Mesa: A geo-replicated online data warehouse for google's advertising system

Ashish Gupta, Fan Yang, Jason Govig, Adam Kirsch, Kelvin Chan, Kevin Lai, Shuo Wu, Sandeep Dhoot, Abhilash Rajesh Kumar, Ankur Agiwal, Sanjay Bhansali, Mingsheng Hong, Jamie Cameron, Masood Siddiqi, David Jones, Jeff Shute, Andrey Gubarev, Shivakumar Venkataraman, Divyakant Agrawal

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

Mesa is a highly scalable analytic data warehousing system that stores critical measurement data related to Google's Internet advertising business. Mesa is designed to satisfy a complex and challenging set of user and systems requirements, including near real-Time data ingestion and retrieval, as well as high availability, reliability, fault tolerance, and scalability for large data and query volumes. Specifically, Mesa handles petabytes of data, processes millions of row updates per second, and serves billions of queries that fetch trillions of rows per day. Mesa is geo-replicated across multiple datacenters and provides consistent and repeatable query answers at low latency, even when an entire datacenter fails. This paper presents the Mesa system and reports the performance and scale that it achieves.

Original languageEnglish
Pages (from-to)117-125
Number of pages9
JournalCommunications of the ACM
Volume59
Issue number7
DOIs
Publication statusPublished - 1 Jul 2016
Externally publishedYes

Fingerprint

Data warehouses
Fault tolerance
Scalability
Marketing
Availability
Internet
Industry

ASJC Scopus subject areas

  • Computer Science(all)

Cite this

Gupta, A., Yang, F., Govig, J., Kirsch, A., Chan, K., Lai, K., ... Agrawal, D. (2016). Mesa: A geo-replicated online data warehouse for google's advertising system. Communications of the ACM, 59(7), 117-125. https://doi.org/10.1145/2936722

Mesa : A geo-replicated online data warehouse for google's advertising system. / Gupta, Ashish; Yang, Fan; Govig, Jason; Kirsch, Adam; Chan, Kelvin; Lai, Kevin; Wu, Shuo; Dhoot, Sandeep; Kumar, Abhilash Rajesh; Agiwal, Ankur; Bhansali, Sanjay; Hong, Mingsheng; Cameron, Jamie; Siddiqi, Masood; Jones, David; Shute, Jeff; Gubarev, Andrey; Venkataraman, Shivakumar; Agrawal, Divyakant.

In: Communications of the ACM, Vol. 59, No. 7, 01.07.2016, p. 117-125.

