Supported Installation and Upgrade Scenarios

MindCluster Ascend Deployer supports the following installation and upgrade scenarios.

In scenarios where a framework needs to be installed, if the GCC version is earlier than 7.3.0, GCC 7.3.0 is automatically installed and the environment variables or soft links are set to ensure that the system can be used properly after the installation. For details, see (Optional) Configuring GCC Environment Variables.

Precautions

  • During the installation, run the date -s command to calibrate the time in the operating environment to the correct UTC time.
  • MindCluster Ascend Deployer can install only basic libraries to ensure that TensorFlow and PyTorch can run properly. If you need to run complex inference services or model training, the model code may contain libraries related to specific services. You need to install the libraries by yourself.
  • MindCluster Ascend Deployer requires many dependencies. You are advised to install or upgrade the software packages uploaded by yourself after the OS dependencies downloaded by this tool are installed or upgraded.
  • Do not run the --install-scene command repeatedly. Otherwise, the installation will fail.

Installation and Upgrade Scenarios

Table 1 Installation and upgrade scenarios

Scenario Type

Scenario Name

Component to Be Installed

Description

Installation

offline_dev

sys_pkg, Python, NPU, Toolkit, kernels

Offline inference in the development environment

offline_run

sys_pkg, Python, NPU, NNRT

Offline inference in the operating environment

mindspore

sys_pkg, Python, NPU, Toolkit, MindSpore, kernels

Scenario where MindSpore is involved

tensorflow_dev

sys_pkg, Python, NPU, Toolkit, TensorFlow, kernels

TensorFlow development scenario

tensorflow_run

sys_pkg, Python, NPU, NNAE, TensorFlow, kernels

TensorFlow operating scenario

pytorch_dev

sys_pkg, Python, NPU, Toolkit, PyTorch, kernels

PyTorch development scenario

pytorch_run

sys_pkg, Python, NPU, NNAE, PyTorch, kernels

PyTorch operating scenario

Installation

and upgrade

dl

Ascend Device Plugin, Ascend Docker Runtime, NodeD, NPU Exporter, Volcano, Ascend Operator, ClusterD

Installation and upgrade of MindCluster cluster scheduling components

Note

The configuration files of the preceding installation scenarios are stored in the scene directory, for example, ascend-deployer/ascend_deployer/playbooks/scene/scene_dl.yml (for DL scenario). To customize an installation scenario, refer to its corresponding configuration file.