Estimating the performance of hypothetical cloud service deployments: A measurement-based approach

Y. Angela Wang, Cheng Huang, Jin Li, Keith Ross

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Abstract

    To optimize network performance, cloud service providers have a number of options available to them, including co-locating production servers in well-connected Internet eXchange (IX) points, deploying data centers in additional locations, or contracting with external Content Distribution Networks (CDNs). Some of these options can be very costly, and some may or may not improve performance significantly. Cloud service providers would clearly like to be able to estimate a priori performance gain of the various options before sinking significant capital expenditures into major infrastructure changes. In this paper we take a measurement-oriented approach and develop methodologies that accurately predict the performance improvement for making major infrastructure changes. Our methodologies leverage active web content, existing large-scale CDN infrastructures, and the SpeedTest network. We then apply our methodologies and a CloudBeacon tool to the problem of locating satellite data centers throughout the world. The results show that for North America, a deployment limited to 11 locations will be sufficient. However, in order to provide good latency and throughput performance on a global scale, somewhere between a total of 36 and 72 cloud-service locations with good peering connections is most likely needed.

    Original languageEnglish (US)
    Title of host publication2011 Proceedings IEEE INFOCOM
    Pages2372-2380
    Number of pages9
    DOIs
    StatePublished - 2011
    EventIEEE INFOCOM 2011 - Shanghai, China
    Duration: Apr 10 2011Apr 15 2011

    Other

    OtherIEEE INFOCOM 2011
    CountryChina
    CityShanghai
    Period4/10/114/15/11

    Fingerprint

    Electric power distribution
    Network performance
    Servers
    Throughput
    Satellites
    Internet

    ASJC Scopus subject areas

    • Computer Science(all)
    • Electrical and Electronic Engineering

    Cite this

    Wang, Y. A., Huang, C., Li, J., & Ross, K. (2011). Estimating the performance of hypothetical cloud service deployments: A measurement-based approach. In 2011 Proceedings IEEE INFOCOM (pp. 2372-2380). [5935057] https://doi.org/10.1109/INFCOM.2011.5935057

    Estimating the performance of hypothetical cloud service deployments : A measurement-based approach. / Wang, Y. Angela; Huang, Cheng; Li, Jin; Ross, Keith.

    2011 Proceedings IEEE INFOCOM. 2011. p. 2372-2380 5935057.

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    Wang, YA, Huang, C, Li, J & Ross, K 2011, Estimating the performance of hypothetical cloud service deployments: A measurement-based approach. in 2011 Proceedings IEEE INFOCOM., 5935057, pp. 2372-2380, IEEE INFOCOM 2011, Shanghai, China, 4/10/11. https://doi.org/10.1109/INFCOM.2011.5935057
    Wang YA, Huang C, Li J, Ross K. Estimating the performance of hypothetical cloud service deployments: A measurement-based approach. In 2011 Proceedings IEEE INFOCOM. 2011. p. 2372-2380. 5935057 https://doi.org/10.1109/INFCOM.2011.5935057
    Wang, Y. Angela ; Huang, Cheng ; Li, Jin ; Ross, Keith. / Estimating the performance of hypothetical cloud service deployments : A measurement-based approach. 2011 Proceedings IEEE INFOCOM. 2011. pp. 2372-2380
    @inproceedings{7fbffc7bc6a94656a25beffbd123f530,
    title = "Estimating the performance of hypothetical cloud service deployments: A measurement-based approach",
    abstract = "To optimize network performance, cloud service providers have a number of options available to them, including co-locating production servers in well-connected Internet eXchange (IX) points, deploying data centers in additional locations, or contracting with external Content Distribution Networks (CDNs). Some of these options can be very costly, and some may or may not improve performance significantly. Cloud service providers would clearly like to be able to estimate a priori performance gain of the various options before sinking significant capital expenditures into major infrastructure changes. In this paper we take a measurement-oriented approach and develop methodologies that accurately predict the performance improvement for making major infrastructure changes. Our methodologies leverage active web content, existing large-scale CDN infrastructures, and the SpeedTest network. We then apply our methodologies and a CloudBeacon tool to the problem of locating satellite data centers throughout the world. The results show that for North America, a deployment limited to 11 locations will be sufficient. However, in order to provide good latency and throughput performance on a global scale, somewhere between a total of 36 and 72 cloud-service locations with good peering connections is most likely needed.",
    author = "Wang, {Y. Angela} and Cheng Huang and Jin Li and Keith Ross",
    year = "2011",
    doi = "10.1109/INFCOM.2011.5935057",
    language = "English (US)",
    isbn = "9781424499212",
    pages = "2372--2380",
    booktitle = "2011 Proceedings IEEE INFOCOM",

    }

    TY - GEN

    T1 - Estimating the performance of hypothetical cloud service deployments

    T2 - A measurement-based approach

    AU - Wang, Y. Angela

    AU - Huang, Cheng

    AU - Li, Jin

    AU - Ross, Keith

    PY - 2011

    Y1 - 2011

    N2 - To optimize network performance, cloud service providers have a number of options available to them, including co-locating production servers in well-connected Internet eXchange (IX) points, deploying data centers in additional locations, or contracting with external Content Distribution Networks (CDNs). Some of these options can be very costly, and some may or may not improve performance significantly. Cloud service providers would clearly like to be able to estimate a priori performance gain of the various options before sinking significant capital expenditures into major infrastructure changes. In this paper we take a measurement-oriented approach and develop methodologies that accurately predict the performance improvement for making major infrastructure changes. Our methodologies leverage active web content, existing large-scale CDN infrastructures, and the SpeedTest network. We then apply our methodologies and a CloudBeacon tool to the problem of locating satellite data centers throughout the world. The results show that for North America, a deployment limited to 11 locations will be sufficient. However, in order to provide good latency and throughput performance on a global scale, somewhere between a total of 36 and 72 cloud-service locations with good peering connections is most likely needed.

    AB - To optimize network performance, cloud service providers have a number of options available to them, including co-locating production servers in well-connected Internet eXchange (IX) points, deploying data centers in additional locations, or contracting with external Content Distribution Networks (CDNs). Some of these options can be very costly, and some may or may not improve performance significantly. Cloud service providers would clearly like to be able to estimate a priori performance gain of the various options before sinking significant capital expenditures into major infrastructure changes. In this paper we take a measurement-oriented approach and develop methodologies that accurately predict the performance improvement for making major infrastructure changes. Our methodologies leverage active web content, existing large-scale CDN infrastructures, and the SpeedTest network. We then apply our methodologies and a CloudBeacon tool to the problem of locating satellite data centers throughout the world. The results show that for North America, a deployment limited to 11 locations will be sufficient. However, in order to provide good latency and throughput performance on a global scale, somewhere between a total of 36 and 72 cloud-service locations with good peering connections is most likely needed.

    UR - http://www.scopus.com/inward/record.url?scp=79960887189&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=79960887189&partnerID=8YFLogxK

    U2 - 10.1109/INFCOM.2011.5935057

    DO - 10.1109/INFCOM.2011.5935057

    M3 - Conference contribution

    AN - SCOPUS:79960887189

    SN - 9781424499212

    SP - 2372

    EP - 2380

    BT - 2011 Proceedings IEEE INFOCOM

    ER -