Skip to content

Latest commit

 

History

History
110 lines (68 loc) · 3.47 KB

README.md

File metadata and controls

110 lines (68 loc) · 3.47 KB
Raw

RCPSP.rs

RCPSP.rs is a tabu search-based local-search algorithm for the Resource-Constrained Project Scheduling Problem in Rust.

Installation

Local Dev Container in VS Code

To run this repo inside a dev container, please install the corresponding Remote Code Extension in VS Code and open the repo inside it. A corresponding dev container manifest is already defined.

Please refer to the official documentation for more extensive and detailed instructions.

Github Codespaces

As this project configures a dev container, running a Github Codespace is easily possible.

To open a cloud-hosted development environment, select the green code button, select "Codespaces," and then select "Create codespace on main."

This operation will take a couple of minutes, after which one is ready to go.

Local Installation

Follow the official Rust language's getting started guide to set up Rust on your local machine.

Testing

One can execute a test suite by running:

cargo test

Graph Visualization

To visualize the PSP Lib problem as a graph, one needs to have previously installed graphviz.

One can generate a graph using:

cargo run -- graph ./examples/file_name.sm ./file_name.dot
dot -Tpng file_name.dot > file_name.png

Benchmarks

One can perform Criterion benchmarks by executing:

cargo bench

The location of the generated report is: target/criterion/report/index.html.

Running the scheduler binary

To run the scheduler in debug mode, without LLVM optimizations applied, one should execute:

cargo run -- schedule ./examples/j30.sm/j3045_10.sm --swr 15 --tls 100 --misb 1500 --noi 3000 -p -v

To run a release build of the scheduler, with LLVM optimizations on, one should:

cargo run --release -- schedule ./examples/j30.sm/j3045_10.sm --swr 15 --tls 100 --misb 1500 --noi 3000 -p

To run a release build of the scheduler using a rayon-based multi-schedule scheduling algorithm, with LLVM optimizations on, one should:

cargo run --release -- schedule ./examples/j30.sm/j3045_10.sm --swr 15 --tls 100 --misb 1500 --noi 3000 -p --algo rayon-multi

To run a release build of the scheduler using a custom multi-threaded scheduling algorithm, with LLVM optimizations on, one should:

cargo run --release -- schedule ./examples/j30.sm/j3045_10.sm --swr 15 --tls 100 --misb 1500 --noi 3000 -p --algo custom

One can evaluate the scheduler quality by running:

cargo run --release -- benchmark ./examples/j30.sm ./j30_results.csv
cargo run --release -- benchmark ./examples/j30.sm ./j30_results_rayon_multi.csv --algo rayon-multi
cargo run --release -- benchmark ./examples/j30.sm ./j30_results_custom.csv --algo custom

Resources

Inspiration was drawn from the following resources:

  • Libor Bukata, Premysl Sucha, and Zdeněk Hanzálek. Solving the resource constrained project scheduling problem using the parallel tabu search designed for the cuda platform. Journal of Parallel and Distributed Computing, 77, 11 2014.