SlideShare uma empresa Scribd logo
1 de 20
Baixar para ler offline
Scaling UP
Challenges Encountered Scaling Up
Recommendation Services @Gravity R&D
Bottyán Németh
Who we are and what we do
Gravity R&D is a recommender system vendor company.
We provide recommendation as a service since 2009 for
our customers all around the globe.
2
How we imagine growth?
3
?
How we imagine growth?
4
How it actually happens?
5
?
How it actually happens?
6
# of requests
7
Vatera.hu largest online marketplace in Hungary
served by one “server”
Alexa TOP100 video chat webpage
(~40M recommendation requests / day):
 Served by 5 application servers and 1 DB
 Too many events to store in MySQL  using
Cassandra (v0.6)
 Training time for IALS too long  speedup by IALS1
 Max. 5 sec latency in “product” availability
Using new/beta technologies
8
Cassandra (v0.6)
Nginx (v0.5) (22% of top 1M sites)
Kafka (v0.8)
MySQL auto. failover
Reaching the limits
9
Even if the technology is widely used if you reach it’s
limits the optimization is very costly / time consuming.
Java GC – service collapsed because increased minor GC
times due to a JVM bug (26th of January 2013)
Maintaining MySQL with lots of data (optimize table,
slave replication lag, faster storage device)
Complexity increases
10
There is always a business request or an algorithmic
development which requires more resources.
Optimizations
11
Infrastructure
12
Currently 200+ hosts and 3500+ services monitored
0
50
100
150
200
250
2008 2009 2010 2011 2012 2013 2014 2015 2016
Number of servers
# of items
13
How to store item model / metadata in memory to serve
requests fast?
# of items
14
How to store item model / metadata in memory to serve
requests fast?
VS.
Auto increment IDs for the items?
231 not enough
Preconceptions
15
More data better results.
If the CTR of a new algorithm is low than the old
algorithm is better.
Daily retrain is enough.
Training frequency
16
CTR decreased in the morning
100+ Algorithms
17
0
10
20
30
40
50
60
0 20 40 60 80 100 120
Number of times an algorithm is used
Now
18
• Performance: Gravity’s performance
oriented architecture enables real-time
response to the always changing
environment and user behavior
• Algorithms: more than 100 different
recommendation algorithm enables true
personalization and to reach the highest
KPIs in different domains
• Infrastructure: fast response times all
around the globe and data security thanks
to the private cloud infrastructure located
in 4 different data centers
• Flexibility: the advanced business rule
engine with intuitive user interface allows
to satisfy various business requirements
Performance
140M requests
served daily
Algorithms
30 man-years
invested
Infrastructure
4 data centers
globally
Flexibility
100s of logics
configurable
Cross the river when you come to it
19
Thank you!
20

Mais conteúdo relacionado

Mais procurados

The challenges of live events scalability
The challenges of live events scalabilityThe challenges of live events scalability
The challenges of live events scalability
Guy Tomer
 
Real time machine learning
Real time machine learningReal time machine learning
Real time machine learning
Vinoth Kannan
 

Mais procurados (20)

GluonCV
GluonCVGluonCV
GluonCV
 
The challenges of live events scalability
The challenges of live events scalabilityThe challenges of live events scalability
The challenges of live events scalability
 
ASTQB washington-sept-2015
ASTQB washington-sept-2015ASTQB washington-sept-2015
ASTQB washington-sept-2015
 
Microsoft AI Platform - AETHER Introduction
Microsoft AI Platform - AETHER IntroductionMicrosoft AI Platform - AETHER Introduction
Microsoft AI Platform - AETHER Introduction
 
Sri Rajan - Driving cloud adoption through DevOps / Unlocked: the Hybrid Clou...
Sri Rajan - Driving cloud adoption through DevOps / Unlocked: the Hybrid Clou...Sri Rajan - Driving cloud adoption through DevOps / Unlocked: the Hybrid Clou...
Sri Rajan - Driving cloud adoption through DevOps / Unlocked: the Hybrid Clou...
 
