/PERT-GNN-KDD23

Official Implementation of our KDD 2023 Paper: PERT-GNN: Latency Prediction for Microservice-based Cloud-Native Applications via Graph Neural Networks.

Primary LanguagePython

Requirements

  • 200G+ Disk space

Steps to reproduce:

  1. Create conda environemnt: conda env create -f environment.yml and activate it: conda activate pert-gnn
  2. Follow https://github.com/alibaba/clusterdata/tree/master/cluster-trace-microservices-v2021 to fetch the data.
  3. Put the MSCallGraph and MSResource directory under data/
  4. mkdir processed
  5. Run python preprocess.py
  6. Run python pert_gnn.py (It takes 10+hrs to generate the data_list for the first time)