Morpheus Integration

NVIDIA UFM Cyber-AI Documentation v2.3.0

NVIDIA Morpheus is an open AI application framework that provides cybersecurity developers with a highly optimized AI developer framework and pre-trained AI capabilities that, for the first time, allows them to inspect all IP traffic across their data center fabric instantaneously. Bringing a new level of security to data centers, Morpheus provides development capabilities around dynamic protection, real-time telemetry, adaptive policies, and cyber defenses for detecting and remediating cybersecurity threats.

The Morpheus Developer Kit allows developers to quickly and easily set up an example pipeline to run inference models provided by NVIDIA and experiment with the features and capabilities available within the Morpheus framework to address their cybersecurity and information security use cases.

  • Built on RAPIDS™

    • Built on the RAPIDS™ libraries, deep learning frameworks, and NVIDIA Triton™ Inference Server, Morpheus simplifies the analysis of logs and telemetry to help detect and mitigate security threats.

  • AI Cybersecurity Capabilities

    • Deploy your models using common deep-learning frameworks. Or get a jump-start in building applications to identify leaked sensitive information, detect malware, and identify errors via logs by using one of NVIDIA’s pre-trained and tested models.

  • Real-Time Telemetry

    • Morpheus can receive rich, real-time network telemetry from every NVIDIA® BlueField® DPU-accelerated server in the data center without impacting performance. Integrating the framework into a third-party cybersecurity offering brings the world’s best AI computing to communication networks.

  • DPU-Connected

    • The NVIDIA BlueField Data Processing Unit (DPU) can be used as a telemetry agent for receiving critical data center communications into Morpheus. As an optional addition to Morpheus, BlueField DPU extends static security logging to a sophisticated dynamic real-time telemetry model that evolves with new policies and threat intelligence.

  1. A Cyber AI machine with T4 or V100 GPU, at least 64GB RAM, eight cores CPU, and 100 GB storage.

  2. Morpheus tarball which contains Morpheus AI Engine Docker image.

  3. Installing Docker engine.

The Integration involves installing and starting the Morpheus AI Engine.

Morpheus tarball is available through this link.

Morpheus tarball Components:

  • Installer and Uninstaller Scripts.

  • The configuration file contains the Morpheus docker image details.

  • Morpheus docker image.

  • Machine Learning models files.

To Integrate Morpheus with CyberAI, follow the next steps:

  • Decompress the morpheus-22.06.tar

  • Run the installer script sh.

  • Installer script loads the Morpheus docker image and enables Morpheus in cfg

    1. Load Morpheus docker image morpheus-22.06.tar.gz

    2. Set [Morpheus] enabled = true inside cfg

    3. Enable Telemetry GPU counters collection by setting [data_prep_telemetry::gpu_counter] skip_collection = false

    4. Copy the models' files under the volumes created for Morpheus.

      Copy
      Copied!
                  

      /opt/ufm/cyber-ai/scripts/e2e_model_script.py /opt/ufm/cyber-ai/datastore/morpheus/output/random_forest_model_crypto_resnet.pkl

After installing the Morpheus AI Engine, restarting Cyber AI creates a Morpheus docker container, which stores GPU Telemetry in a shared volume accessed by the Morpheus docker container, where you can run the ML model and inference Crypto-Mining activities and generate output files with events.

© Copyright 2023, NVIDIA. Last updated on Sep 5, 2023.