Skip to content

project-structure.md

Latest commit

 

History

History
85 lines (81 loc) · 3.85 KB

project-structure.md

File metadata and controls

85 lines (81 loc) · 3.85 KB

Project Structure and Architecture

The project is organized as follows:

.
├── configs
│   ├── models
│   │   └── ton.json
│   └── topos
│       ├── fattreetopo.json
│       └── topo.json
├── controller.py
├── convert_RF_and_populate_tables.py
├── dune.lua
├── Makefile
├── mininet
│   ├── dune.py
│   └── p4
│       ├── link.py
│       └── node.py
├── models
│   └── ton
│       ├── ToN_IoT_ClID0_T3_F5_L41_N4_With_Others.sav
│       ├── ToN_IoT_ClID1_T1_F5_L41_N2_With_Others.sav
│       ├── ToN_IoT_ClID2_T1_F6_L85_N3_With_Others.sav
│       └── ToN_IoT_ClID3_T1_F12_L129_N3_With_Others.sav
├── p4include
│   ├── dune_compute_checksum.p4
│   ├── dune_egress_deparser.p4
│   ├── dune_egress.p4
│   ├── dune_headers.p4
│   ├── dune_inference.p4
│   ├── dune_ingress.p4
│   ├── dune_ingress_parser.p4
│   ├── dune_verify_checksum.p4
│   └── network_headers.p4
├── p4sources
│   ├── no_inference.p4
│   ├── template
│   ├── ton_iot_m1.p4
│   ├── ton_iot_m2.p4
│   ├── ton_iot_m3.p4
│   ├── ton_iot_m4.p4
│   └── TON_IOT.md
├── populate_forwarding_tables.py
├── README.md
├── requirements.txt
├── upload_p4prog_to_switch.py
├── logs
├── pcaps
└── utils
    ├── params.ini
    ├── prepare_pcap_traces.py
    └── process_result_pcaps.sh

Critical Files

  • controller.py: The main controller script for managing the Mininet switches.
  • Makefile: Central script to build and run the project. It compiles the P4 programs, builds the parameter strings, and starts the Mininet environment.
  • mininet/: Contains the custom Mininet classes for creating a simple_switch_grpc-based network that supports DUNE-based inference models.
    • dune.py: Main script to set up the topologies and test, i.e., traffic injection patters.
    • p4/: Contains custom mininet classes for handling links and p4-based switches.
  • p4include/: Contains P4 header files and definitions used in the project. You should not need to modify these files.
  • p4sources/: Contains the P4 source files:
    • no_inference.p4: A P4 program without inference capabilities.
    • template: A template for creating new P4 programs. You should use this as a starting point for new P4 programs.
    • ton_iot_m*.p4: P4 programs for different models of the ToN IoT dataset.
    • TON_IOT.md: Documentation for the ToN IoT dataset.
  • utils/:
    • prepare_pcap_traces.py: Script to prepare PCAP traces for being injected in the Mininet topo.
    • process_result_pcaps.sh: Script to process the result PCAPs generated by the Mininet tests.

Output Files

  • logs/: Contains the logs generated by the controller and switches during the execution of the Mininet tests.
  • pcaps/: Contains the PCAP files generated by each switch in the Mininet network. Each file is named according to the switch and port, e.g., pe_l0-eth5_out.pcap for the output of the first leaf switch from the egress pod on the fifth port.

Advanced Usage Files

  • configs/: Contains configuration files for models and topologies.
  • convert_RF_and_populate_tables.py: Script to convert previously pickled RF models into table data, and populate tables.
  • dune.lua: dissector script to help Wireshark parse the Dune protocol.
  • models/: Contains the trained models used for inference.
  • populate_forwarding_tables.py: Script to populate the forwarding tables (not related to inference).
  • upload_p4prog_to_switch.py: Script to upload P4 programs to the switch.
  • utils/:
    • params.ini: Configuration file for the prepare_pcap_traces.py script.