Import system images

System-level container images are used by all users. You can obtain images from Lenovo salesperson or import images into LiCO as system-level container images. This section describes how to create and import system-level container images.

Image bootstrap files

LiCO is released with image bootstrap files for commonly-used AI frameworks. The image bootstrap files for commonly-used AI frameworks in the compressed package you obtained. Users can use these files to create images.

The table below lists the image bootstrap files.

File name Framework CPU/GPU Comments
caffe-1.0-cpu Caffe CPU
caffe-1.0-gpu-cuda92 Caffe CUDA 9.2
  • Supports P100 and V100
  • Caffe does not support CUDA 9.0 officially
    		•
    chainer-6.2.0-gpu-cuda100 Chainer CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
    		•
    intel-caffe-1.1.3-cpu Intel-caffe CPU
    intel-python Other CPU
    jupyter-py27-cpu Jupyter CPU
    jupyter-py27-gpu Jupyter CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
    		•
    jupyter-py36-cpu Jupyter CPU
    jupyter-py36-gpu Jupyter CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
    		•
    jupyter-py37-cpu Jupyter CPU
    jupyter-py37-gpu Jupyter CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
    		•
    letrain-1.2-cpu LeTrain CPU
    letrain-1.2-gpu-cuda100 Caffe CPU
  • Supports P100, V100, RTX5000, and T4
    		•
    lico-file-manager Other CPU
  • Indispensable
    		•
    lico-k8s-tools Other CPU
  • Indispensable
    		•
    mxnet-1.5.0-cpu-mkl Mxnet CPU
    mxnet-1.5.0-gpu-mkl-cuda100 Mxnet CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
    		•
    neon-2.6-cpu Neon CPU
    pytorch-1.1.0-gpu-cuda100 PyTorch CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
    		•
    scikit-single-cpu Scikit CPU
    tensorflow-1.13.1-cpu TensorFlow CPU
    tensorflow-1.13.1-gpu-cuda100 TensorFlow CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
    		•
    tensorflow-1.13.1-gpu-cuda100-hbase TensorFlow CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
  • Supports HBase
    		•
    tensorflow-1.13.1-gpu-cuda100-keras TensorFlow CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
  • Supports Keras(2.2.4)
    		•
    tensorflow-1.13.1-gpu-cuda100-mongodb TensorFlow CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
  • Supports MongoDB
    		•
    tensorflow-1.13.1-mkl TensorFlow CPU
    tensorflow-2.0.0-cpu TensorFlow CPU
    tensorflow-2.0.0-gpu-cuda100 TensorFlow CUDA 10.0
  • Supports P100, V100, RTX5000, and T4
    		•

    Create images

    Step 1. Check to ensure that the lico-auth-internal.key file is under /etc/lico on the LiCO node.

    Step 2. Prepare a build node with a minimum storage of 100 GB.

    Step 3. To the build node, copy the /etc/lico/lico-auth-internal.key file from the LiCO node. For example, copy the lico-auth-internal.key file to the new directory /opt/images. If the new directory cannot be found, create it manually.

    step 4. To the build node, upload the compressed image bootstrap file you obtained which named image_bootstrap.zip. For example, upload the compressed package to the new directory /opt/images.

    Step 5. To the build node, run the following commands to configure the image:

    	cd /opt/images
	unzip image_bootstrap.zip
	cp /opt/images/lico-auth-internal.key /opt/images/image_bootstrap/lico-file-manager
	cp /opt/images/lico-auth-internal.key /opt/images/image_bootstrap/lico-k8s-tools
	chmod 744 /opt/images/image_bootstrap/lico-k8s-tools/lico-auth-internal.key
	chmod 744 /opt/images/image_bootstrap/lico-file-manager/lico-auth-internal.key

    Step 6. To the build node, do one of the following to create image.

    Step 7. Import the ready image into the Docker Repository.

    Import images into LiCO as system-level images

    Run the following commands on the LiCO node to import images from Docker Repository:

    	lico import_system_image kube-tools <LiCO-K8s-Client-IMAGE> other
	lico import_system_image lico-file-manager <LiCO-File-Manager-IMAGE> other
	lico import_system_image caffe-cpu <Caffe-CPU-IMAGE> caffe
	lico import_system_image caffe-gpu <Caffe-GPU-IMAGE> caffe
	lico import_system_image tensorflow-cpu <TensorFlow-CPU-IMAGE> tensorflow
	lico import_system_image tensorflow-gpu <TensorFlow-GPU-IMAGE> tensorflow
	lico import_system_image tensorflow2-cpu <TensorFlow2-CPU-IMAGE> tensorflow2
	lico import_system_image tensorflow2-gpu <TensorFlow2-GPU-IMAGE> tensorflow2
	lico import_system_image tensorflow-mkl <TensorFlow-MKL-IMAGE> tensorflow
	lico import_system_image tensorflow-gpu-hbase <TensorFlow-HBase-IMAGE> tensorflow
	lico import_system_image tensorflow-gpu-keras <TensorFlow-Keras-IMAGE> tensorflow
	lico import_system_image tensorflow-gpu-mongodb <TensorFlow-MongoDB-IMAGE> tensorflow
	lico import_system_image intel-caffe <Intel-Caffe-IMAGE> intel-caffe
	lico import_system_image intel-python <Intel-Python-IMAGE> other
	lico import_system_image pytorch <PyTorch-IMAGE> pytorch
	lico import_system_image neon <NEON-CPU-IMAGE> neon
	lico import_system_image chainer-gpu <Chainer-GPU-IMAGE> chainer
	lico import_system_image letrain-cpu <LeTrain-CPU-IMAGE> letrain
	lico import_system_image letrain-gpu <LeTrain-GPU-IMAGE> letrain
	lico import_system_image mxnet-cpu <MXNet-CPU-IMAGE> mxnet
	lico import_system_image mxnet-gpu <MXNet-GPU-IMAGE> mxnet
	lico import_system_image jupyter-py27-cpu <Jupyter-py27-CPU-IMAGE> jupyter -t py27 -t cpu
	lico import_system_image jupyter-py27-gpu <Jupyter-py27-CPU-IMAGE> jupyter -t py27 -t gpu
	lico import_system_image jupyter-py36-cpu <Jupyter-py36-CPU-IMAGE> jupyter -t py36 -t cpu
	lico import_system_image jupyter-py36-gpu <Jupyter-py36-GPU-IMAGE> jupyter -t py36 -t gpu
	lico import_system_image jupyter-py37-cpu <Jupyter-py37-CPU-IMAGE> jupyter -t py37 -t cpu
	lico import_system_image jupyter-py37-gpu <Jupyter-py37-GPU-IMAGE> jupyter -t py37 -t gpu
	lico import_system_image scikit-cpu <Scikit-CPU-IMAGE> scikit
    Attention: Modify <*-IMAGE> to the actual path in Docker repository.