SpaceWire Hot Modules International SpaceWire Conference 2007

Slides:



Advertisements
Similar presentations
A Switch-Based Approach to Starvation in Data Centers Alex Shpiner and Isaac Keslassy Department of Electrical Engineering, Technion. Gabi Bracha, Eyal.
Advertisements

1 Maintaining Packet Order in Two-Stage Switches Isaac Keslassy, Nick McKeown Stanford University.
Best of Both Worlds: A Bus-Enhanced Network on-Chip (BENoC) Ran Manevich, Isask har (Zigi) Walter, Israel Cidon, and Avinoam Kolodny Technion – Israel.
The 9th Israel Networking Day 2014 Scaling Multi-Core Network Processors Without the Reordering Bottleneck Alex Shpiner (Technion/Mellanox) Isaac Keslassy.
* Mellanox Technologies LTD, + Technion - EE Department
Handling Global Traffic in Future CMP NoCs Ran Manevich, Israel Cidon, and Avinoam Kolodny. Group Research QNoC Electrical Engineering Department Technion.
1 Multi-layer Active Queue Management and Congestion Control for Scalable Video Streaming Kang, S.-R.; Zhang, Y.; Dai, M.; Loguinov, D.; Distributed Computing.
Module R R RRR R RRRRR RR R R R R Efficient Link Capacity and QoS Design for Wormhole Network-on-Chip Zvika Guz, Isask ’ har Walter, Evgeny Bolotin, Israel.
Frame-Aggregated Concurrent Matching Switch Bill Lin (University of California, San Diego) Isaac Keslassy (Technion, Israel)
Network based System on Chip Final Presentation Part B Performed by: Medvedev Alexey Supervisor: Walter Isaschar (Zigmond) Winter-Spring 2006.
Mitigating Bandwidth- Exhaustion Attacks using Congestion Puzzles XiaoFeng Wang Michael K. Reiter.
Network based System on Chip Part A Performed by: Medvedev Alexey Supervisor: Walter Isaschar (Zigmond) Winter-Spring 2006.
GeNoLator – Generic Network Simulator Final Presentation Students: Gal Ben-Haim, Dan Blechner Supervisor: Isask'har Walter Winter 08/09 18/08/2009.
1 Link Division Multiplexing (LDM) for NoC Links IEEE 2006 LDM Link Division Multiplexing Arkadiy Morgenshtein, Avinoam Kolodny, Ran Ginosar Technion –
A Switch-Based Approach to Starvation in Data Centers Alex Shpiner Joint work with Isaac Keslassy Faculty of Electrical Engineering Faculty of Electrical.
1 E. Bolotin – The Power of Priority, NoCs 2007 The Power of Priority : NoC based Distributed Cache Coherency Evgeny Bolotin, Zvika Guz, Israel Cidon,
A novel switching paradigm for buffer-less WDM networks Myungsik Yoo and Chunming Qiao EE and CSE Departments University at Buffalo (SUNY)
1 Evgeny Bolotin – ICECS 2004 Automatic Hardware-Efficient SoC Integration by QoS Network on Chip Electrical Engineering Department, Technion, Haifa, Israel.
Lecture 5: Congestion Control l Challenge: how do we efficiently share network resources among billions of hosts? n Last time: TCP n This time: Alternative.
School of Information Technologies IP Quality of Service NETS3303/3603 Weeks
SpaceWire Hot Modules Asaf Baron, Isask'har Walter, Israel Cidon, Ran Ginosar, Isaac Keslassy EE Department, Technion, Haifa, Israel Ofer Lapid Israel.
Benchmarking SpaceWire Networks Asaf Baron, Isask'har Walter, Ran Ginosar, Isaac Keslassy EE Department, Technion, Haifa, Israel Ofer Lapid Israel Ministry.
1 Robust Transport Protocol for Dynamic High-Speed Networks: enhancing XCP approach Dino M. Lopez Pacheco INRIA RESO/LIP, ENS of Lyon, France Congduc Pham.
PC1 LAN GW SP RTR1 SP RTR2 DST 4 * 25 ms 21 ms dst [ ] 4. A third packet is sent with TTL=3, which decrements at each hop, and expires after RTR2,
Multipath Routing in Wireless Mesh Networks Mobile Adhoc and Sensor Systems (MASS), 2006 IEEE International Conference on Author: Nagesh S. Nandiraju,
50 th Annual Allerton Conference, 2012 On the Capacity of Bufferless Networks-on-Chip Alex Shpiner, Erez Kantor, Pu Li, Israel Cidon and Isaac Keslassy.
Selection and Navigation of Mobile sensor Nodes Using a Sensor Network Atul Verma, Hemjit Sawant and Jindong Tan Department of Electrical and Computer.
IEEE HPSR 2014 Scaling Multi-Core Network Processors Without the Reordering Bottleneck Alex Shpiner (Technion / Mellanox) Isaac Keslassy (Technion) Rami.
Overload Prediction Based on Delay in Wireless OFDMA Systems E. O. Lucena, F. R. M. Lima, W. C. Freitas Jr and F. R. P. Cavalcanti Federal University of.
Deadline-based Resource Management for Information- Centric Networks Somaya Arianfar, Pasi Sarolahti, Jörg Ott Aalto University, Department of Communications.
Dynamic Traffic Distribution among Hierarchy Levels in Hierarchical Networks-on-Chip Ran Manevich, Israel Cidon, and Avinoam Kolodny Group Research QNoC.
Module R R RRR R RRRRR RR R R R R Access Regulation to Hot-Modules in Wormhole NoCs Isask’har (Zigi) Walter Supervised by: Israel Cidon, Ran Ginosar and.
Péter Hága Eötvös Loránd University, Hungary European Conference on Complex Systems 2008 Jerusalem, Israel.
Intel Slide 1 A Comparative Study of Arbitration Algorithms for the Alpha Pipelined Router Shubu Mukherjee*, Federico Silla !, Peter Bannon $, Joel.
Prolonging the Lifetime of Wireless Sensor Networks via Unequal Clustering Stanislava Soro Wendi B. Heinzelman University of Rochester IPDPS 2005.
Providing QoS in IP Networks
Improving OBS Efficiency Li, Shuo, Meiqian Wang. Eric W. M. Wong, Moshe Zukerman City University of Hong Kong 1.
MAC Protocols for Sensor Networks
How to Train your Dragonfly
Simulation and Exploration of
QoS & Queuing Theory CS352.
Network Address Translation (NAT)
Analysis of Delay and Jitter of Voice Traffic Over the Internet
Scalable Load-Distance Balancing
CS 268: Lecture 6 Scott Shenker and Ion Stoica
Capacity Allocation Paradox Isaac Keslassy
Traffic Engineering with AIMD in MPLS Networks
Benchmarking SpaceWire Networks
Network Address Translation (NAT)
Tapping Into The Unutilized Router Processing Power
OpenSMART: Single-cycle Multi-hop NoC Generator in BSV and Chisel
Load Balancing Memcached Traffic Using SDN
EE382C Lecture 15 Quality of Service 5/19/11
Israel Cidon, Ran Ginosar and Avinoam Kolodny
CIS679: Anycast Review of Last lecture Anycast.
Coexistence Mechanism
Multi-hop Coflow Routing and Scheduling in Data Centers
OPNET Simulation of Different Queuing Mechanisms
Kevin Lee & Adam Piechowicz 10/10/2009
Course Modules: Demo in Class or
The University of Adelaide, School of Computer Science
Computer Science Division
Memento: Making Sliding Windows Efficient for Heavy Hitters
COMP/ELEC 429/556 Fall 2017 Homework #2
Capacity Allocation Paradox Isaac Keslassy
Congestion Control Reasons:
A Fair Scheduling Scheme for HCF
Call Duration Control with Dynamic Pricing for Congestion Management in Cellular Networks Problem: Heavy Congestion in Cellular Networks at peak hours/hot.
Network Address Translation (NAT)
EECS 122: Introduction to Computer Networks Packet Scheduling and QoS
Presentation transcript:

