1 / 88

Course Matters

Course Matters. Recent Papers. from journals. Good Posters Bad Posters. Improving I/O Performance of Intermediate Multimedia Storage Node. Pal Halvorsen Thomas Plagemann Vera Goebel Multimedia Systems 03. Problem. Improve the performance of I/O in integrated multimedia storage node

ddelgado
Download Presentation

Course Matters

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Course Matters

  2. Recent Papers from journals

  3. Good PostersBad Posters

  4. Improving I/O Performance of Intermediate Multimedia Storage Node Pal Halvorsen Thomas Plagemann Vera Goebel Multimedia Systems 03

  5. Problem • Improve the performance of I/O in integrated multimedia storage node • 3 areas of improvement are identified • reduce memory copy • checksum computation • FEC computation

  6. Reduce Memory Copy • File system maintain pointer to an area in memory • Communication system maintain pointer to same area in memory • Memory copy avoided!

  7. Network Level Framing • Packet payloads are stored with checksum • When packets are retrieved for sending, destination address in header is updated • Checksum is updated with new destination • No need to recompute checksum for payload

  8. Integrated Error Management • Data are stored on RAID-4 (single parity checks) • Use the same error correcting code for RAID and packets • Avoid multiple computation of error correcting code

  9. Experiment: Memory Copy • Read 28662512B file 38 times in a loop • The time to transmit data through the storage node is reduced by 45-50% when there is no CPU load, and by 70-73% when CPU load is high

  10. Experiment: Network Level Framing • Transmit 255MB file • Time to calculate checksum is reduced by 95-99% • Time spent in kernel is reduced by 51-61%

  11. Experiment: Integrated Error Management • With encoding FEC, the maximum throughput is 22-24 Mbps. • Without encoding FEC, the maximum throughput is 1Gbps

  12. Let’s try again..

  13. Improving I/O Performance of Intermediate Media Storage Node Pal Halvorsen Thomas Plagemann Vera Goebel

  14. Contributions • Improve performance by • reducing memory copy • reducing checksum computation • reducing ECC computation

  15. Reduce Memory Copy • One shared copy of data for different OS component Memory Network System File System data

  16. Network Level Framing store payload with payload’s checksum • Reduce time to packetize data and compute checksum for data 1 3 update header and checksum 2 read payload with checksum

  17. Integrated Error Management • Avoid multiple computation of error correcting code reuse RAID ECC as ECC packet RAID ECC data data data ECC

  18. Results: Zero Copy

  19. Results: Network Level Framing Accumulated UDP Protocol Execution Time for sending 225MB file

  20. Results: Integrated Error Management Maximum Throughput with/without Encoding using Cauchy-based Reed Solomon Erasure Code

  21. Ad Hoc Networks Session 1

  22. Mobile Ad Hoc Network Radio Router Host

  23. Animation • http://www-i4.informatik.rwth-aachen.de/~mesut/manet/manet_en.html

  24. Radio Radio Radio Radio Router Router Router Router Host Host Host Host Mobile Ad Hoc Network

  25. Examples • Battlefield • Highway • Disaster Zone

  26. Challenges • All the difficulties of wireless LAN • Plus • Nodes can move • Connections can go up/down • No fix route

  27. Two Papers • IEEE JSAC • Special Issues in Wireless Multimedia • Baochun Li from U. of Toronto • Shiwen Mao from Polytechnic U.

  28. NonStop: Continuous Streaming Service on MANET Baochun Li IEEE JSAC 2004

  29. Streaming over MANET

  30. Network Partition Problem

  31. Only Solution .. • Predict Partition • Replicate Service

  32. Partition Prediction

  33. Network Partition Problem

  34. How to Predict Partition? • given velocity of each node • cluster nodes into “mobile groups” • find mean group velocity

  35. Clustering Algorithm

  36. Clustering Algorithm

  37. Clustering Algorithm

  38. Clustering Algorithm

  39. Choosing Server ?

  40. How to Choose Server? • find “stable group” • choose server within stable group with the most similar velocity

  41. Stable Group probability A B distance mean <= radio range and variance is not too large

  42. Stable Group B A C H D G E F

  43. Stable Group BCD are in my group AGH are in my group B A C H D G E F

  44. Stable Group BCDGH are in my group ABDGH are in my group B A C H D G E F

  45. How to Choose Server? • find “stable group” • choose server within stable group with the most similar velocity

  46. Summary • Server construct mobile group by clustering nodes using velocity • Use mean mobile group velocity to predict network partition • Replicate service before partition to ensure continuous service

  47. Summary • Node construct stable group by comparing distance over time • Choose server within stable group with most similar velocity

  48. Multipath Transport and Multistream Coding for MANET Shiwen Mao et. al. IEEE JSAC

  49. Single Path Transport

  50. Multipath Transport

More Related