A copy of this work was available on the public web and has been preserved in the Wayback Machine. The capture dates from 2015; you can also visit the original URL.
The file type is application/pdf
.
Filters
Towards minimal-delay deadline-driven data center TCP
2013
Proceedings of the Twelfth ACM Workshop on Hot Topics in Networks - HotNets-XII
This paper presents MCP, a novel distributed and reactive transport protocol for data center networks (DCNs) to achieve minimal per-packet delay while providing guaranteed transmission rates to meet flow ...
To design MCP, we first formulate a stochastic packet delay minimization problem with constraints on deadline completion and network stability. ...
CONCLUSION AND FUTURE WORK We have proposed MCP, a novel transport protocol to minimize packet delay while providing guaranteed rates to meet flow deadlines. ...
doi:10.1145/2535771.2535788
dblp:conf/hotnets/ChenH0WT13
fatcat:r77z2qiiwjcnhones2zalvfeiy
Finishing flows quickly with preemptive scheduling
2012
Proceedings of the ACM SIGCOMM 2012 conference on Applications, technologies, architectures, and protocols for computer communication - SIGCOMM '12
Through extensive packet-level and flow-level simulation, we demonstrate that PDQ significantly outperforms TCP, RCP and D 3 in data center environments. ...
Today's data centers face extreme challenges in providing low latency. ...
Now we consider D 3 , a recently proposed deadline-aware protocol for data center networks [20] . ...
doi:10.1145/2342356.2342389
dblp:conf/sigcomm/HongCG12
fatcat:4rh4g7hdwbeupaxratxkgh6aty
Finishing flows quickly with preemptive scheduling
2012
Computer communication review
Through extensive packet-level and flow-level simulation, we demonstrate that PDQ significantly outperforms TCP, RCP and D 3 in data center environments. ...
Today's data centers face extreme challenges in providing low latency. ...
Now we consider D 3 , a recently proposed deadline-aware protocol for data center networks [20] . ...
doi:10.1145/2377677.2377710
fatcat:zcmnqmctg5d2jkif66msk5b6ea
Low Latency Datacenter Networking: A Short Survey
[article]
2014
arXiv
pre-print
Datacenters are the cornerstone of the big data infrastructure supporting numerous online services. ...
DCTCP DCTCP (Data Center TCP) [1] is arguably the first work targeting at reducing latency in datacenter networks. It lays down the groundwork by making two seminal contributions. Alizadeh et al. ...
Deadline as priority 1) D 3 : D 3 (Deadline Driven Delivery) [11] is an early attempt to perform deadline-aware flow prioritization. ...
arXiv:1312.3455v2
fatcat:qldidvdtsbcqlh3xf37aqg7ggy
Finishing Flows Quickly with Preemptive Scheduling
[article]
2012
arXiv
pre-print
Through extensive packet-level and flow-level simulation, we demonstrate that PDQ significantly outperforms TCP, RCP and D3 in data center environments. ...
Today's data centers face extreme challenges in providing low latency. ...
from (a, b) a production data center of a large commercial cloud service [12] and (c) a university data center located in Midwestern United States (EDU1 in [6] ). duction compared with TCP. ...
arXiv:1206.2057v2
fatcat:naar5tvxmzcczkmdyft4g4r7ly
Survey on Traffic Management in Data Center Network: From Link Layer to Application Layer
2021
IEEE Access
INDEX TERMS Data center network, congestion control, load balancing, RDMA. ...
In the meantime, many surveys have emerged currently to introduce the current research of data center networks. ...
As a result, when clients request data from multiple servers, flows that encounter RTO is delayed by 200ms in a data center network, resulting in TCP throughput collapse. ...
doi:10.1109/access.2021.3064008
fatcat:jnwnjjv65fgbrf2wxnwdyc2c6u
Deadline is not Enough: How to Achieve Importance-aware Server-centric Data Centers via a Cross Layer Approach
[article]
2013
arXiv
pre-print
semantic data of a query. ...
Since both of them are too complicated to be deployed in the real world, we propose the importance-aware delivery protocol, which is a distributed event-driven rate-based delivery control protocol, for ...
Global Optimization in Data Center Networks In a data center network, once a flow cannot meet its deadline, all response units of that flow are regarded as lost. ...
arXiv:1302.0210v1
fatcat:am4fyz5ol5ei3moyvufpfrix4e
SDTCP: Towards Datacenter TCP Congestion Control with SDN for IoT Applications
2017
Sensors
In cloud data center networks (DCN), TCP incast usually happens when multiple senders simultaneously communicate with a single receiver. ...
Therefore, SDTCP is an effective and scalable solution for the TCP incast problem. delay-insensitive, while burst flows consists of short control and web search flows that require low delay [9] . ...
