valzam / switch Goto Github PK

switch/server.py

• Creates a sub socket for producers to connect to
• Creates a pub socket for consumers to connect to
• Gets instantiated with a specific ML model that will be used during message handling

switch/predictor.py

Currently a mock implementation of a ML model

switch/discovery.py

• Simple HTTP interface for querying information about the running server
• Implemented as zmq.ROUTER

• Setup virtualenv or any other python 3 installion
• Install requirements.txt
• Create and source .env file
• Use Makefile to run everything

# Set necessary paths
PYTHONPATH=~/dev/python/switch
PROTOC=~/dev/protobuf/bin/protoc

# Server configuration
INCOMING_PORT=5556
OUTGOING_PORT=5566
INCOMING_TOPIC=Data_1A
OUTGOING_TOPIC=Predictions_1A