Programmable Network Devices Research Articles

Databases on Modern Networks: A Decade of Research That Now Comes into Practice

Modern cloud networks are a fundamental pillar of data-intensive applications. They provide high-speed transaction (packet) rates and low overhead, enabling, for instance, truly scalable database designs. These networks, however, are fundamentally different from conventional ones. Arguably, the two key discerning technologies are RDMA and programmable network devices. Today, these technologies are not niche technologies anymore and are widely deployed across all major cloud vendors. The question is thus not if but how a new breed of data-intensive applications can benefit from modern networks, given the perceived difficulty in using and programming them. This tutorial addresses these challenges by exposing how the underlying principles changed as the network evolved and by presenting the new system design opportunities they opened. In the process, we also discuss several hard-earned lessons accumulated by making the transition first-hand.

Scalable Deep Reinforcement Learning-Based Online Routing for Multi-Type Service Requirements

Emerging applications raise critical QoS requirements for the Internet. The improvements in flow classification technologies, software-defined networks (SDN), and programmable network devices make it possible to fast identify users' requirements and control the routing for fine-grained traffic flows. Meanwhile, the problem of optimizing the forwarding paths for traffic flows with multiple QoS requirements in an online fashion is not addressed sufficiently. To address the problem, we propose DRL-OR-S, a highly scalable online routing algorithm using multi-agent deep reinforcement learning. DRL-OR-S adopts a comprehensive reward function, an efficient learning algorithm, and a novel deep neural network structure to learn appropriate routing strategies for different types of flow requirements. In order to enhance the generalization and scalability, we propose a novel graph-based actor-critic network architecture and a carefully designed input state for DRL-OR-S. To accelerate the training process and guarantee reliability, we further introduce an NN-simulator for efficient offline training and a safe learning mechanism to avoid unsafe routes during the online routing process. We implement DRL-OR-S under SDN architecture and conduct Mininet-based experiments using real network topologies and traffic traces. The results validate that DRL-OR-S can well satisfy the requirements of latency-sensitive, throughput-sensitive, latency-throughput-sensitive, and latency-loss-sensitive flows at the same time, while exhibiting great adaptiveness and reliability under the scenarios of link failure, traffic change, unseen large topology and partial deployment.

PTA: Finding Hard-to-Find Data Plane Bugs

Bugs in network hardware can cause tremendous problems. However, programmable network devices have the potential to provide greater visibility into the internal behavior of devices, allowing us to more quickly find and identify problems. In this paper, we provide a taxonomy of data plane bugs, and use the taxonomy to derive a Portable Test Architecture (PTA) which offers essential abstractions for testing on a variety of network hardware devices. PTA is implemented with a novel data plane design that (i) separates target-specific from target-independent components, allowing for portability, and (ii) allows users to write a test program once at compile time, but dynamically alter the behavior via runtime configuration. We report 12 diverse bugs on different hardware targets, and their associated software, exposed using PTA.

MAP4: A Pragmatic Framework for In-Network Machine Learning Traffic Classification

Self-driving networks guided by machine-learning (ML) algorithms are the driving force for building networks of the future. ML is effective at making inferences about data that is too complex or too unpredictable for humans. The network softwarization enabled by a deep programmability approach opens up new opportunities to deploy ML at the programmable data plane. In this paper, we introduce the MAP4 as a framework that explores the feasibility of mapping ML models in programmable network devices. To achieve this, we rely on the P4 language to deploy a pre-trained model into a programmable switch, utilizing the ML model to accurately classify flows at line rate. Our approach demonstrates that ML models working as classifiers can better fit the data by using the new levels of network programmability from the P4 language. The results showed that with few packets, most of the flows are properly classified. In some use cases, with two packets in the flow, 97% of traffic can be correctly classified, and all classes are properly labeled with a maximum of four packets.

