Virtual switches, used for end-host networking, drop packets when the receiving application is not fast enough to consume them. This is called the slow receiver problem, and it is important because packet loss hurts tail communication latency and wastes CPU cycles, resulting in application-level performance degradation. Further, solving this problem is challenging because application throughput is highly variable over short timescales as it depends on workload, memory contention, and OS thread scheduling.
This paper presents Backdraft, a new lossless virtual switch that addresses the slow receiver problem by combining three new components: (1) Dynamic Per-Flow Queuing (DPFQ) to prevent HOL blocking and provide on-demand memory usage; (2) Doorbell queues to reduce CPU overheads; (3) A new overlay network to avoid congestion spreading. We implemented Backdraft on top of BESS and conducted experiments with real applications on a 100 Gbps cluster with both DCTCP and Homa, a state-of-the-art congestion control scheme. We show that an application with Backdraft can achieve up to 20x lower tail latency at the 99th percentile.
more »
« less
Line-Speed and Scalable Intrusion Detection at the Network Edge via Federated Learning
Intrusion detection through classifying incoming packets is a crucial functionality at the network edge, requiring accuracy, efficiency and scalability at the same time, introducing a great challenge. On the one hand, traditional table-based switch functions have limited capacity to identify complicated network attack behaviors. On the other hand, machine learning based methods providing high accuracy are widely used for packet classification, but they typically require packets to be forwarded to an extra host and therefore increase the network latency. To overcome these limitations, in this paper we propose an architecture with programmable data plane switches. We show that Binarized Neural Networks (BNNs) can be implemented as switch functions at the network edge classifying incoming packets at the line speed of the switches. To train BNNs in a scalable manner, we adopt a federated learning approach that keeps the communication overheads of training small even for scenarios involving many edge network domains. We next develop a prototype using the P4 language and perform evaluations. The results demonstrate that a multi-fold improvement in latency and communication overheads can be achieved compared to state-of the-art learning architectures.
more »
« less
- Award ID(s):
- 1815676
- NSF-PAR ID:
- 10393089
- Date Published:
- Journal Name:
- IFIP
- ISSN:
- 1571-5736
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Advanced high-speed network cards have made packet processing in host operating systems a major performance bottleneck. The kernel network stack gives rise to various sources of overheads that limit the throughput and lengthen the per-packet processing latency. The problem is further exacerbated for short-lived, latency-sensitive network flows such as control packets, online gaming, database requests, etc. — in a highly utilized system, especially in virtualized (containerized) cloud environments, short flows can experience excessively long in-kernel queuing delays. As a consequence, recent research works propose to bypass the kernel network stack to enable lightweight, custom userspace network stacks for improved performance, but at a heavy cost of compatibility and security. In this paper, we take a different approach: We first analyze various sources of inefficiencies in the kernel network stack and propose ways to mitigate them without compromising systems compatibility, security, or flexibility. Further, we propose PRISM, a novel mechanism in the kernel network stack to differentiate incoming packets based on their performance requirements and streamline the processing stages of multi-stage packet processing pipelines (e.g., in container overlay networks). Our evaluation demonstrates that PRISM can significantly improve the latency of high-priority flows in container overly networks in the presence of heavy low-priority background traffic.more » « less
-
The 5G user plane function (UPF) is a critical inter-connection point between the data network and cellular network infrastructure. It governs the packet processing performance of the 5G core network. UPFs also need to be flexible to support several key control plane operations. Existing UPFs typically run on general-purpose CPUs, but have limited performance because of the overheads of host-based forwarding. We design Synergy, a novel 5G UPF running on SmartNICs that provides high throughput and low latency. It also supports monitoring functionality to gather critical data on user sessions for the prediction and optimization of handovers during user mobility. The SmartNIC UPF efficiently buffers data packets during handover and paging events by using a two-level flow-state access mechanism. This enables maintaining flow-state for a very large number of flows, thus providing very low latency for control and data planes and high throughput packet forwarding. Mobility prediction can reduce the handover delay by pre-populating state in the UPF and other core NFs. Synergy performs handover predictions based on an existing recurrent neural network model. Synergy's mobility predictor helps us achieve 2.32× lower average handover latency. Buffering in the SmartNIC, rather than the host, during paging and handover events reduces packet loss rate by at least 2.04×. Compared to previous approaches to building programmable switch-based UPFs, Synergy speeds up control plane operations such as handovers because of the low P4-programming latency leveraging tight coupling between SmartNIC and host.more » « less
-
Vehicle-to-pedestrian communication could significantly improve pedestrian safety at signalized intersections. However, it is unlikely that pedestrians will typically be carrying a low latency communication-enabled device with an activated pedestrian safety application in their hand-held device all the time. Because of this, multiple traffic cameras at a signalized intersection could be used to accurately detect and locate pedestrians using deep learning, and broadcast safety alerts related to pedestrians to warn connected and automated vehicles around signalized intersections. However, the unavailability of high-performance roadside computing infrastructure and the limited network bandwidth between traffic cameras and the computing infrastructure limits the ability of real-time data streaming and processing for pedestrian detection. In this paper, we describe an edge computing-based real-time pedestrian detection strategy that combines a pedestrian detection algorithm using deep learning and an efficient data communication approach to reduce bandwidth requirements while maintaining high pedestrian detection accuracy. We utilize a lossy compression technique on traffic camera data to determine the tradeoff between the reduction of the communication bandwidth requirements and a defined pedestrian detection accuracy. The performance of the pedestrian detection strategy is measured in relation to pedestrian classification accuracy with varying peak signal-to-noise ratios. The analyses reveal that we detect pedestrians by maintaining a defined detection accuracy with a peak signal-to-noise ratio 43 dB while reducing the communication bandwidth from 9.82 Mbits/sec to 0.31 Mbits/sec, a 31× reduction.more » « less
-
Hanus, Michael ; Inclezan, Daniela (Ed.)The development of programmable switches such as the Intel Tofino has allowed network designers to implement a wide range of new in-network applications and network control logic. However, current switch programming languages, like P4, operate at a very low level of abstraction. This paper introduces SwitchLog, a new experimental logic programming language designed to lift the level of abstraction at which network programmers operate, while remaining amenable to efficient implementation on programmable switches. SwitchLog is inspired by previous distributed logic programming languages such as NDLog, in which programmers declare a series of facts, each located at a particular switch in the network. Logic programming rules that operate on facts at different locations implicitly generate network communication, and are updated incrementally, as packets pass through a switch. In order to ensure these updates can be implemented efficiently on switch hardware, SwitchLog imposes several restrictions on the way programmers can craft their rules. We demonstrate that SwitchLog can be used to express a variety of networking applications in a mere handful of lines of code.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- The DOI is not currently available.
