In remote direct memory access (RDMA)-based remote procedure call (RPC) systems, using reliable connection (RC) transport mode with one-sided verbs might face challenges from excessive queue pairs (QPs) and per-client request regions when there is high client concurrency. This can cause cache thrashing on the RDMA network interface card (RNIC) and CPU and restrict server-side scalability. To address this issue, this paper proposesRdmax, a highly-scalable RC-based RPC system incorporating two innovations. First, it leverages QP multiplexity to serve multiple independent clients simultaneously with one RC QP, breaking the one-to-one connection constraint of RC. This also allows to post responses for different clients in one batch. Second, it uses in-network request dispatching to consolidate per-client request regions into a shared one, while still ensuring conflict-free client writes.Rdmaxrealizes these innovations with a programmable network, which manages the states of RC QPs and the request region and modifies relevant fields in RDMA packets accordingly. In a dummy RPC benchmark,Rdmaxachieves up to$10.4\times $higher throughput. For online transaction processing,Rdmaxoutperforms two state-of-the-arts based on client grouping and unreliable datagram schemes, improving throughput by up to 44.3% and 45.7%, while reducing 99% tail latency by up to 66.8% and 40.0%, respectively.

The rapid development of network function virtualization (NFV) has notably increased the implementation of virtual network functions (vNFs), especially the stateful ones, on high-performance programmable data plane (PDP) switches (e.g., those based on P4 and Tofino ASICs). This facilitates offloading stateful service function chains (SFCs) to PDP switches. However, the capability of PDP switches to carry stateful SFCs is still constrained by their limited hardware resources. In this work, we propose to address this issue with adaptive table merging, formulate an integer linear programming (ILP) model to optimize the deployment of stateful SFCs on P4-based PDP switches with table merging, and design a time-efficient heuristic to solve the problem quickly. Simulation results validate that our proposals successfully enhance the deployment efficiency of stateful SFCs, outperforming existing benchmarks significantly.

As network monitoring is crucial for ensuring the performance of network operations, one key challenge is how to optimize the tradeoff between its overheads and accuracy. This paper proposes a smart reporting mechanism, namely SmtRPTG, which continuously optimizes the scheme of network status collecting and reporting to properly balance the tradeoff. SmtRPTG estimates the distributions of status data of various types based on sampled results, and lets network elements make local decisions on whether and what type of status data should be reported based on the estimations. We formulate a probabilistic model with hidden variables, based on which an algorithm is designed to estimate data distributions for SmtRPTG. Extensive simulations verify the effectiveness of our proposal on balancing the tradeoff between overheads and accuracy of network monitoring.

Data plane programmable (PDP) switches are becoming increasingly popular for network function virtualization (NFV), for their programmability and high packet processing performance. However, the inherent limitations of PDP switches, such as limited memory space, make it challenging to implement certain types of network functions (NFs) (i.e., the stateful ones) on them. This paper proposes SFCache, which combines PDP switches and commodity servers to achieve self-adaptive SFC deployment. SFCache aims to exploit the high packet processing performance of PDP switches while supporting the flexible deployment of a wide range of SFCs (including the stateful ones) with servers. Specifically, SFCache can dynamically improve the packet processing performance of the SFCs that were deployed on servers by selectively caching SFC-level packet processing rules on PDP switches. We design a few key components to facilitate SFCache, including an NF-destructed P4 pipeline that allows customizing packet processing rules in a match-rewrite pattern, a runtime NF synthesis method that can transform a set of NF-level match-rewrite rules into an equivalent SFC-level rule, and a count-min selection strategy to choose the best synthesized rule for being cached in PDP switch pipeline. We prototype SFCache with a PDP switch based on Tofino ASIC and a server, and demonstrate the effectiveness of our proposal experimentally.

With the development of programmable data plane (PDP), in-band network telemetry (INT) has become a promising network monitoring technique to visualize network operations in a fine-grained and real-time way. In this work, to better balance the tradeoff between INT overheads and monitoring accuracy, we design and optimize an information-sensitive INT system (namely, P4InfoSen-INT), which makes each PDP switch decide locally whether and what type(s) of telemetry data should be inserted in a packet based on the “information content” of the data, and implement it in P4-based PDP switches. We first realize the basic principle of P4InfoSen-INT with P4 programs. Then, we propose algorithms to estimate the information content of telemetry data accurately in a dynamic network and optimize the tradeoff between INT overheads and monitoring accuracy. Finally, we further optimize the implementation of P4InfoSen-INT by proposing table merging to reduce stage occupation in each switch. Experimental results verify that our proposed P4InfoSen-INT can balance the tradeoff between INT overheads and monitoring accuracy better than existing benchmarks.

Runtime re-optimization of service function chains (SFCs) has been considered as a must-have feature to enable cost-effective SFC provisioning for the multi-access edge computing (MEC) in beyond 5G (B5G) networks. Therefore, in this work, we propose SRv6-INT, which time-multiplexes the fields of segment routing over IPv6 (SRv6) and in-band network telemetry (INT) in packets to monitor and adjust SFCs timely and efficiently. We also prototype a closed-loop network control and management (NC&M) system based on SRv6-INT to re-optimize SFC provisioning in runtime for precisely balancing the tradeoff between quality-of-service (QoS) and energy usage of SFCs. Experimental results show that with SRv6-INT, the provisioning of SFCs can be adjusted in runtime to save switch ports and CPU frequency, leading to effective energy-saving without QoS violation.

In-band network telemetry (INT) can monitor networks in a flow-oriented and realtime way. The bandwidth overheads caused by per-packet INT can be effectively reduced by selective INT (Sel-INT), i.e., sampling packets or/and types of telemetry data for INT. For the sampling in Sel-INT, there is a tradeoff between bandwidth overheads and monitoring accuracy, which can be adjusted by the sampling ratio. In this paper, we investigate how to adjust the sampling ratio of Sel-INT adaptively such that the tradeoff can be balanced precisely. We first develop a theoretical model to analyze the accuracy of Sel-INT for traffic trace reconstruction under different sampling ratios. We assume that the data samples of a flow's traffic trace (i.e., bandwidth usage) follow the Gaussian random process, and the traffic trace is reconstructed with linear interpolation based on the samples collected by Sel-INT. We analyze this process theoretically and derive the expected reconstruction error (RCE) between the original and reconstructed traces. Then, we use RCE to determine whether the sampling ratio of Sel-INT is properly set and propose two algorithms, namely, RCESA and pRCESA, to adjust the sampling ratio adaptively. Extensive simulations with realistic traffic traces verify the effectiveness of our proposal.