Menu

Out-of-Band Management for Compute Systems on the Edge

September 13, 2022

OOB System Management

By Ray DeMoss, Support Manager

Out-of-Band (OOB) system management is a staple of the modern datacenter. The tools used by administrators have matured over the last 20 years to provide a variety of mission critical features for managing, monitoring, configuring, and updating computer systems independent of the host system’s resources. Intel led the way in the early days by defining a set of interface specifications for basic system management in 1998 called the Intelligent Platform Management Interface (IPMI). Over the next 20 years, 200+ computer vendors joined support for the IPMI specifications. While IPMI has served us well over the years, it is showing its age with a myriad of security concerns and use of legacy standards that make managing modern hybrid infrastructures difficult and complicated.

The successor to IPMI is a new standard suite of specifications called Redfish, first published by the Distributed Management Task Force (DMTF) in 2015. Redfish is an Application Program Interface (API) specification designed to encourage the use of modern protocols, intuitive data structures, and solid security methods that are implemented using HTTPS, JSON, and a commonly known REST API software architecture. The adoption of Redfish has support from all major hardware vendors.

Edge Computing

Most servers use a Baseboard Management Console (BMC) network endpoint for their OOB management. This is the component that monitors the system’s sensor information and creates alerts for critical conditions. Accessing this content is commonly done by pointing your browser to the BMC hostname or IP address and connecting to its web management interface. Administrating endpoint systems in this way does not scale well if there are hundreds of devices or more. The larger the IT environment, the more dependent administrators become on centralized management systems. Centralized servers consume updates from endpoint systems, then organize and analyze the data into usable content for administrators to view and take actions as needed. The major vendors like IBM, HPE, Dell and Lenovo understand this and have many years of experience developing their various system management suite of tools.

Redfish implementations use the common HTTP client/server connection to transfer data from the endpoint device, like the BMC, to a centralized server. The network payload for these updates is small using text formatted as JSON. Small packet sizes and low network utilization are ideal for remote and mobile systems where bandwidth may be limited. High Performance Computing (HPC) systems on the Edge in the recent past have had limited OOB support. This is changing as 5G network convergence improves and remote and mobile endpoints at the Edge are getting connected. Instead of abandoning that content, the use of Redfish standards is increasing the availability of OOB management at the Edge. As the quality and quantity of information improves, there will be new findings about how these environments differ from the datacenters and drive innovation to ruggedization for mobile and Edge cooling, controls, and vibration issues.

U-BMC

One Stop System (OSS) recently introduced a Redfish compliant Unified-BMC (U-BMC) component to our server and expansion systems. This means that OSS products can integrate into any existing OOB management tools that are also compliant with Redfish standards. We are also immensely proud of our easy-to-use U-BMC management web-interface. This tool is not just an endpoint management tool, it can connect to other U-BMC devices and manage them as well. This feature reduces the number of endpoint logins necessary to administrate systems. This will not replace centralized management in large IT environments but can improve ease of management for small and medium-sized IT environments.

U-BMC is planned for our Rigel Edge Supercomputer, Intel and AMD Short Depth Servers (SDS), the Gen5 4U Professional Accelerator System and many more future products. OSS is excited to bring these products to market with modern and competitive capabilities for OOB management in mobile and Edge servers and PCIe Expansion.

Click the buttons below to share this blog post!

Return to the main Blog page




Leave a comment

Comments will be approved before showing up.


Also in One Stop Systems Blog

The Challenges of Designing with PCIe Gen 5 for AI Transportables
The Challenges of Designing with PCIe Gen 5 for AI Transportables

November 22, 2022

In the world of digital computing, the fundamental determinant of performance is the ability to distinguish between an electrical signal representing a 1 or a 0.  The speed at which these transitions can be recognized on a system’s internal PCI Express (PCIe) interconnect determines the bandwidth of data that can be transmitted and acted upon. With PCIe Gen5, transitions from 1’s to 0’s must be recognized 32 billion times per second.  

Continue Reading

VPX Takes a Back Seat to PCIe and NVLink for Military AI
VPX Takes a Back Seat to PCIe and NVLink for Military AI

November 15, 2022

The need to keep US and allied troops out of harm’s way, while still pursuing battlefield superiority, increasingly requires a need for battlefield assets throughout the military theater to become fully autonomous.  Currently, most unmanned military vehicles are controlled remotely, but the military is expanding the role of autonomy within surface ships, submarine vessels, aircraft, and land vehicles to identify and take action on current and future threats. 

Continue Reading

Features and Benefits of the Rigel Edge Supercomputer
[VIDEO] Features and Benefits of the Rigel Edge Supercomputer

November 08, 2022

In this video, Tom Fries, Government Sales Manager at OSS, does a quick walkthrough of the Rigel Edge Supercomputer. The Rigel Edge Supercomputer brings the power of NVIDIA® HGX™ A100 SXM GPUs to the rugged edge. The HGX A100 4-GPU backplane delivers 78 teraFLOPS of FP64 HPC performance using third generation NVIDIA NVLink™ technology. The GPUs are integrated with OSS PCIe Gen 4.0 expansion technology to take advantage of the latest AMD 3rd Gen EPYC processors while offering four PCIe Gen 4.0 x16 expansion slots for high-speed network interconnect, NVMe storage, or FPGA sensor capture. 

Continue Reading

You are now leaving the OSS website