Functional Split of In-Network Deep Learning for 6G: A Feasibility Study

In existing mobile network systems, the data plane (DP) is mainly considered a pipeline consisting of network elements end-to-end forwarding user data traffics. With the rapid maturity of programmable network devices, however, mobile network infrastructure mutates towards a programmable computing platform. Therefore, such a programmable DP can provide in-network computing capability for many application services. In this paper, we target to enhance the data plane with in-network deep learning (DL) capability. However, in-network intelligence can be a significant load for network devices. Then, the paradigm of the functional split is applied so that the deep neural network (DNN) is decomposed into sub-elements of the data plane for making machine learning inference jobs more efficient. As a proof-of-concept, we take a Blind Source Separation (BSS) problem as an example to exhibit the benefits of such an approach. We implement the proposed enhancement in a full-stack emulator and we provide a quantitative evaluation with professional datasets. As an initial trial, our study provides insightful guidelines for the design of the future mobile network system, employing in-network intelligence (e.g., 6G).

Evaluation of Hash Join Operations Performance Executing on SDN Switches: A Cost Model Approach

Distributed database systems store and manipulate data on multiple machines. In these systems, the processing cost of query operations is mainly impacted by the data access latency between machines over the network. With recent technology advances in programmable network devices, the network switches provide new opportunities for dynamically managing the network topology, enabling the data processing on these devices with the same network throughput. In this paper, we explore the programmable network switches in query processing, evaluating the processing performance of a cost model in executing the hash join operation. We assume the storage of the hash table built from outer relation and the materialization of the join probing are made in switches using advanced matching techniques similar to package inspections enabled by Ternary Content-Addressable Memories (TCAM) or SRAM via hashing. Our results show that processing the hash join operation using network switches achieved the best results compared to traditional servers, with an average time reduction of 91.82% (Query-10 from TPC-H) and 96.52% (Query-11 from TPC-H).

A cybersecurity assessment framework for virtual operational technology in power system automation

Assessment of cybersecurity in power system automation (PSA) requires a framework to study and analyze the complex relationships between the cyber-based control systems and power systems. A production system is not ideal or available for such assessment due to potential impacts and disruptions. In this paper,11Preliminary idea of this paper has been presented in part at Winter Simulation Conference 2018, Gothenburg, Sweden, December 2018. we propose a framework constituents of power system, process network, communication network, automation network, and enterprize network for cybersecurity assessment in PSA. Both real and virtual components supportability are incorporated in the framework for covering good enough functionalities of power systems maintaining hardware diversity and scalability. A specific instance of the proposed framework, Virtual Operational Technology Network (VOTNet), is presented explaining simulated and emulated systems such as programmable logic controllers (PLCs), network devices, computing systems, software, and tools. The VOTNet consists of a centralized control center deployed with computing devices, an emulated communication network, a substation control center, and power systems. We evaluate and assess the VOTNet for cybersecurity and scalability issues and its cyber–physical impacts under different cyberattacks such as unauthorized access, denial of service (DoS), modbus protocol scanning and data reading, data manipulation/injection, and session hijacking. We also present risk assessment and mitigation against all the demonstrated attacks. Situational awareness and coordination under cyberattacks are also demonstrated. Finally, the usefulness of a virtual testbed in terms of different research applications and lessons learnt from its usage are also presented.

Scalable Traffic Control Using Programmable Data Planes in a Space Information Network

The past several decades have witnessed wide deployment of space information networks (SINs). More than 8000 satellites have been launched into Earth's orbit recently. Meanwhile, the space communications requirements have grown by a factor of 10 in the last decade. The sharply increasing traffic volume puts tremendous pressure on the network traffic control in SINs. Current end-host-based schemes have to collect a massive amount of data (e.g., explicit congestion notification) to respond to one network event (e.g., topology change, network congestion), which is unacceptable in SINs due to the large end-to-end delay. Recently, with the emergence of programmable network devices, it has become possible to perform flexible control strategies as the traffic is flowing through the network (i.e., in-network traffic control). By implementing the control directly inside the network, the in-network schemes are more effective at scale and more responsive to network dynamics. Therefore, in this article, we design an in-network traffic control powered SIN architecture to enhance the network performance. In addition, we present two use cases, in-network load balance and in-network congestion control, to demonstrate the feasibility of our architecture. Extensive simulations are performed to evaluate our proposed schemes.

The Programmable Data Plane

Programmable data plane technologies enable the systematic reconfiguration of the low-level processing steps applied to network packets and are key drivers toward realizing the next generation of network services and applications. This survey presents recent trends and issues in the design and implementation of programmable network devices, focusing on prominent abstractions, architectures, algorithms, and applications proposed, debated, and realized over the past years. We elaborate on the trends that led to the emergence of this technology and highlight the most important pointers from the literature, casting different taxonomies for the field, and identifying avenues for future research.

