Introduction
NVIDIA DOCA™ services are DOCA-based products, wrapped in a container for deployment on the NVIDIA® BlueField® DPU. DOCA services use DPU capabilities to offer telemetry, time synchronization, networking solutions, and more.
DOCA service containers are available in the NGC catalog, labeled "DOCA" and "DPU." You can also find them through the built-in NVIDIA platform option ("DOCA") in the NGC catalog.
The following services are not available in the NGC catalog:
-
DOCA Management Service
-
OVS-DOCA
For information on the deployment of the services, see the DOCA Container Deployment Guide.
DOCA Services Summary and Quality Levels
|
DOCA Service |
Description |
Quality Level |
|---|---|---|
|
DOCA Platform Framework (DPF) |
Provides fleet-scale provisioning and lifecycle management for BlueField DPUs and orchestrates DOCA services and compatible third-party DPU services across the fleet. |
GA |
|
DOCA Argus |
Provides agentless runtime threat detection from BlueField's isolated trust domain. |
Beta |
|
DOCA Blueman |
Provides a web dashboard for BlueField inventory, health, and telemetry. |
GA |
|
DOCA Firefly |
Synchronizes BlueField and host clocks using Precision Time Protocol (PTP). |
GA |
|
DOCA Flow Inspector |
Analyzes mirrored network traffic and sends real-time packet telemetry to DOCA Telemetry Service. |
GA |
|
DOCA HBN |
Provides hardware-accelerated BGP/EVPN routing, ECMP, L4 firewall, and NAT on BlueField. |
GA |
|
DOCA Management Service (DMS) |
Provides centralized configuration and management for BlueField and NVIDIA networking devices. |
Alpha |
|
DPL Service |
Provides a P4-derived language and tools for building custom hardware-accelerated packet-processing pipelines. |
Beta |
|
SNAP Virtio-fs |
Presents networked file storage to hosts as accelerated Virtio-fs devices on BlueField-3. |
Beta |
|
SNAP-4 |
Presents networked block storage to hosts as NVMe or virtio-blk devices on BlueField-3. |
GA |
|
SNAP-3 |
Presents networked block storage to hosts as NVMe or virtio-blk devices on BlueField-2. |
GA |
|
DOCA Telemetry Service (DTS) |
Collects and exports real-time telemetry from BlueField DPUs, hosts, and workloads. |
GA |
|
DOCA UROM |
Offloads HPC and AI parallel-computing operations from hosts to BlueField. |
GA |
|
DOCA Virtio-net |
Provides hardware-accelerated host and VM networking through BlueField-emulated virtio-net devices. |
GA |
|
OVS-DOCA |
Offloads Open vSwitch packet processing to NVIDIA hardware while retaining the standard OVS control plane. |
GA |
|
DOCA XPlane |
Manages failover between local and remote OVS-DOCA planes in Spectrum-X deployments. |
GA |
Development Lifecycle
DOCA-based containers fall into two categories:
-
DOCA Base Images: containerized DOCA environments for both runtime and development. Developers use these images either as their development environment or to containerize a DOCA-based solution.
-
DOCA Services: containerized DOCA-based products.
Development
Before containerizing a product, you must first design and develop it using the same process as for a bare-metal deployment on the BlueField DPU.
This process consists of the following steps:
-
Identify the requirements for the DOCA-based solution.
-
Review the feature set offered by the DOCA SDK libraries, as shown in their respective programming guides.
-
Start the development process by following the best practices detailed in the Developer Guide.
-
Test the developed solution.
After development and testing, you can containerize the product.
Containerization
For this process, NVIDIA recommends using DOCA's base images, available on DOCA's NGC page.
There are three image varieties:
-
base-rt: includes the DOCA runtime, using the most basic runtime environment required by DOCA's SDK. -
full-rt: builds on the previous image and includes the full list of runtime packages, which are all user-mode components that can be found under thedoca-runtimepackage. -
devel: builds on the previous image and adds headers and development tools for developing and debugging DOCA applications. This image is particularly useful for multi-stage builds.
All images are preconfigured to use the DOCA repository of the matching DOCA version. This means that you can install an additional DOCA package as part of a Docker file or within the development container by running the following commands:
apt update
apt install <package-name>,
For DOCA and CUDA environments, the images follow a similar pattern and are combined with CUDA's images:
-
base-rt(DOCA) +base(CUDA) -
full-rt(DOCA) +runtime(CUDA) -
devel(DOCA) +devel(CUDA)
After you containerize the product, you can start profiling it to prepare for a production-grade deployment.
DOCA provides base images for both the DPU and the host. For host-related DOCA base images, refer to the image tag suffixed with -host.
Profiling
As mentioned in the DOCA Container Deployment Guide, the current deployment model of containers on the DPU is based on kubelet-standalone. This Kubernetes-based deployment uses YAML files to describe the resources that the pod requires, such as the following:
-
CPU
-
RAM
-
Huge pages
NVIDIA recommends profiling your product to estimate the resources it requires (under regular deployments, as well as under stress testing) so that the YAML file includes an accurate resources section. An accurate resources section helps administrators better understand the requirements for deploying your service and also enables the Kubernetes infrastructure to ensure that the service is not malfunctioning after deployment.
After the profiling is complete, the containerized DOCA-based product is ready for the final testing and deployment in production environments.
Last updated: