Network congestion avoidance through Speculative Reservation

Jiang, Nan; Becker, Daniel U.; Michelogiannakis, George; Dally, William J.

doi:10.1109/hpca.2012.6169047

Cited by 58 publications

(30 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [16] and [17] the authors propose several reservation mechanism for dragonflies, which avoid congestion by pre-reserving bandwidth for each flow. Alternative proposals include the use of dynamically allocated side-buffers in the network switches [9].…”

Section: Related Workmentioning

confidence: 99%

Network unfairness in dragonfly topologies

et al. 2016

View full text Add to dashboard Cite

Dragonfly networks arrange network routers in a two-level hierarchy, providing a competitive cost-performance solution for large systems. Nonminimal adaptive routing (adaptive misrouting) is employed to fully exploit the path diversity and increase the performance under adversarial traffic patterns. Network fairness issues arise in the dragonfly for several combinations of traffic pattern, global misrouting and traffic prioritization policy. Such unfairness prevents a balanced use of the resources across the network nodes and degrades severely the performance of any application running on an affected node. This paper reviews the main causes behind network unfairness in dragonflies, including a new adversarial traffic pattern which can easily occur in actual systems and congests all the global output links of a single router. A solution for the observed unfairness is evaluated using age-based arbitration. Results show that age-based arbitration mitigates fairness issues especially when using in-transit adaptive routing. However, when using source adaptive routing, the saturation of the new traffic pattern interferes with the mechanisms employed to detect remote congestion, and the problem grows with the network size. This makes source adaptive routing in dragonflies based on remote notifications prone to reduced performance, even when using age-based arbitration.

show abstract

Section: Related Workmentioning

confidence: 99%

Network unfairness in dragonfly topologies

et al. 2016

View full text Add to dashboard Cite

show abstract

“…CRP uses speculative packets to mitigate the requestgrant round-trip latency under low and medium loads, as proposed by SRP [28]. Each flow initiates packet transmission speculatively without a grant.…”

Section: Protocol Considerationsmentioning

confidence: 99%

“…While ECN can be effective [37,21], it is also considerably sensitive to its configuration parameters which depend on numerous and sometimes unpredictable factors, including the traffic pattern [37]. Past work has also highlighted ECN's slow adaptation to changes in the traffic pattern [15,28]. This is based on ECN's reactive nature, since ECN acts only after congestion forms, which is also true for TCP.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Channel reservation protocol for over-subscribed channels and destinations

Jiang

Becker

Dally

2013

Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis

Self Cite

View full text Add to dashboard Cite

Channels in system-wide networks tend to be over-subscribed due to the cost of bandwidth and increasing traffic demands. To make matters worse, workloads can overstress specific destinations, creating hotspots. Lossless networks offer attractive advantages compared to lossy networks but suffer from tree saturation. This led to the development of explicit congestion notification (ECN). However, ECN is very sensitive to its configuration parameters and acts only after congestion forms. We propose channel reservation protocol (CRP) to enable sources to reserve bandwidth in multiple resources in advance of packet transmission and with a single request, but without idling resources like circuit switching. CRP prevents congestion from ever occurring and thus reacts instantly to traffic changes, whereas ECN requires 300,000 cycles to stabilize in our experiments. Furthermore, ECN may not prevent congestion formed by short-lived flows generated by a large combination of source-destination pairs.

show abstract

“…First, low-latency datacenter-wide networks [22] and faster storage (e.g., Flash) have turned many I/O-bound workloads into compute-bound, often making compute the major source of latency [40,43]. Second, single-node latencies must be small and tightly distributed, since servicing each user request involves hundreds to thousands of nodes, and the slowest nodes often determine end-to-end latency.…”

Section: Background and Related Workmentioning

confidence: 99%

Ubik

Kasture

Sánchez

2014

Proceedings of the 19th International Conference on Architectural Support for Programming Languages and Operating Systems

114

View full text Add to dashboard Cite

Chip-multiprocessors (CMPs) must often execute workload mixes with different performance requirements. On one hand, user-facing, latency-critical applications (e.g., web search) need low tail (i.e., worst-case) latencies, often in the millisecond range, and have inherently low utilization. On the other hand, compute-intensive batch applications (e.g., MapReduce) only need high long-term average performance. In current CMPs, latency-critical and batch applications cannot run concurrently due to interference on shared resources. Unfortunately, prior work on quality of service (QoS) in CMPs has focused on guaranteeing average performance, not tail latency.In this work, we analyze several latency-critical workloads, and show that guaranteeing average performance is insufficient to maintain low tail latency, because microarchitectural resources with state, such as caches or cores, exert inertia on instantaneous workload performance. Last-level caches impart the highest inertia, as workloads take tens of milliseconds to warm them up. When left unmanaged, or when managed with conventional QoS frameworks, shared last-level caches degrade tail latency significantly. Instead, we propose Ubik, a dynamic partitioning technique that predicts and exploits the transient behavior of latency-critical workloads to maintain their tail latency while maximizing the cache space available to batch applications. Using extensive simulations, we show that, while conventional QoS frameworks degrade tail latency by up to 2.3×, Ubik simultaneously maintains the tail latency of latency-critical workloads and significantly improves the performance of batch applications.

show abstract

Network congestion avoidance through Speculative Reservation

Cited by 58 publications

References 23 publications

Network unfairness in dragonfly topologies

Network unfairness in dragonfly topologies

Channel reservation protocol for over-subscribed channels and destinations

Ubik

Contact Info

Product

Resources

About