Data Stream Processing for Packet-Level Analytics.

One of the most challenging tasks for network operators is implementing accurate per-packet monitoring, looking for signs of performance degradation, security threats, and so on. Upon critical event detection, corrective actions must be taken to keep the network running smoothly. Implementing this mechanism requires the analysis of packet streams in a real-time (or close to) fashion. In a softwarized network context, Stream Processing Systems (SPSs) can be adopted for this purpose. Recent solutions based on traditional SPSs, such as Storm and Flink, can support the definition of general complex queries, but they show poor performance at scale. To handle input data rates in the order of gigabits per seconds, programmable switch platforms are typically used, although they offer limited expressiveness. With the proposed approach, we intend to offer high performance and expressive power in a unified framework by solely relying on SPSs for multicores. Captured packets are translated into a proper tuple format, and network monitoring queries are applied to tuple streams. Packet analysis tasks are expressed as streaming pipelines, running on general-purpose programmable network devices, and a second stage of elaboration can process aggregated statistics from different devices. Experiments carried out with an example monitoring application show that the system is able to handle realistic traffic at a 10 Gb/s speed. The same application scales almost up to 20 Gb/s speed thanks to the simple optimizations of the underlying framework. Hence, the approach proves to be viable and calls for the investigation of more extensive optimizations to support more complex elaborations and higher data rates.

In-Network Computing Powered Mobile Edge: Toward High Performance Industrial IoT

Recently, the industrial Internet of Things (IoT) has quickly become a disruptive force reshaping how we live and work. Compared to the consumer Internet, the industrial IoT puts forward much higher performance requirements in terms of network and computing capacity. The industrial IoT system needs to process tons of data generated by millions of IoT sensors in real-time. Recently, with the advent of programmable network devices (e.g., SmartNIC, programmable switch), the in-network computing (INC) paradigm has received a large amount of attention. INC refers to offload application-specific tasks from end-host to network devices. Benefiting from the line-rate processing capacity of network devices, the INC paradigm presents superior performance in terms of high-throughput low-latency computing. Therefore, INC is considered a promising technique to meet performance requirements in the industrial IoT. However, while network devices are capable of performing application-specific tasks, they are still far from the universal computing platform. In this article, we propose an INC powered mobile edge architecture, where we offload lightweight critical tasks to the INC devices and leave the rest to the MEC. To identify critical tasks, we introduce the complex event processing (CEP) tool in our architecture. Also, we present two industrial IoT use cases to evaluate the feasibility of our architecture.

Green SDN: Trends of Energy Conservation in Software Defined Network

Because of the exponential growth of the data in the Internet, storing and computing the data become a challenging issue. Therefore the data center networks are used to provide the infrastructure for storing and computing the data. Most of the network devices present in the data center network is comparatively idle for most of the time and resulting in a waste of energy. To overcome it, the Software Defined Network is proposed which allows the administrators to manage the network devices using a centralized controller and enables programmable network devices. It optimizes the utilization of network resources efficiently and results in significant amount of energy saving. There are several approaches proposed for optimizing the energy in the Software Defined Network which takes a network towards green energy and lower carbon print. This paper presents a survey on energy conservation techniques used in the software defined networks which makes the networks more capable and productive. We also provide a brief comparison of possible energy conservation techniques and guidelines for future research.

IncBricks

The emergence of programmable network devices and the increasing data traffic of datacenters motivate the idea of in-network computation. By offloading compute operations onto intermediate networking devices (e.g., switches, network accelerators, middleboxes), one can (1) serve network requests on the fly with low latency; (2) reduce datacenter traffic and mitigate network congestion; and (3) save energy by running servers in a low-power mode. However, since (1) existing switch technology doesn't provide general computing capabilities, and (2) commodity datacenter networks are complex (e.g., hierarchical fat-tree topologies, multipath communication), enabling in-network computation inside a datacenter is challenging. In this paper, as a step towards in-network computing, we present IncBricks, an in-network caching fabric with basic computing primitives. IncBricks is a hardware-software co-designed system that supports caching in the network using a programmable network middlebox. As a key-value store accelerator, our prototype lowers request latency by over 30% and doubles throughput for 1024 byte values in a common cluster configuration. Our results demonstrate the effectiveness of in-network computing and that efficient datacenter network request processing is possible if we carefully split the computation across the different programmable computing elements in a datacenter, including programmable switches, network accelerators, and end hosts.

