skip to main content


Title: Leveraging SONiC Functionalities in Disaggregated Network Switches
Ever since the inception of the networking industry, routing and switching devices have been limited to tightly-coupled hardware and software components. Vendors provide closed source proprietary stacks, restraining network operators from utilizing customized features, and hence hindering innovation. This aggregated model is costly, time consuming, and unscalable as changes in the devices require vendor's intervention. As a result, the industry started manufacturing white-box switches and developing Network Operating Systems (NOSs) that support multiple vendors and Application Specific Integrated Circuits (ASICs). This model is referred to as ”disaggregated” as the software and hardware are decoupled; essentially, vendors' switching silicons (e.g., Broadcom) are compatible with different NOS (e.g., SONiC). In this paper, we discuss the lessons learned while designing and implementing a testbed that consists of disaggregated network devices. We iterate over several open source Internet Protocol (IP) routing suites and NOSs that are vendor-agnostic. Additionally, we highlight a novel type of forwarding data planes that are programmable and explore their features. The testbed consists of two white-box switches provided by Edgecore that use programmable switching silicon (Tofino) manufactured by Barefoot Networks, an Intel Company. We installed SONiC NOS on top of the switches and tested static and BGP routing protocols. We report the configuration process and the prerequisites needed to deploy a working disaggregated environment. Finally, we discuss how open source NOSs and programmable switches can be extended to support campus networks, rather than being data center-oriented only.  more » « less
Award ID(s):
1925484
NSF-PAR ID:
10252946
Author(s) / Creator(s):
; ; ;
Date Published:
Journal Name:
2020 43rd International Conference on Telecommunications and Signal Processing (TSP)
Page Range / eLocation ID:
457 to 460
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Network monitoring and measurement have always been critical components of network management. Recent developments in sketch-based monitoring techniques and the deployment opportunities arising from the increasing programmability of network elements (e.g., programmable switches, SmartNICs, and software switches) have made the possibility of accurate, detailed, network-wide telemetry tantalizingly within reach. However, the wide heterogeneity of the programmable hardware and dynamic changes in both resources available and resources needed for monitoring over time make existing approaches to network-wide monitoring impractical. We present HeteroSketch, a framework that consists of two main components: (1) a profiling tool that automatically quantifies the capabilities of arbitrary hardware by predicting their performance for sketching algorithms, and (2) an optimization framework that decides placement of measurement tasks and resource allocation for devices to meet monitoring goals while considering heterogeneous device capabilities. HeteroSketch enables optimized deployments for large networks (> 40,000 nodes) using a novel clustering approach and enables prompt responses to network topology, traffic, query, and resource dynamics. Our evaluation shows that HeteroSketch reduces resource overheads by 20-60% compared to prior art, while maintaining monitoring performance, coverage, and accuracy. 
    more » « less
  2. Network monitoring and measurement have always been critical components of network management. Recent developments in sketch-based monitoring techniques and the deployment opportunities arising from the increasing programmability of network elements (e.g., programmable switches, SmartNICs, and software switches) have made the possibility of accurate, detailed, network-wide telemetry tantalizingly within reach. However, the wide heterogeneity of the programmable hardware and dynamic changes in both resources available and resources needed for monitoring over time make existing approaches to network-wide monitoring impractical. We present HeteroSketch, a framework that consists of two main components: (1) a profiling tool that automatically quantifies the capabilities of arbitrary hardware by predicting their performance for sketching algorithms, and (2) an optimization framework that decides placement of measurement tasks and resource allocation for devices to meet monitoring goals while considering heterogeneous device capabilities. HeteroSketch enables optimized deployments for large networks (> 40,000 nodes) using a novel clustering approach and enables prompt responses to network topology, traffic, query, and resource dynamics. Our evaluation shows that HeteroSketch reduces resource overheads by 20−60% compared to prior art, while maintaining monitoring performance, coverage, and accuracy. 
    more » « less
  3. Network monitoring and measurement have always been critical components of network management. Recent developments in sketch-based monitoring techniques and the deployment opportunities arising from the increasing programmability of network elements (e.g., programmable switches, SmartNICs, and software switches) have made the possibility of accurate, detailed, network-wide telemetry tantalizingly within reach. However, the wide heterogeneity of the programmable hardware and dynamic changes in both resources available and resources needed for monitoring over time make existing approaches to network-wide monitoring impractical. We present HeteroSketch, a framework that consists of two main components: (1) a profiling tool that automatically quantifies the capabilities of arbitrary hardware by predicting their performance for sketching algorithms, and (2) an optimization framework that decides placement of measurement tasks and resource allocation for devices to meet monitoring goals while considering heterogeneous device capabilities. HeteroSketch enables optimized deployments for large networks (> 40,000 nodes) using a novel clustering approach and enables prompt responses to network topology, traffic, query, and resource dynamics. Our evaluation shows that HeteroSketch reduces resource overheads by 20-60% compared to prior art, while maintaining monitoring performance, coverage, and accuracy. 
    more » « less
  4. This paper presents Virginia Tech’s wireless testbed supporting research on long-term evolution (LTE) signaling and radio frequency (RF) spectrum coexistence. LTE is continuously refined and new features released. As the communications contexts for LTE expand, new research problems arise and include operation in harsh RF signaling environments and coexistence with other radios. Our testbed provides an integrated research tool for investigating these and other research problems; it allows analyzing the severity of the problem, designing and rapidly prototyping solutions, and assessing them with standard-compliant equipment and test procedures. The modular testbed integrates general-purpose software-defined radio hardware, LTE-specific test equipment, RF components, free open-source and commercial LTE software, a configurable RF network and recorded radar waveform samples. It supports RF channel emulated and over-the-air radiated modes. The testbed can be remotely accessed and configured. An RF switching network allows for designing many different experiments that can involve a variety of real and virtual radios with support for multiple-input multiple-output (MIMO) antenna operation. We present the testbed, the research it has enabled and some valuable lessons that we learned and that may help designing, developing, and operating future wireless testbeds. 
    more » « less
  5. Baseband processing is one of the most time-consuming and computationally expensive tasks in radio access networks (RANs), which is typically realized in dedicated hardware. The concept of virtualizing the RAN functions by moving their computation to edge data centers can significantly reduce the deployment cost and enable more flexible use of the network resources. Recent studies have focused on software-based baseband processing for large-scale sub-6 GHz MIMO systems, while 5G also embraces the millimeter-wave (mmWave) frequency bands to achieve further improved data rates leveraging the widely available spectrum. Therefore, it is important to build a platform for the experimental investigation of software-based baseband processing for mmWave MIMO systems. In this paper, we implement programmable mmWave MIMO radios equipped with real-time baseband processing capability, leveraging the open-access PAWR COSMOS testbed. We first develop Agora-UHD, which enables UHD-based software-defined radios (SDRs) to interface with Agora, an open-source software realization of real-time massive MIMO baseband processing. Next, we integrate Agora-UHD with the USRP SDRs and IBM 28 GHz phased array antenna module (PAAM) subsystem boards deployed in the PAWR COSMOS testbed. We demonstrate a 2×2 28 GHz polarization MIMO link with a bandwidth of 122.88 MHz, and show that it can meet the real-time processing deadline of 0.375 ms (3 transmission time intervals for numerology 3 in 5G NR FR2) using only 8 CPU cores. The source code of Agora-UHD and its integration with the programmable 28 GHz radios in the COSMOS testbed with example tutorials are made publicly available. 
    more » « less