What is changed in products/service licensing with Cloud?
What is changed in products/service licensing with Cloud?What is changed in products/service licensing with Cloud?
What is changed in products/service licensing with Cloud?
 
Industrial Data Science
Industrial Data ScienceIndustrial Data Science
Industrial Data Science
 
ReStream: Accelerating Backtesting and Stream Replay with Serial-Equivalent P...
ReStream: Accelerating Backtesting and Stream Replay with Serial-Equivalent P...ReStream: Accelerating Backtesting and Stream Replay with Serial-Equivalent P...
ReStream: Accelerating Backtesting and Stream Replay with Serial-Equivalent P...
 
Rail Performance in the Cloud - Opening
Rail Performance in the Cloud - OpeningRail Performance in the Cloud - Opening
Rail Performance in the Cloud - Opening
 
SolidWorks Design Automation Using the SolidWorks API, Microsoft Excel and VBA
SolidWorks Design Automation Using the SolidWorks API, Microsoft Excel and VBASolidWorks Design Automation Using the SolidWorks API, Microsoft Excel and VBA
SolidWorks Design Automation Using the SolidWorks API, Microsoft Excel and VBA
 
EIA2017Italy - Danny Lange - Artificial Intelligence - A Game Changer in App ...
EIA2017Italy - Danny Lange - Artificial Intelligence - A Game Changer in App ...EIA2017Italy - Danny Lange - Artificial Intelligence - A Game Changer in App ...
EIA2017Italy - Danny Lange - Artificial Intelligence - A Game Changer in App ...
 
AWS Webcast - Tibco Jaspersoft
AWS Webcast - Tibco JaspersoftAWS Webcast - Tibco Jaspersoft
AWS Webcast - Tibco Jaspersoft
 
12 Ways to Manage Cloud Costs and Optimize Cloud Spend
12 Ways to Manage Cloud Costs and Optimize Cloud Spend12 Ways to Manage Cloud Costs and Optimize Cloud Spend
12 Ways to Manage Cloud Costs and Optimize Cloud Spend
 
Real time machine learning
Real time machine learningReal time machine learning
Real time machine learning
 
Creating a Culture of Cost Management in Your Organization – J.R. Storment, C...
Creating a Culture of Cost Management in Your Organization – J.R. Storment, C...Creating a Culture of Cost Management in Your Organization – J.R. Storment, C...
Creating a Culture of Cost Management in Your Organization – J.R. Storment, C...
 
Big Data in Production: Lessons from Running in the Cloud
Big Data in Production: Lessons from Running in the CloudBig Data in Production: Lessons from Running in the Cloud
Big Data in Production: Lessons from Running in the Cloud
 
SnapLogic Overview: Are You Feeling SMACT?
SnapLogic Overview: Are You Feeling SMACT?SnapLogic Overview: Are You Feeling SMACT?
SnapLogic Overview: Are You Feeling SMACT?
 
How Companies are Using Cloud-Based Data Visualization & Analytics to Transfo...
How Companies are Using Cloud-Based Data Visualization & Analytics to Transfo...How Companies are Using Cloud-Based Data Visualization & Analytics to Transfo...
How Companies are Using Cloud-Based Data Visualization & Analytics to Transfo...
 
AWS Webcast - Journey through the Cloud - Cost Optimization
AWS Webcast - Journey through the Cloud - Cost OptimizationAWS Webcast - Journey through the Cloud - Cost Optimization
AWS Webcast - Journey through the Cloud - Cost Optimization
 
Big Data Day LA 2015 - Building a Big Data Culture in the Entertainment Indus...
Big Data Day LA 2015 - Building a Big Data Culture in the Entertainment Indus...Big Data Day LA 2015 - Building a Big Data Culture in the Entertainment Indus...
Big Data Day LA 2015 - Building a Big Data Culture in the Entertainment Indus...
 