While the previous works are deadline-agnostic and focus on fair-share bandwidth allocation among flows, D 3 [12] , a deadline-driven delivery control protocol, uses explicit rate control to apportion ...
doi:10.3390/s17010109
pmid:28075347
pmcid:PMC5298682
fatcat:o2cead3xbzazpf4myjxwbjan5q
Pangolin
2011
Proceedings of the Seventh COnference on emerging Networking EXperiments and Technologies on - CoNEXT '11
The engine optimally controls the transmission of redundant Forward Error Correction packets to combat data loss. ...
Trace-driven emulation demonstrates that Pangolin reduces the 99.9-percentile latency from more than 4 seconds to about 1 second with negligible overhead. ...
console and the data center. ...
doi:10.1145/2079296.2079319
dblp:conf/conext/ZhangHCLMRCLT11
fatcat:vw3xyx7aqjcujbhucwbpnkwwy4
Timecard
2013
Proceedings of the Twenty-Fourth ACM Symposium on Operating Systems Principles - SOSP '13
These delays include location lookup, sensor data acquisition, radio wake-up, network transmissions, and processing on both the client and server. ...
Experiments with Timecard incorporated into two mobile applications show that the end-to-end delay is within 50 ms of the target delay of 1200 ms over 90% of the time. ...
LR and HB were supported in part by the National Science Foundation under Grant 0931508 and the MIT Center for Wireless Networks and Mobile Computing (Wireless@MIT). ...
doi:10.1145/2517349.2522717
dblp:conf/sosp/RavindranathPMB13
fatcat:vqv3ferphffzbedtuoucp2w7qa
Timely Result-Data Offloading for Improved HPC Center Scratch Provisioning and Serviceability
2011
IEEE Transactions on Parallel and Distributed Systems
window can be minimized while also meeting center-user service level agreements. ...
Modern High-Performance Computing (HPC) centers are facing a data deluge from emerging scientific applications. ...
Furthermore, a delayed offload renders output-data vulnerable to center purge policies. ...
doi:10.1109/tpds.2010.190
fatcat:lacplavzmnfkrcrb3xjhovhdx4
Congestion Control Mechanisms for Inter-Datacenter Networks
[article]
2022
arXiv
pre-print
Firstly, the inter-datacenter communication involves both data center networks (DCNs) and wide-area networks (WANs) connecting multiple data centers. ...
Driven by the data communication need of these applications, inter-datacenter network (IDN) is getting increasingly important. ...
D2TCP [109] and L2DCT [87] modulate the window adjustment function of DCTCP to meet deadlines and minimize FCT, respectively. DCQCN [124] is built on the top of DCTCP and QCN [92] . ...
arXiv:2201.03734v1
fatcat:xrnbhxhh3vgy7o46rueslwgfoe
Packet Slicing for Highly Concurrent TCPs in Data Center Networks with COTS Switches
2015
2015 IEEE 23rd International Conference on Network Protocols (ICNP)
Modern data center nowadays leverages highly concurrent TCP connections between thousands of computer servers to achieve high system performance and service reliability. ...
We integrate Packet Slicing transparently (i.e., without modification) with three state-of-theart TCP protocols designed for data centers on NS2 simulation and a physical testbed, respectively. ...
When congestion occurs, far-deadline flows release some bandwidth to near-deadline flows. D 2 TCP delivers high application throughput by meeting more deadlines. ...
doi:10.1109/icnp.2015.39
dblp:conf/icnp/HuangHWH15
fatcat:6bzwgtm2k5bodhqrqsxdlrevhi
CASPR: Judiciously Using the Cloud for Wide-Area Packet Recovery
[article]
2019
arXiv
pre-print
When receivers detect packet loss, they recover packets with the help of the nearby data center, not the sender, thus providing quick and reliable packet recovery for latency-sensitive applications. ...
We propose CASPR, a system that judiciously leverages the cloud to recover from lost or delayed packets. ...
Inter-Data Center Networking. Our work complements the large body of work on inter-data center networking. ...
arXiv:1812.10835v2
fatcat:vw3am5aidff35dfpbdww45qrvm
Quality of Service guarantees and fault-tolerant TCP services in mobile wireless optical networks
2008
International Journal of Ad Hoc and Ubiquitous Computing
We also present a mechanism for deploying QoS-enabled dependable TCP services in this network. ...
We propose an architecture for end-to-end statistical delay guarantees. ...
adverse atmospheric effects are minimal. ...
doi:10.1504/ijahuc.2008.018401
fatcat:2jsfhfbycrhwtbieew5hzwh4z4
« Previous
Showing results 1 — 15 out of 849 results