refer to Figure 1 to deploy the Rec SDK TensorFlow development environment in a container.

Figure 1 Configuring the development environment and building a training image in a container

Prepare the host.
Deploy the host environment by referring to Dependency Installation.
Obtain the training image and start the container. You can refer to the Ascend Hub to create the base image (search for rec_sdk and select the appropriate sample) and complete the installation of the Rec SDK TensorFlow.
Optional: To use the dynamic capacity expansion function in the container, compile and install the corresponding operator package by referring to (Optional) Installing the Dynamic Capacity Expansion Operator Package of the on-chip memory.
Optional: To use the Hadoop distributed file system, deploy the environment and set up a cluster by referring to Hadoop official documents. Hadoop 2.7.5 is recommended.

After the environment is deployed, the owner of the /usr/local/hadoop-2.7.5/sbin file is 20415 (non-root user). The owner has the permission to rename and create a file to replace the executable file in the PATH environment variable of the root user, which may cause unauthorized operations.

Parent topic: Deploying the Development Environment in Containers