Distributed Systems Practice - Containerized

This repository demonstrates multiple distributed system communication patterns. Each section now has an isolated Docker / Docker Compose setup so the host Python environment remains untouched.

Prerequisites

• Docker Engine + Docker Compose plugin

Sections & How to Run

All compose files are under compose/.

Pattern	Compose File	Run Command	Notes
REST (Flask)	compose/rest.yml	docker compose -f compose/rest.yml up -d	Manual exec inside
JSON-RPC	compose/rpc.yml	docker compose -f compose/rpc.yml up -d	Manual server/client
SOAP	compose/soap.yml	docker compose -f compose/soap.yml up -d	Exec zeep client
MQTT (Mosquitto)	compose/mqtt.yml	docker compose -f compose/mqtt.yml up -d	Internal local broker
TCP Req/Resp	compose/reqresp.yml	docker compose -f compose/reqresp.yml up -d	Interactive echo
UDP One-way	compose/udp.yml	docker compose -f compose/udp.yml up -d	UDP send/recv
ZeroMQ	compose/zmq.yml	docker compose -f compose/zmq.yml up -d	REQ/REP + PUB/SUB
ZooKeeper Election	compose/zookeeper_election.yml	docker compose -f compose/zookeeper_election.yml up -d	Nodes auto run
Processes & Threads / Ice	compose/code_process.yml	docker compose -f compose/code_process.yml up -d	Demos + Ice
CORBA	compose/corba.yml	docker compose -f compose/corba.yml up -d	ORB only
Upcall Demo	compose/upcall.yml	docker compose -f compose/upcall.yml up -d	Manual client
Lamport Clock	compose/lamport.yml	docker compose -f compose/lamport.yml up -d	Manual client

Stop with Ctrl+C or:

docker compose -f compose/<file>.yml down

Container Details

Each service has its own image build (Python 3.11 slim) with minimal dependencies via per-module requirements.txt when needed.

Logs & Interaction

Use a second terminal for logs or interactive runs:

docker compose -f compose/rpc.yml exec rpc-server python rpcserver.py
docker compose -f compose/rpc.yml exec rpc-client python rpcclient.py

Extending

• Add healthchecks for production usage.
• Combine into a single multi-service docker-compose.yml if orchestration across patterns is desired.

Manual Execution Cheat Sheet

All clients start idle (tail -f). Exec commands inside containers after up -d.

REST

docker compose -f compose/rest.yml up -d
docker compose -f compose/rest.yml exec rest-server python server.py
docker compose -f compose/rest.yml exec rest-client python client.py --op both -a 2 -b 3

JSON-RPC

docker compose -f compose/rpc.yml up -d
docker compose -f compose/rpc.yml exec rpc-server python rpcserver.py
docker compose -f compose/rpc.yml exec rpc-client python rpcclient.py

SOAP

docker compose -f compose/soap.yml up -d
docker compose -f compose/soap.yml exec soap-server python server.py
docker compose -f compose/soap.yml exec soap-client python client.py

MQTT (internal broker service)

docker compose -f compose/mqtt.yml up -d
# Start subscriber (listens on topic)
docker compose -f compose/mqtt.yml exec mqtt-sub python sub.py
# Publish messages
docker compose -f compose/mqtt.yml exec mqtt-pub python pub.py

TCP Req/Resp

docker compose -f compose/reqresp.yml up -d
docker compose -f compose/reqresp.yml exec reqresp-server python server.py
docker compose -f compose/reqresp.yml exec reqresp-client python client.py

UDP One-way

docker compose -f compose/udp.yml up -d
docker compose -f compose/udp.yml exec udp-server python serverUDP.py
docker compose -f compose/udp.yml exec udp-client python clientUDP.py

ZeroMQ

docker compose -f compose/zmq.yml up -d
docker compose -f compose/zmq.yml exec zmq-req python clientzmq.py
docker compose -f compose/zmq.yml exec zmq-sub python subzmq.py
# PUSH/PULL pattern (producer auto-runs). Open a worker:
docker compose -f compose/zmq.yml exec zmq-pull python pullzmq.py

ZooKeeper Election

docker compose -f compose/zookeeper_election.yml up -d
docker compose -f compose/zookeeper_election.yml logs -f election-node-1

Processes / Threads / Ice

docker compose -f compose/code_process.yml up -d
docker compose -f compose/code_process.yml exec multiprocess-demo python multiprcs.py
docker compose -f compose/code_process.yml exec multithread-demo python multithreads.py
docker compose -f compose/code_process.yml exec ice-server sh -c "slice2py Demo.ice || true; python ice_server.py"
docker compose -f compose/code_process.yml exec ice-client python ice_client.py

CORBA (Naming Service only server currently)

The original example uses the CORBA Naming Service. Due to missing Python omniORB bindings in the base image, only the server stub (which attempts NameService binding) is present and runs for demonstration; a separate client script is not included.

Current state:

• corba-server container builds environment with omniORB tools.
• Server registers object with NameService (requires an external naming service to actually resolve).
• No working Python client provided (bindings unavailable in minimal container without heavier source build).

Run the container (server will start and then block):

docker compose -f compose/corba.yml up -d
docker compose -f compose/corba.yml exec corba-server python server.py

If you later add omniORBpy bindings, you can create a client that resolves the name Hello via the naming context.

Upcall

docker compose -f compose/upcall.yml up -d
docker compose -f compose/upcall.yml exec upcall-server python servercall.py
docker compose -f compose/upcall.yml exec upcall-client python clientcall.py

Lamport Clock

docker compose -f compose/lamport.yml up -d
docker compose -f compose/lamport.yml exec lamport-server python s_lamp.py
docker compose -f compose/lamport.yml exec lamport-client python c_lamp.py

Limitations / Future Work

• Add healthchecks & simple retry loops (e.g. RPC client) instead of fixed sleeps.
• Security (TLS, auth) intentionally omitted for clarity.

Cleanup

Remove all images:

docker image prune -f

License

Educational use.