Research output: Contribution to journalArticle

Gupta, A, Yang, F, Govig, J, Kirsch, A, Chan, K, Lai, K, Wu, S, Dhoot, S, Kumar, AR, Agiwal, A, Bhansali, S, Hong, M, Cameron, J, Siddiqi, M, Jones, D, Shute, J, Gubarev, A, Venkataraman, S & Agrawal, D 2016, 'Mesa: A geo-replicated online data warehouse for google's advertising system', Communications of the ACM, vol. 59, no. 7, pp. 117-125. https://doi.org/10.1145/2936722
Gupta, Ashish ; Yang, Fan ; Govig, Jason ; Kirsch, Adam ; Chan, Kelvin ; Lai, Kevin ; Wu, Shuo ; Dhoot, Sandeep ; Kumar, Abhilash Rajesh ; Agiwal, Ankur ; Bhansali, Sanjay ; Hong, Mingsheng ; Cameron, Jamie ; Siddiqi, Masood ; Jones, David ; Shute, Jeff ; Gubarev, Andrey ; Venkataraman, Shivakumar ; Agrawal, Divyakant. / Mesa : A geo-replicated online data warehouse for google's advertising system. In: Communications of the ACM. 2016 ; Vol. 59, No. 7. pp. 117-125.
@article{82962710e35e4a90bd03043aca620a86,
title = "Mesa: A geo-replicated online data warehouse for google's advertising system",
abstract = "Mesa is a highly scalable analytic data warehousing system that stores critical measurement data related to Google's Internet advertising business. Mesa is designed to satisfy a complex and challenging set of user and systems requirements, including near real-Time data ingestion and retrieval, as well as high availability, reliability, fault tolerance, and scalability for large data and query volumes. Specifically, Mesa handles petabytes of data, processes millions of row updates per second, and serves billions of queries that fetch trillions of rows per day. Mesa is geo-replicated across multiple datacenters and provides consistent and repeatable query answers at low latency, even when an entire datacenter fails. This paper presents the Mesa system and reports the performance and scale that it achieves.",
author = "Ashish Gupta and Fan Yang and Jason Govig and Adam Kirsch and Kelvin Chan and Kevin Lai and Shuo Wu and Sandeep Dhoot and Kumar, {Abhilash Rajesh} and Ankur Agiwal and Sanjay Bhansali and Mingsheng Hong and Jamie Cameron and Masood Siddiqi and David Jones and Jeff Shute and Andrey Gubarev and Shivakumar Venkataraman and Divyakant Agrawal",
year = "2016",
month = "7",
day = "1",
doi = "10.1145/2936722",
language = "English",
volume = "59",
pages = "117--125",
journal = "Communications of the ACM",
issn = "0001-0782",
publisher = "Association for Computing Machinery (ACM)",
number = "7",

}

TY - JOUR

T1 - Mesa

T2 - A geo-replicated online data warehouse for google's advertising system

AU - Gupta, Ashish

AU - Yang, Fan

AU - Govig, Jason

AU - Kirsch, Adam

AU - Chan, Kelvin

AU - Lai, Kevin

AU - Wu, Shuo

AU - Dhoot, Sandeep

AU - Kumar, Abhilash Rajesh

AU - Agiwal, Ankur

AU - Bhansali, Sanjay

AU - Hong, Mingsheng

AU - Cameron, Jamie

AU - Siddiqi, Masood

AU - Jones, David

AU - Shute, Jeff

AU - Gubarev, Andrey

AU - Venkataraman, Shivakumar

AU - Agrawal, Divyakant

PY - 2016/7/1

Y1 - 2016/7/1

N2 - Mesa is a highly scalable analytic data warehousing system that stores critical measurement data related to Google's Internet advertising business. Mesa is designed to satisfy a complex and challenging set of user and systems requirements, including near real-Time data ingestion and retrieval, as well as high availability, reliability, fault tolerance, and scalability for large data and query volumes. Specifically, Mesa handles petabytes of data, processes millions of row updates per second, and serves billions of queries that fetch trillions of rows per day. Mesa is geo-replicated across multiple datacenters and provides consistent and repeatable query answers at low latency, even when an entire datacenter fails. This paper presents the Mesa system and reports the performance and scale that it achieves.

AB - Mesa is a highly scalable analytic data warehousing system that stores critical measurement data related to Google's Internet advertising business. Mesa is designed to satisfy a complex and challenging set of user and systems requirements, including near real-Time data ingestion and retrieval, as well as high availability, reliability, fault tolerance, and scalability for large data and query volumes. Specifically, Mesa handles petabytes of data, processes millions of row updates per second, and serves billions of queries that fetch trillions of rows per day. Mesa is geo-replicated across multiple datacenters and provides consistent and repeatable query answers at low latency, even when an entire datacenter fails. This paper presents the Mesa system and reports the performance and scale that it achieves.

UR - http://www.scopus.com/inward/record.url?scp=84977156667&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84977156667&partnerID=8YFLogxK

U2 - 10.1145/2936722

DO - 10.1145/2936722

M3 - Article

AN - SCOPUS:84977156667

VL - 59

SP - 117

EP - 125

JO - Communications of the ACM

JF - Communications of the ACM

SN - 0001-0782

IS - 7

ER -