Now Available, Network and Cloud Edge Reference System Architectures Portfolio Release v22.05
Lindsay Hiebert
Product Manager, Invisinet, First Packet® Authentication (FPA) Zero Trust Network Access security technology and capabilities. Where Zero Trust Begins! Let's Connect. Gen AI Expert, CISSP certified, M.B.A.; M.S.;
Now Available, Intel® Network and Cloud Edge Reference System Architectures Portfolio Release v22.05
Please check out the new resources of Network and Cloud Edge Reference System Architectures Portfolio available for the Network and Cloud Edge Reference System Architectures Portfolio Release v22.05
Source: Network and Cloud Edge Reference System Architectures Portfolio Release v22.05
This updated manual provides an overview of the Bare Metal and Virtual Machine Reference Architectures (BMRA and VMRA) release 22.05 content and terminologies, including: the newly supported 4th Generation Intel® Xeon® Scalable processor and new Cloud Native software (e.g., Cloud Native Data Plane (CNDP), Key Management Reference Application (KMRA) v2.1).
Source: Network and Cloud Edge Reference System Architectures Portfolio.
Now available from Intel, on the Intel Network Builders Website that provide a very efficient way to learn about Intel Network and Cloud Reference Architectures, how to deploy solutions and how to take advantage of Intel platforms for optimized operations. The platforms are based on 2nd Generation, 3rd Generation, and 4th Generation Intel® Xeon® Scalable processors plus the Intel® Xeon® D processor. These platforms include the Taylors Falls Reference Design. The platforms integrate Intel® Ethernet Controller 700 Series and 800 Series, Intel® QuickAssist Technology (Intel® QAT), Intel® Server GPU (Graphic Processor Unit), Intel® Optane™ technology, and more.
A brief excerpt of topics covered are listed below.
Experience Kits
Guidelines delivered in the form of—manuals, user guides, application notes, solution briefs, training videos—for best-practice implementation of cloud native and Kubernetes technologies to ease developments and deployments.
Network and Cloud Edge Reference System Architectures Portfolio
A templated system-level blueprint for a range of locations in enterprise and cloud infrastructure with automated deployment tools. The portfolio integrates the latest Intel platforms and cloud-native technologies for multiple deployment models to simplify and accelerate deployments of key workloads across a service infrastructure.
Deployment Model
Provides flexibility to deploy solutions according to IT needs. The portfolio offers two deployment models:
· Container Bare Metal Reference System Architecture (BMRA) – A deployment model of a Kubernetes cluster with containers on a bare metal platform. A special version of BMRA is the BMRA for Storage that supports MinIO Object Storage.
· Virtual Machine Reference System Architecture (VMRA)– A deployment model of a virtual cluster on a physical node. The virtual cluster can be a Kubernetes containers- based cluster.
Configuration Profiles
A prescribed set of components—hardware, software modules, hardware/software configuration specifications—designed for a deployment for specific workloads at a network location (such as Access Edge). Configuration Profiles define the components for optimized performance, usability, and cost per network location and workload needs.1 In addition, generic Configuration Profiles are available for developers’ flexible deployments.
Reference Architecture Flavor
A Reference Architecture deployment using a Configuration Profile.
Ansible Playbook
A set of validated scripts that prepare, configure, and deploy a Reference Architecture Flavor by implementing a Configuration Profile.
Configuration Profile Ansible Scripts
Automates quick, repeatable, and predictive deployments using Ansible playbooks. Various Configuration Profiles and Ansible scripts allow automated installations that are application-ready, depending on the workload and network location.
TERM
DESCRIPTION
Kubernetes Cluster
A deployment that installs at least one worker node running containerized applications. Pods are the components of the application workload that are hosted on worker nodes. Control nodes manage the pods and worker nodes.
Intel Platforms
Prescribes Intel platforms for optimized operations. The platforms are based on 2nd Generation, 3rd Generation, and 4th Generation Intel® Xeon® Scalable processors plus the Intel® Xeon® D processor. These platforms include the Taylors Falls Reference Design. The platforms integrate Intel® Ethernet Controller 700 Series and 800 Series, Intel® QuickAssist Technology (Intel® QAT), Intel® Server GPU (Graphic Processor Unit), Intel® Optane™ technology, and more.
Note: This release of VMRA does not support the Intel Xeon D processor.
A brief list of configuration profiles (excerpt from the source document above).
