Software package intended for research teams, for distributed computations. Software at current version supports computations that has sequential nature. Software supports unlimited number of participants, that can request at anytime. All requests will be placed into queue, and executed in the order they arrived. From here can be inferred that the what kind of models can be computed by this software package. That is, only models which has state at every given moment, and can be updated at any given step, regardless of requests waiting in queue. Nevertheless, this is the default behaviour expected from software, and there is freedom to design the application is a way, so it can be used for parallel and non sequential models. Sequential modeling is available off-the-shelf. At current version request supported only by HTTP protocol.
System provides logging available in server, where the software is running. All logs are collected through the most known logging stack that is ELK, hence, all logs are visualized by Kibana. To have easy experience with environment we've wrapped the software into Docker containers, which is known enterprise container platform. This gives freedom to operate with any operating system without infrastructure lock-ins. All is needed for application to work is the Docker Community Edition.
Requirements: Docker Community Edition
To start the app run: docker-compose up.
To start the app in background run: docker-compose up -d.
runing on {YOUR_IP}:3000 HTTP:GET /toJson returns json of the model HTTP:GET /export returns model as encrypted string HTTP:POST /fromJson recovers model from given json HTTP:POST /import recovers model from given encrypted string