bg: a scalable benchmark for interactive social networking...

1
5. Results 3. Key Features 1. Motivation 6. Related Work BG: A Scalable Benchmark for Interactive Social Networking Actions http://BGBenchmark.org dblab.usc.edu 2. BG Benchmark Conceptual Data Model Software Architecture 4. Execution Paradigms SoAR Social Action Rating ,SoAR, is defined as the highest throughput provided by a data store while satisfying a pre- specified service level agreement, SLA Unpredictable Reads Performance of Memcached augmented SQL-X system configured with a TTL based invalidation mechanism -Mixed workload of 10% write actions. Visualization Deck Client-side cache Architecture Non-Partitioned Implementations Yazeed Alabdulkarim, Sumita Barahmand, Shahram Ghandeharizadeh Dept. Of Computer Science, University Of Southern California, Microsoft 1 1 2 1 2 BGCoord Data Store Output DB Calls DB Calls BG Visualization Deck Input + SLA SoAR/Socialites BGClient Threads BGClient 1 Data Generator Action Generator Data Store Interface Generator BGClient Threads BGClient N Data Generator Data Store Interface Generator Validation Measurement Validation Measurement Action Generator 1- S. Barahmand and S. Ghandeharizadeh, “BG: A Benchmark to Evaluate Interactive Social Networking Actions,” Proceedings of 2013 CIDR, January 2013 2- T. G. Armstrong and et. al., “LinkBench: A Database Benchmark Based on the Facebook Social Graph,” in SIGMOD, 2013, pp. 1185–1196 3- D. E. Difallah and et. al., “OLTP-Bench: An Extensible Testbed for Benchmarking Relational Databases,” PVLDB, vol. 7, no. 4, pp. 277–288, 2013. 4- O. Erling and et. al., “The LDBC Social Network Benchmark: Interactive Workload,” in SIGMOD, 2015. How do we compare one data store over another? Which components of a data store result in it becoming the bottleneck? Which systems perform best for what workloads? How accurate is the information being offered by the vendors of data stores? R. Cattell, “Scalable SQL and NoSQL Data Stores,” SIGMOD Rec., vol. 39, pp. 12–27, May 2011. S. Barahmand and S. Ghandeharizadeh, “BG: A Benchmark to Evaluate Interactive Social Networking Actions,” Proceedings of 2013 CIDR, January 2013 Y. Alabdulkarim, S. Barahmand, S. Ghandeharizadeh. BG: A Scalable Benchmark for Interactive Social Networking Actions. Submitted for publication Partitioned Non-Partitioned

Upload: others

Post on 28-Sep-2020

0 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: BG: A Scalable Benchmark for Interactive Social Networking ...acmsocc.github.io/2015/posters/socc15posters-final2.pdf- O. Erling and et. al., ^The LD Social Network enchmark: Interactive

5. Results

3. Key Features

1. Motivation

6. Related Work

BG: A Scalable Benchmark for Interactive Social Networking Actions

http://BGBenchmark.org dblab.usc.edu

2. BG Benchmark

Conceptual Data Model Software Architecture

4. Execution Paradigms

SoAR Social Action Rating ,SoAR, is defined as the highest throughput provided by a data store while satisfying a pre-specified service level agreement, SLA

Unpredictable Reads Performance of Memcached augmented SQL-X system configured with a TTL based invalidation mechanism -Mixed workload of 10% write actions.

Visualization Deck

Client-side cache Architecture

Non-Partitioned Implementations

Yazeed Alabdulkarim, Sumita Barahmand, Shahram Ghandeharizadeh Dept. Of Computer Science, University Of Southern California, Microsoft

1 1 2

1 2

BGCoord

Data Store

Output

DB Calls DB Calls

BG Visualization Deck

Input + SLA

SoAR/Socialites

BGClient

Threads

BGClient 1

Data

Generator

Action

Generator

Data Store Interface

Generator

BGClient

Threads

BGClient N

Data

Generator

Data Store Interface

Generator

… …

Validation Measurement Validation Measurement

Action

Generator

1- S. Barahmand and S. Ghandeharizadeh, “BG: A Benchmark to Evaluate Interactive Social Networking Actions,” Proceedings of 2013 CIDR, January 2013 2- T. G. Armstrong and et. al., “LinkBench: A Database Benchmark Based on the Facebook Social Graph,” in SIGMOD, 2013, pp. 1185–1196 3- D. E. Difallah and et. al., “OLTP-Bench: An Extensible Testbed for Benchmarking Relational Databases,” PVLDB, vol. 7, no. 4, pp. 277–288, 2013. 4- O. Erling and et. al., “The LDBC Social Network Benchmark: Interactive Workload,” in SIGMOD, 2015.

• How do we compare one data store over another? • Which components of a data store result in it becoming the bottleneck? • Which systems perform best for what workloads? • How accurate is the information being offered by the vendors of data stores?

R. Cattell, “Scalable SQL and NoSQL Data Stores,” SIGMOD Rec., vol. 39, pp. 12–27, May 2011.

S. Barahmand and S. Ghandeharizadeh, “BG: A Benchmark to Evaluate Interactive Social Networking Actions,” Proceedings of 2013 CIDR, January 2013 Y. Alabdulkarim, S. Barahmand, S. Ghandeharizadeh. BG: A Scalable Benchmark for Interactive Social Networking Actions. Submitted for publication

Partitioned Non-Partitioned