SpaceWire Hot Modules International SpaceWire Conference 2007 17-19 September 2007 SpaceWire Hot Modules Asaf Baron, Isask'har Walter, Israel Cidon, Ran Ginosar, Isaac Keslassy EE Department, Technion, Haifa, Israel Ofer Lapid Israel Ministry of Defense

An end unit may receive heavy traffic from several sources Hot Module An end unit may receive heavy traffic from several sources The heavy load congests the network Blocks other (light) traffic Enhancing the network will not help Problem is with the end unit, not the network Such unit is a HOT MODULE The heavy load congests the network as long worms spread across it

Example network D1 D2 D3 D4 D5 DHS S5 S4 S3 S2 S1 Storage 29 25 26 27 28 22 23 21 24 18 19 17 20 14 15 13 16 10 11 9 12 6 7 5 8 2 3 1 4 Should say that parts and traffic requirements were taken from the paper “Benchmarking SpaceWire Networks” S5 S4 S3 S2 S1 Storage

Two main problems System Performance Source Fairness Router 5 Router 4 DHS Router 5 Router 4 Router 3 Router 2 Router 1 Reaction Wheel Storage S5 S4 S3 S2 S1

Solution #1 Use Packet Level Priority (PLP) Assign low priority to packets that are traveling to the Hot Module Improves System Performance But does not solve Fairness DHS Router 5 Router 4 Router 3 Router 2 Router 1 Reaction Wheel Need to mention that there is a separate Round Robin arbitration for each priority Storage S5 S4 S3 S2 S1

Solution #2 Problem: traffic sent to Storage (Hot Module), congesting the network Introduce Credits: Before sending to Hot Module, ask for credit No need to modify routers

Credits Improve Performance & Fairness DHS Router 5 Reaction Wheel Router 4 Router 3 Router 2 Router 1 Storage S5 S4 S3 S2 S1 Improves System Performance Solves Fairness

Simulations Use our new SpW simulator and benchmark No more than 2 Sensors have credits at a time

Traffic to Hot Module PLP CREDITS PLP + CREDITS Sensor 5 Sensor 1 Equally getting heart Average ETE delay to Hot Module remains more or less the same

Traffic to DHS (non-Hot-Module) PLP CREDITS PLP + CREDITS A curve here represents traffic that needs to traverse over X hops. For example …

Conclusions: Hot Modules incur two effects Source Fairness Solved using Credits Performance Improved using PLP and Credits Using PLP + Credits yields the best results