Skip to content

Latest commit

 

History

History
74 lines (58 loc) · 1.86 KB

HowToMakeQiskitLayer.md

File metadata and controls

74 lines (58 loc) · 1.86 KB
Raw

How to make Qiskit as AWS Lambda layers

Prepare an environment

If you already have an environment, skip this step.
For example, prepare with the following command.

sudo yum update -y
sudo yum install gcc -y
sudo yum install python3-devel -y

pip3 install --user virtualenv
python3 -m venv venv
source ./venv/bin/activate

Get the files to make layers

Use pip to get the files.

mkdir -p qiskit-layer/python
cd qiskit-layer/python
pip3 install qiskit-ibmq-provider==0.4.5 -t .
cd ../../
mkdir -p terra-tmp
cd terra-tmp
pip3 install qiskit-terra==0.11.1 -t .
cp -pr ./* ../qiskit-layer/python

cd ../qiskit-layer/python

rm -rf numpy/ numpy-1.18.5.dist-info/ numpy.libs/ scipy/ scipy-1.4.1.dist-info/ sympy/ sympy-1.6.dist-info/ urllib3/ urllib3-1.25.9.dist-info/

Modify qiskit-layer/python/qiskit/providers/basicaer/basicaerjob.py because ProcessPoolExecutor is not available on AWS Lambda.

Before:

    if sys.platform in ['darwin', 'win32']:
        _executor = futures.ThreadPoolExecutor()
    else:
        _executor = futures.ProcessPoolExecutor()

After:

    _executor = futures.ThreadPoolExecutor()
    #if sys.platform in ['darwin', 'win32']:
    #    _executor = futures.ThreadPoolExecutor()
    #else:
    #    _executor = futures.ProcessPoolExecutor()

deploy

Copy qiskit-layer directory to serverless/common directory and deploy AWS Lambda Layers.

cp -pr qiskit-layer <qctokyo>/serverless/common
cd <qctokyo>/serverless/common
sls deploy --stage <stage>

Deploy applications(source codes and web contents).

cd <qctokyo>/serverless/apps
sls deploy --stage <stage>

Limits

AWS Lambda has a limit of 250MB on the size of a deployment package.
To work around this limit, qctokyo excludes sympy from AWS Lambda layers.
Therefore, you cannot use any qiskit features that use sympy (e.g. parameterized circuits).