Python serverless communication framework for distributed AWS Lambda applications. Part of my master's thesis.
Implemented communication methods:
- S3
- EFS
- DynamoDB
- Redis
- Simple relay implementation
- P2P w/ NAT traversal
python -m build
virtualenv venv
venv/bin/pip install -e .
venv/bin/python
aws s3 cp dist/bulletin-0.0.5.tar.gz s3://bulletin-py/bulletin/
aws s3 cp dist/bulletin-0.0.5-py3-none-any.whl s3://bulletin-py/bulletin/
# Install bucket-dir first
bucket-dir bulletin-py
Use like this in requirements.txt.
--extra-index-url https://dxsxsb220w9y2.cloudfront.net
bulletin==0.0.5
src/
- Bulletin source.tf/benchmark/
- Terraform config for point to point benchmarks of event exchange methods.tf/e2e/
- Terraform config for e2e benchmark: federated machine learning.tf/bulletin/
- Publish bulletin on S3.tf/ec2/
- Single VM to run benchmarks with low latency to AWS API.notebooks/
- Jupiter notebooks with benchmark result analysis.