IncBricks

The emergence of programmable network devices and the increasing data traffic of datacenters motivate the idea of in-network computation. By offloading compute operations onto intermediate networking devices (e.g., switches, network accelerators, middleboxes), one can (1) serve network requests on the fly with low latency; (2) reduce datacenter traffic and mitigate network congestion; and (3) save energy by running servers in a low-power mode. However, since (1) existing switch technology doesn't provide general computing capabilities, and (2) commodity datacenter networks are complex (e.g., hierarchical fat-tree topologies, multipath communication), enabling in-network computation inside a datacenter is challenging. In this paper, as a step towards in-network computing, we present IncBricks, an in-network caching fabric with basic computing primitives. IncBricks is a hardware-software co-designed system that supports caching in the network using a programmable network middlebox. As a key-value store accelerator, our prototype lowers request latency by over 30% and doubles throughput for 1024 byte values in a common cluster configuration. Our results demonstrate the effectiveness of in-network computing and that efficient datacenter network request processing is possible if we carefully split the computation across the different programmable computing elements in a datacenter, including programmable switches, network accelerators, and end hosts.

Towards an automated framework to instantiate virtual networks in OpenFlow-based infrastructures

The explosion of cloud applications and data-centres technologies has recently renewed the interest of both academic and industrial research community on network virtualisation techniques. The increasing availability of programmable network devices and the pervasive adoption of a software-defined networking approach in all segments of the network are paving the way toward novel virtualisation approaches which aim at improving network management operations while guaranteeing a better utilisation of resources especially when compared to traditional overlay-based techniques widely used in data-centre settings. In this paper, the experience gained by the authors in the development and deployment of an innovative network virtualisation framework on an OpenFlow-based programmable experimental facility is provided and discussed in detail. Despite its specific application to a Future Internet testbed scenario, the proposed architecture is a first step toward a completely automatic management of virtual networks in OpenFlow-based software-defined networks.

Towards an automated framework to instantiate virtual networks in OpenFlow-based infrastructures

The explosion of cloud applications and data-centres technologies has recently renewed the interest of both academic and industrial research community on network virtualisation techniques. The increasing availability of programmable network devices and the pervasive adoption of a software-defined networking approach in all segments of the network are paving the way toward novel virtualisation approaches which aim at improving network management operations while guaranteeing a better utilisation of resources especially when compared to traditional overlay-based techniques widely used in data-centre settings. In this paper, the experience gained by the authors in the development and deployment of an innovative network virtualisation framework on an OpenFlow-based programmable experimental facility is provided and discussed in detail. Despite its specific application to a Future Internet testbed scenario, the proposed architecture is a first step toward a completely automatic management of virtual networks in OpenFlow-based software-defined networks.

Scaling network services using programmable network devices

The NEon system offers an integrated approach to architecting, operating, and managing network services. NEon uses policy rules defining the operation of individual network services and produces a unified set of rules that generic packet-processing engines enforce.

An Extensible, Programmable, Commercial-Grade Platform for Internet Service Architecture

With their increasingly sophisticated applications, users promote the notion that there is more to a network (be it an intranet, or the Internet) than mere L1-3 connectivity. In what shapes a next generation service contract between users and the network, users want the network to offer services that are as ubiquitous and dependable as dial tones. Typical services include application-aware firewalls, directories, nomadic support, virtualization, load balancing, alternate site failover, etc. To fulfill this vision, a service architecture is needed. That is, an architecture wherein end-to-end services compose, on-demand, across network domains, technologies, and administration boundaries. Such an architecture requires programmable mechanisms and programmable network devices for service enabling, service negotiation, and service management. The bedrock foundation of the architecture, and also the key focus of the paper, is an open-source programmable service platform that is explicitly designed to best exploit commercial-grade network devices. The platform predicates a full separation of concerns, in that control-intensive operations are executed in software, whereas, data-intensive operations are delegated to hardware. This way, the platform is capable of performing wire-speed content filtering, and activating network services according to the state of data and control flows. The paper describes the platform and some distinguishing services realized on the platform.