Configuration Profiles Overview
A Configuration Profile describes a specific hardware and software BOM and configurations, applicable for a specific deployment (Figure 3). Configuration Profiles take into consideration the best-known configuration (BKC) validated by Intel for optimized performance.4
Each Configuration Profile includes installation scripts to deploy the required components for a Reference Architecture Flavor.
Installation scripts are available to deploy the required components for a Reference Architecture Flavor. Each Reference Architecture is built on the following:
Recommended by LinkedIn
• Intel Platform foundation with Intel processors and technologies.
• Hardware BOM optimized for delivering an application at a specific location using a deployment model. For example, to support a UPF workload at the Remote CO, the BMRA deployment is populated with the maximum available network interface cards (NICs).
• Software BOM leverages the Intel platform and enables cloud-native adoption.
• Installation (Ansible) Playbook automates the installation of a Reference Architecture Flavor per a Configuration Profile specification.
This version of the Network and Cloud Edge Reference System Architectures Portfolio includes the following Configuration Profiles that support per network locations.
• On-Premises Edge Configuration Profile (Customer Premises)
• Access Edge (Far Edge)
• Remote Central Office-Forwarding Configuration Profile (Near Edge)
• Regional Data Center Configuration Profile (Regional POP)
Additional Configuration Profiles support non-location-specific deployments: Basic, Full, and the newly introduced Build-Your-Own Configuration Profile.
Note: The BMRA with Storage is available for deployment using a dedicated Storage Configuration Profile.
On-Premises Edge Configuration Profile (Customer Premises)
This Configuration Profile is designed for a small cluster of stationary or mobile server platforms, ranging from one to four servers. Usage scenarios include data collection from sensors, local (edge) processing, and upstream data transmission. Sample locations include hospitals, factory floors, law enforcement, media, cargo transportation, and power utilities.
This Configuration Profile recommends a Kubernetes cluster hardware configuration, software capabilities, and specific hardware and software configurations that typically support enterprise edge workloads used in Smart City deployments, CDN, and Ad-insertion. This Configuration Profile is designed to meet the following:
• Optimized data processing (to gain insights and reduce upstream transmission volumes) is priority
• Power efficiency is important in most use cases
• Automated infrastructure orchestration capabilities are mandatory
Access Edge Configuration Profile (Far Edge)
This Configuration Profile is designed for a small cluster of one to four servers for data collection, aggregation, and, in some cases, data processing for high-speed and low-latency services. These nodes represent a large portion of the total number of network elements for an operator with the equipment possibly being in the outside plant in harsh, minimally controlled temperature cabinets. Use cases are: Internet of Things (IoT), virtual Radio Access Networks (vRAN) for 5G distribution unit (DU), virtual Broadband Network Gateway (vBNG), Multi-access Edge Control (MEC), video security, and Smart City.
This Configuration Profile is designed to meet the following:
• Optimal balance of power and packet performance for the given application
• Platform and network security
• High-speed packet processing for MEC application
• Automated infrastructure orchestration capabilities are mandatory
Remote Central Office-Forwarding Configuration Profile (Near Edge)
This Configuration Profile is designed for clusters ranging from a half rack to a few racks of servers, typically in a pre-existing, repurposed, unmanned structure. Usage scenarios include running latency-sensitive applications near the user (for example, real-time gaming, stock trading, video conferencing).
This Configuration Profile addresses an installation using Kubernetes cluster hardware, software capabilities, and configurations that help enable high performance for packet forwarding. This category could contain workloads, such as UPF, vBNG, vCMTS, and vCDN. This Configuration Profile is designed to meet the following:
• Power efficiency is a requirement
• Resource usage efficiency and sharing are a priority
• Scaling and redundancy/reliability through scale-out; performance is defined at the cluster level
• Multitenancy support may be required
• Operational automation is important (site is unmanned)
2.3.5 Regional Data Center Configuration Profile (Regional POP)
The Regional Data Center consists of a management domain with many racks of servers, typically managed and orchestrated by a single instance of resource orchestration. Usage scenarios include content delivery, media, mobile connectivity, and cloud services. This Configuration Profile is exclusively tailored and defined for Media Visual Processing workloads such as CDN Transcoding. It is designed to meet the following:
• Automation is mandatory due to scale
• High connectivity (in the aggregate - not individual data plane performance)
• Scaling and redundancy/reliability through scale-out
• Multitenancy support may be required
Source: Network and Cloud Edge Reference System Architectures Portfolio Release v22.05
Source: Network and Cloud Edge Reference System Architectures Portfolio.