This is the meta repository of the Price Wars platform that allows the simulation of a dynamic marketplace similar to online marketplaces like www.amazon.com. Using the simulation, one can test different pricing strategies and assess their performance compared to other pricing strategies.
The simulation is built using a microservice architecture and consists of multiple components representing important players in the simulation. Each component has its own repository and documentation and the links can be found in the following section.
For more information about the platform and publications, see the our chair's project site on Dynamic Pricing under Competition.
On the master branch, this repository contains the docker-setup that allows running the simulation locally. On the gh-pages branch one can find the swagger-documentation of the REST-APIs of all components.
Due to the current github bug that submodules with dependencies to private repositories cannot be resolved, the github page build process fails, enforcing us to separate the API specification from the submodules for rendering those via github.io.
The API specification can be found here.
Repositories
- Management UI: https://github.com/hpi-epic/pricewars-mgmt-ui
- Consumer: https://github.com/hpi-epic/pricewars-consumer
- Producer: https://github.com/hpi-epic/pricewars-producer
- Marketplace: https://github.com/hpi-epic/pricewars-marketplace
- Merchant: https://github.com/hpi-epic/pricewars-merchant
- Kafka RESTful API: https://github.com/hpi-epic/pricewars-kafka-rest
Please note, that we cannot make the repository for the management UI public (we use a free bootstrap template that we are not allowed to distribute). But you can ask us anytime for access and we will add you to the access list. The management UI is not required but eases the first steps on the platform.
If you are working on Windows, make sure to set the following git-settings:
git config --global core.eol lf
git config --global core.autocrlf input
The first setting will ensure that files keep their lf-line endings when checking out a repository. Otherwise the line endings of all files will be converted to cr+lf under Windows, causing problems with script-execution within the Docker-containers that expect lf-endings. The second setting ensures that line endings are converted back to what they were on checkout when pushing a change.
(These settings are global so they will affect all repositories, make sure to change them again if you do not want this behaviour on other repositories!)
Then go ahead and clone the repository:
git clone git@github.com:hpi-epic/pricewars.git
cd pricewars
git submodule update --init --recursive
Build docker images and containers with the following command. Bring a fast internet line and some time. This can take up to 30 minutes to finish.
docker-compose up --no-start
This command might not be available if you are on an older docker-compose version.
Use docker-compose create instead.
Once the containers are created, you can start the Pricewars platform:
docker-compose up
You can shut down the platform with CTRL + C or docker-compose stop.
If you want to use the management-ui add the following lines to your host file. This allows you to access container by domain name via your local docker ip address.
- open "/etc/hosts" as root on Linux / Unix or "C:\Windows\System32\drivers\etc\hosts" on Windows as Administrator
- and don't forget additional names if you specify more containers!
172.29.0.1 postgres redis zookeeper kafka kafka-reverse-proxy
172.29.0.1 flink-jobmanager flink-taskmanager analytics management-ui
172.29.0.1 marketplace producer consumer merchant
Warning: There might be routing problems if the docker network (172.29.0.0/24) overlaps with your local network.
If this is the case, change the ip address in docker-compose.yml under the networks entry.
After that change the addresses in the host file (172.29.0.1) to your new docker host ip address.
After starting the Pricewars platform with docker-compose up, it can be controlled with the Management UI
- [Optional] Configure available products in the Config/Producer section
- Start the Consumer
- Merchants are trading products now. The Dashboard shows graphs about sales, profits and more.
Run the following commands to run the platform in a clean state.
docker-compose rm --stop
docker-compose up
First, stop your running containers.
docker-compose stop
Update repositories.
git pull
git submodule update
Rebuild all images that have changed:
docker-compose build
For Windows versions that don't have Docker Native support, like Home etc., another solution is to use an Ubuntu virtual machine with Docker installed. We recommend to use Ubuntu Server as the GUI of other Ubuntu versions might consume to much main memory. The main memory should be at least 8GB, since your Windows consumes about 2GB, Ubuntu Server 0.5GB and the simulation about 4GB. Also, your browser will consume a few hundred MB, too.
On your guest system, you have to adjust your DNS settings like mentioned before. If you want to access your simulation from your host system, i.e. with the managment ui in the browser or with a locally deployed merchant, you also need to adjust the DNS settings of this, too. You could also use the Windows Subsystem for Linux (WSL) to run local merchants. Be aware, that you need to adjust the DNS settings a third time, this time in the hosts file of the WSL. Also, you might need to add an exception to your firewall to allow your merchant to access the virtual machine.
First, the analytics container is expected to be stopped after a short moment - it's just used to compile and export the flink jobs. If this is the only container not running, you're fine. In case any other container is not running, analyse the first container (except for analytics) that stopped.
- Postgres: Bring the platform in a clean state with
docker-compose rm --stopand run it again. - Zookeeper / Kafka: If you just stopped some older containers: Wait! There is a timeout for Zookeeper to notice that Kafka has been stopped (timestamp-based, so it works even if Zookeeper is not running). Bring the platform in a clean state with
docker-compose rm --stopand run it again. - Others: Try to read the logs or read on.
- Reset the containers and the network:
docker system prune(and restart the Docker service). - Terminate Docker and ensure, that all docker processes are stopped (especially the network service).
- Restart your computer and wait (even though it might be slow) for about five to ten minutes.
- Reset Docker to factory defaults (should be your last attempt, as this requires re-building of all images):
- macOS: Click on "Preferences" > "Reset" > "Reset to factory defaults"
For details regarding the deployment of the component, we kindly refer to the deployment section of the microservice specific README.md file. The links can be found above.
You can run a benchmark on the Pricewars platform with the benchmark tool benchmark.py. This tool allows to run the platform in a given configuration for a specific time period. Afterwards, results of this run are written to the output directory.
Example command:
python3 helper_scripts/benchmark.py --duration 30 --output <output directory> --merchants <merchant A command> <merchant B command> --consumer <consumer command>
This starts the whole platform and two merchants to compete against each other for 30 minutes.
Run python3 helper_scripts/benchmark.py --help to see all arguments.