This project was part of the distributed systems lab during the summer semester 2019 at Hochschule Darmstadt, University of Applied Sciences. The task was to implement a distributed weather service consisting of

• sensors (temperature, humidity, rain, wind) : sensors send randomly generated weather data via UDP/ MQTT to their respective weather stations
• weather stations: receive weather data from at least three sensors. Weather stations include a HTTP- server which displays the received sensor data via a REST API. Furthermore, the weather stations transmit the aggregated sensor data via Thrift to a connected weather service.
• weather services: weather services gather data from multiple weather stations. They implement a master slave architecture and consist of at least three replicated servers which synchronize the received data continuously.

gradle installDist

Launch the demo mode to run the application on your host system.

docker-compose up

Note: if you run the code on macOS, remove the last line from the docker-compose.yml (network_mode: host)

./test_weatherservice.sh

The test_weatherservice.sh script launches a weather service instance with three servers.

./test_weatherpacket.sh

The test_weatherpacket.sh script launches three weather stations and for each station it launches four sensors, that send their data to the corresponding station. Each weather station transmits its data to the weather service server it is assigned to.

Check the transmitted sensor values on ports 5554, 5564, 5574 according to the REST API documentation.

docker-compose down

docker-compose up

Note: if you run the code on macOS, remove the last line from the docker-compose.yml (network_mode: host)

Update the start.sh file:

• replace ip0=0.0.0.0 with your local IP address
• replace ip1=0.0.0.0 with a remote IP address you want to communicate with

The start.sh script will launch a weather service instance, three weather stations and their respective sensors.

./start.sh

./stop.sh

docker-compose down

sensors/sensortype/whichdata

sensortypes:

• temperature
• rain
• wind
• humidity

whichdata:

• current
• history