Exploring XRd: A Containerized Solution for Network Operations
XRd manifests as a containerized iteration of IOS-XR, enabling users to delve into platform-agnostic functionalities while facilitating deployment on various cloud or on-premises infrastructures. This approach harnesses the advantages inherent in container technology for network management.
Enhanced Programmability and Developer Support
This innovative platform encompasses all the programmability features from IOS-XR, such as Telemetry and YANG models, positioning it as an optimal resource for both developers and network engineers seeking advanced capabilities in their projects.
The XRd sandbox offers an exceptional starting point for those wishing to familiarize themselves with XRd’s environment while investigating the depths of network programmability.
Variants: Control Plane vs. vRouter
XRd is available in two distinct iterations: XRd Control Plane and XRd vRouter. The former suits computational tasks such as Virtual Route Reflectors (vRR) or Path Compute Elements (PCE), while the latter emphasizes efficient traffic forwarding alongside control plane operations.
The sandbox leverages the capabilities of the XRd Control Plane variant, making it ideal for users engaging in exploratory endeavors and experimentation.
Kernel Specifications for Optimal Performance
A prerequisite for efficient operation is proper kernel configuration on the host machine; however, this has already been pre-configured within the sandbox environment, ensuring seamless use upon setup.
Simplified Deployment Strategies
XRD can be deployed using various methods including Terraform, Helm, or Packer. For our specific application within this context, we adopted Docker Compose due to its straightforward management capabilities.
Create Your Topology Using XYZ Technology
This scenario utilizes a sample segment-routing topology sourced from xrd-tools GitHub repository. Our implementation involves a customized compose format titled XR-YAML aimed at deploying within a singular virtual machine through Docker Compose integration.
Modifications Made to Optimize Functionality- IP Addressing Update: We realigned IP configurations pertaining to the management network according to available resources within our testing sandbox environment.
- Laying Down macvlan Driver: Implementation of the macvlan Docker driver granted each individual ARD container its own dedicated IP address instead of sharing one with its host VM unit.
The Deployment Process Made Effortless
()
- Your lab set-up contains all required elements; you only need to construct a new compose file utilizing Docker Compose’s functionality.
- The initial step involves invoking
xr-compose
from xrd-tools which initializes your composition document. - Edit this generated document so that it reflects specifications referring back correctly towards interfaces attributed by your chosen macvlan driver
{
.