Fighting the bufferbloat: On the coexistence of AQM and low priority congestion control

Gong, Yixi; Rossi, Dario; Testa, Claudio; Valenti, Silvio; Taht, M. D.

doi:10.1109/infcom.2013.6567153

Cited by 33 publications

(46 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This kind of traffic has been pointed out as a root cause of Bufferbloat. However, the introduction of AQM degrades the Less-thanBest-Effort aspect of this protocol [12], by restoring fairness, as also suggested by our results with TCP Vegas. Therefore, we believe that RED is a good candidate to reduce Internet latency, and has the additional advantage of being well studied.…”

Section: Discussionsupporting

confidence: 78%

Revisiting old friends

Kuhn

Lochin

Mehani

2014

Proceedings of the 2014 ACM SIGCOMM Workshop on Capacity Sharing Workshop

View full text Add to dashboard Cite

We use ns-2 simulations to compare RED's gentle_ mode to CoDel in terms of their ability to reduce the latency for various TCP variants. We use a common dumbbell topology with Pareto background traffic, and measure the packet delays and transmission time of a 10 MB FTP transfer.In our scenarios, we find that CoDel reduces the latency by 87 %, but RED still manages to reduce it by 75 %. However, the use of CoDel results in a transmission time 42 % longer than when using RED. In light of its maturity, we therefore argue that RED could be considered as a good candidate to tackle Bufferbloat.

show abstract

Section: Discussionsupporting

confidence: 78%

Revisiting old friends

Kuhn

Lochin

Mehani

2014

Proceedings of the 2014 ACM SIGCOMM Workshop on Capacity Sharing Workshop

View full text Add to dashboard Cite

show abstract

“…Low priority TCP traffic is tested together with AQM in [17]. The main conclusion was that the use of AQM makes low priority TCP variants to become elevated in priority equal to normal flows.…”

Section: Related Workmentioning

confidence: 99%

“…For comparison we use HRED (Harsh RED) [26] which we built earlier on top of RED. Our study significantly differs from earlier studies on PIE and CoDel performance [28], [46], [47], [17], [18], as we focus on very short-term transient effects that are typical with the traffic patterns of today and do not introduce long-running TCP flows at all. The dynamics during load transients have in general received too little attention in AQM research, whereas steady-state behavior is much better covered.…”

Section: Introductionmentioning

confidence: 96%

Evaluating CoDel, PIE, and HRED AQM techniques with load transients

Järvinen

Kojo

2014

39th Annual IEEE Conference on Local Computer Networks

View full text Add to dashboard Cite

In the past, networks have been mainly optimized for good system throughput but recently achieving low delay has also gained notable traction. Active Queue Management (AQM) has long been recognized necessary for operating Internet routers with shorter standing queues but only limited deployment has occurred. The recent interest in AQM has resulted in new AQM proposals. In this paper we evaluate CoDel (Controlled Delay) and PIE (Proportional Integral controller Enhanced), both being new AQM proposals, and compare the performance against an aggressive RED (Random Early Detection) variant called HRED (Harsh RED). We focus on AQM behavior during load transients typically occurring at the network edge with the common traffic types of today such as Web transactions. We discover that CoDel auto-tuning does not scale well with the load. With the high-end delays experienced, HRED is better than PIE and CoDel when more than a few simultaneous flows share the bottleneck.

show abstract

“…Finally, the coexistence of both AQM and LPCC-based solutions has been recently studied by Gong et al [14]. Ideally, both approaches should be able to coexist transparently.…”

Section: Discussion On Existing Solutionsmentioning

confidence: 99%

Bufferbloat systematic analysis

Cardozo

Silva

Vieira

et al. 2014

2014 International Telecommunications Symposium (ITS)

View full text Add to dashboard Cite

There is a recent interest in the bufferbloat phenomenon as a possible explanation for the increased network latency observed in the Internet. The bufferbloat phenomenon is related to the excessive packet queuing in over-sized buffers inside the network that may lead to network performance degradation. In this context, we observe a lack of experimental results considering the practical aspects of off-the-shelf network devices. Therefore, in this paper, we present a systematic analysis of the bufferbloat phenomenon considering the microscopic view of the insides in the buffer architecture of typical network devices. Our experimental results show that bufferbloat might not be a significant problem in practice. First, the phenomenon is only observed in specific cases. Second, changes made to the queues under the control of the operating system have negligible effects. Finally, recent versions of the Linux kernel avoid bufferbloat, even in the specific cases in which it was observed.Index Terms-Increased latency; network traffic; buffering 978-1-4799-3743-

show abstract

Fighting the bufferbloat: On the coexistence of AQM and low priority congestion control

Cited by 33 publications

References 14 publications

Revisiting old friends

Revisiting old friends

Evaluating CoDel, PIE, and HRED AQM techniques with load transients

Bufferbloat systematic analysis

Contact Info

Product

Resources

About