f2189a0468
After function autoscaling support, it's impossible for qinling-engine to get execution log because it doesn't know which pod it is talking to. So, it's neccessary for the runtime to return execution logs to qinling engine. The qinling client is not affected. Change-Id: I96dfd00cc83d8b8a5e8c601ee3800b1ef1a45b1b |
||
---|---|---|
.. | ||
Dockerfile | ||
README.md | ||
requirements.txt | ||
server.py |
Qinling: Python Environment
This is the Python environment for Qinling.
It's a Docker image containing a Python 2.7 runtime, along with a dynamic loader. A few common dependencies are included in the requirements.txt file. End users need to provide their own dependencies in their function packages through Qinling API or CLI.
Rebuilding and pushing the image
You'll need access to a Docker registry to push the image, by default it's docker hub. After modification, build a new image and upload to docker hub:
docker build -t USER/python-runtime . && docker push USER/python-runtime
Using the image in Qinling
After the image is ready in docker hub, create a runtime in Qinling:
http POST http://127.0.0.1:7070/v1/runtimes name=python2.7 image=USER/python-runtime