1 / 25

A Provider-side View of Web Search Response Time

A Provider-side View of Web Search Response Time. Yingying Chen, Ratul Mahajan, Baskar Sridharan , Zhi -Li Zhang (Univ. of Minnesota) Microsoft. Web services are the dominant way to find and access information. Web service latency is critical to service providers as well. revenue -20%.

chul
Download Presentation

A Provider-side View of Web Search Response Time

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. A Provider-side View ofWeb Search Response Time Yingying Chen, Ratul Mahajan, BaskarSridharan, Zhi-Li Zhang (Univ. of Minnesota) Microsoft

  2. Web services are the dominant way to find and access information

  3. Web service latency is critical to service providers as well revenue -20% Latency +2 sec Latency +0.5 sec revenue -4.3% • Google • Bing

  4. Understanding SRT behavior is challenging 200+t 300+t SRT (ms) t t peak M W Th off-peak S F Su T SRT (ms)

  5. Our work • Explaining systemic SRT variation • Identify SRT anomalies • Root cause localization

  6. Client- and server-side instrumentation query on-load HTML header Brand header Query results BoP scripts Embedded images Referenced content

  7. Impact Factors of SRT network browser query server

  8. Primary factors of SRT variation • Apply Analysis of Variance (ANOVA) on the time intervals ƞ Unexplained variance SRT variance Variance explained by time interval k

  9. 60 40 20 0 Explained variance (%) • Primary factors: network characteristics, browser speed, query type • Server-side processing time has a relatively small impact network browser server query

  10. Variation in network characteristics RTT

  11. Explaining network variations • Residential networks send a higher fraction of queries during off-peak hours than peak hours • Residential networks are slower

  12. residential enterprise unknown 1.25t 25% t RTT (ms) enterprise residential Residential networks are slower Residential networks send a higher fraction of queries during off-peak hours than peak hours

  13. Variation in query type • Impact of query on SRT • Server processing time • Richness of response page • Measure: number of image

  14. Explaining query type variation Off-peak hours Peak hours

  15. Javascript exec time Browser variations 1.82t 82% • Two most popular browsers: X(35%), Y(40%) • Browser-Y sends a higher fraction of queries during off-peak hours • Browser-Y has better performance t Browser-Y Browser-X

  16. Summarizing systemic SRT variation • Server: Little impact • Network: Poorer during off-peak hours • Query: Richer during off-peak hours • Browser: Faster during off-peak hours

  17. Detecting anomalous SRT variations • Challenge: interference from systemic variations

  18. Week-over-Week (WoW) approach + Seasonality + Noise

  19. Comparison with approaches that do not account for systemic variations

  20. Conclusions • Understanding SRT is challenging • Changes in user demographics lead to systemic variations in SRT • Debugging SRT is challenging • Must factor out systemic variations

  21. Implications • Performance monitoring • Should understand performance-equivalent classes • Performance management • Should consider the impact of network, browser, and query • Performance debugging • End-to-end measures are tainted by user behavior changes

  22. Questions?

More Related