data flow simulation of the alice computing system with
play

Data Flow Simulation of the ALICE Computing System with OMNET++ - PowerPoint PPT Presentation

Data Flow Simulation of the ALICE Computing System with OMNET++ Rifki Sadikin, Furqon H.Mutaqien, Iosif Legrand November 25, 2016 1/17 Table of contents Design and Implementation Results Small Size FLP=6*6, EPN=6*36 Big Size


  1. Data Flow Simulation of the ALICE Computing System with OMNET++ Rifki Sadikin, Furqon H.Mutaqien, Iosif Legrand November 25, 2016 1/17

  2. Table of contents Design and Implementation Results Small Size FLP=6*6, EPN=6*36 Big Size FLP=42*6=252,EPN=42*36=1512 2/17

  3. Design and Implementation 3/17

  4. Design 4/17

  5. Network Topology 5/17

  6. Behavior Design 6/17

  7. State Diagram 7/17

  8. Results 8/17

  9. Buffer (with Service Time in FLP Processing Varied) 9/17

  10. Accumulated Dropped Event 10/17

  11. Buffer (with Buffer Size Varied) 11/17

  12. Total Buffer at FLP#0 and #40 with FLP Service Time { 100,200,300,400,500 } ms and Max Buffer Size 8 GB 12/17

  13. Drop Event at FLP#0 and #40 with FLP Service Time { 100,200,300,400,500 } ms and Max Buffer Size 8 GB 13/17

  14. 14/17

  15. Network Latency from FLP to ELP 15/17

  16. Time to Gather STFs to TF in an EPN 16/17

  17. Future Plan 1. Current simulation can produce buffer size and data flow timing for ALICE-DAQ simulation 2. The results depends on service time in FLP and EPN, and maximum buffer size 3. More refine behavior on FLPs and EPNs internal data flow are needed 4. Validation against previous real data 17/17

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend


More recommend