Abstract-Modern data centres provide large aggregate network capacity and multiple paths among servers. Traffic is very diverse; most of the data is produced by long, bandwidth hungry flows but the large majority of flows, which commonly come with strict deadlines regarding their completion time, are short. It has been shown that TCP is not efficient for any of these types of traffic in modern data centres. More recent protocols such MultiPath TCP (MPTCP) are very efficient for long flows, but are ill-suited for short flows.In this paper, we present AMPTCP, a novel transport protocol which, compared to TCP and MPTCP, reduces short flows' completion times, while providing excellent goodput to long flows. To do so, AMPTCP runs in two phases; initially, it randomly scatters packets in the network under a single congestion window exploiting all available paths. This is beneficial to latency-sensitive flows. After a specific amount of data is sent, AMPTCP switches to a regular MultiPath TCP mode. AMPTCP is incrementally deployable in existing data centres as it does not require any modifications outside the transport layer and behaves well when competing with legacy TCP and MPTCP flows. Our extensive experimental evaluation in simulated FatTree topologies shows that all design objectives for AMPTCP are met.
MPTCP and its ECN-capable variants such as XMP and DCM have recently been introduced to effectively exploit the path diversity of modern data center networks (DCNs). Although these multipath schemes improve overall network throughput compared to single-path schemes due to their fast, host-based, load balancing ability, they failed to address the following two problems: TCP incast and last hop unfairness. Firstly, these mechanisms cause frequent TCP incast collapses when used for workloads with a many-to-one communication pattern, commonly found in DCNs. Secondly, the last hop unfairness problem severely violates network fairness as single-path flows achieve 2-5 times less throughput than multipath flows.To effectively tackle these problems, we propose the Adaptive MultiPath (AMP) congestion control mechanism that quickly detects the onset of these problems and transforms its multipath flow into a single-path flow. Once these problems disappear, AMP safely reverses this transformation and continues data transmission via multiple paths. Our evaluation results under a diverse set of scenarios in a large-scale fat-tree topology demonstrate that AMP is robust to the TCP incast problem and improves network fairness between multipath and single-path flows significantly with no performance loss.
In recent years several multipath data transport mechanisms, such as MPTCP and XMP, have been introduced to effectively exploit the path diversity of data center networks (DCNs). However, these multipath schemes have not been widely deployed in DCNs. We argue that two key factors among others impeded their adoption: TCP incast and minimum window syndrome. First, these mechanisms are ill-suited for workloads with a many-to-one communication pattern, commonly found in DCNs, causing frequent TCP incast collapses. Second, the syndrome we discover for the first time, results in 2-5 times lower throughput for single-path flows than multipath flows, thus severely violating network fairness.To effectively tackle these problems, we propose AMP: an adaptive multipath congestion control mechanism that quickly detects the onset of these problems and transforms its multipath flow into a single-path flow. Once these problems disappear, AMP safely reverses this transformation and continues its data transmission via multiple paths. Our evaluation results under a diverse set of scenarios in a fat-tree topology with realistic workloads demonstrate that AMP is robust to the TCP incast problem and improves network fairness between multipath and single-path flows significantly with little performance loss.
Modern data centres provide large aggregate network capacity and multiple paths among servers. Traffic in data centres is very diverse; most of the data is produced by long, bandwidth hungry flows but the large majority of flows, which commonly come with stringent deadlines regarding their completion time, are short. It has been shown that TCP is not efficient for any of these types of traffic in modern data centres. MultiPath TCP (MPTCP) employs multipath data transport and is efficient for long flows but ill-suited for short flows. In this paper, we present Maximum MultiPath TCP (MMPTCP), a novel transport protocol which extends MPTCP and, compared to TCP and MPTCP, reduces short flows' completion times, while providing excellent goodput to long flows. To do so, MMPTCP runs in two phases; initially, it randomly scatters packets in the network under a single congestion window exploiting all available paths. This is beneficial to latency-sensitive flows. After a specific amount of data is sent, MMPTCP switches to a regular MultiPath TCP mode. MMPTCP is incrementally deployable in existing data centres as it does not require any modifications outside the transport layer and behaves well when competing with MPTCP flows. We also present a topology-specific extension of MMPTCP that adjusts the numbers of subflows during the second phase of the protocol based on knowledge about the location of the receiver in the data centre. We present extensive evaluation that shows that MMPTCP's design objectives are met. We have implemented MMPTCP (along with MPTCP and packet spraying) in ns-3 and evaluated our protocol in simulated FatTree topologies. We have evaluated how MMPTCP performs compared to TCP and MPTCP and how its performance is affected by transient hotspots in the network. We have also experimented with different thresholds for duplicate acknowledgements and fast retransmissions and shown that MMPTCP performs well when the size of short flows is widely ranged. Finally, we have evaluated how MMPTCP performs under conditions that result in Incast, when different congestion control algorithms are used in its second phase and when varying the overall network load.
Internet protocols have developed significantly over the last 50 years but have reached a point where the further improvements in performance, resilience, security and privacy cannot be achieved by simple incremental changes. This paper proposes a new IP protocol that puts the user's end host at the centre of major algorithmic decisions. It consist of three new mechanisms: a private source routing establishment protocol that allows inter-domain traffic routes to be decided by the user and kept private from the providers whilst allowing for anonymous connections where two node can communicate without knowing the identity/address of the other end point; a mechanism to control reception of packets that mitigates denial-of-service attacks and a new directory system that puts the end user at the core of the decisions enabling anycast and mobility with a pub-sub mechanism with fine grain capabilities for describe resources. These changes allow end nodes to have a much tighter control of how they send and receive their traffic and provide a paradigm shift for the Internet ecosystem.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.