Destaque

Gravity rd corporate introduction - nlp matiné 2014
Gravity rd corporate introduction  - nlp matiné 2014Gravity rd corporate introduction  - nlp matiné 2014
Gravity rd corporate introduction - nlp matiné 2014
Zoltan Varju
 
Gravity personalizaton intro
Gravity personalizaton introGravity personalizaton intro
Gravity personalizaton intro
Eszter Nagy
 
Understanding How CQL3 Maps to Cassandra's Internal Data Structure
Understanding How CQL3 Maps to Cassandra's Internal Data StructureUnderstanding How CQL3 Maps to Cassandra's Internal Data Structure
Understanding How CQL3 Maps to Cassandra's Internal Data Structure
DataStax
 
Dynamically Allocate Cluster Resources to your Spark Application
Dynamically Allocate Cluster Resources to your Spark ApplicationDynamically Allocate Cluster Resources to your Spark Application
Dynamically Allocate Cluster Resources to your Spark Application
DataWorks Summit
 

Destaque (16)

Recommenders on video sharing portals - business and algorithmic aspects
Recommenders on video sharing portals - business and algorithmic aspectsRecommenders on video sharing portals - business and algorithmic aspects
Recommenders on video sharing portals - business and algorithmic aspects
 
Gravity rd corporate introduction - nlp matiné 2014
Gravity rd corporate introduction  - nlp matiné 2014Gravity rd corporate introduction  - nlp matiné 2014
Gravity rd corporate introduction - nlp matiné 2014
 
Xây dựng mạng lưới tài năng trẻ trong sáng tạo – khởi nghiệp
Xây dựng mạng lưới tài năng trẻ trong sáng tạo – khởi nghiệpXây dựng mạng lưới tài năng trẻ trong sáng tạo – khởi nghiệp
Xây dựng mạng lưới tài năng trẻ trong sáng tạo – khởi nghiệp
 
Gravity personalizaton intro
Gravity personalizaton introGravity personalizaton intro
Gravity personalizaton intro
 
Entrepreneurship & Innovation: Dual-core Engine
Entrepreneurship & Innovation: Dual-core EngineEntrepreneurship & Innovation: Dual-core Engine
Entrepreneurship & Innovation: Dual-core Engine
 
The rise of Recommendation Engines
The rise of Recommendation EnginesThe rise of Recommendation Engines
The rise of Recommendation Engines
 
Lessons learnt at building recommendation services at industry scale
Lessons learnt at building recommendation services at industry scaleLessons learnt at building recommendation services at industry scale
Lessons learnt at building recommendation services at industry scale
 
RecSys 2015: Large-scale real-time product recommendation at Criteo
RecSys 2015: Large-scale real-time product recommendation at CriteoRecSys 2015: Large-scale real-time product recommendation at Criteo
RecSys 2015: Large-scale real-time product recommendation at Criteo
 
Understanding How CQL3 Maps to Cassandra's Internal Data Structure
Understanding How CQL3 Maps to Cassandra's Internal Data StructureUnderstanding How CQL3 Maps to Cassandra's Internal Data Structure
Understanding How CQL3 Maps to Cassandra's Internal Data Structure
 
Dynamically Allocate Cluster Resources to your Spark Application
Dynamically Allocate Cluster Resources to your Spark ApplicationDynamically Allocate Cluster Resources to your Spark Application
Dynamically Allocate Cluster Resources to your Spark Application
 
Creating an end-to-end Recommender System with Apache Spark and Elasticsearch...
Creating an end-to-end Recommender System with Apache Spark and Elasticsearch...Creating an end-to-end Recommender System with Apache Spark and Elasticsearch...
Creating an end-to-end Recommender System with Apache Spark and Elasticsearch...
 
Organizational-culture
Organizational-cultureOrganizational-culture
Organizational-culture
 
Centralization and Decentralization
Centralization and DecentralizationCentralization and Decentralization
Centralization and Decentralization
 
Using Docker for GPU Accelerated Applications
Using Docker for GPU Accelerated ApplicationsUsing Docker for GPU Accelerated Applications
Using Docker for GPU Accelerated Applications
 
LDA Beginner's Tutorial
LDA Beginner's TutorialLDA Beginner's Tutorial
LDA Beginner's Tutorial
 
10 Lessons Learned from Building Machine Learning Systems
10 Lessons Learned from Building Machine Learning Systems10 Lessons Learned from Building Machine Learning Systems
10 Lessons Learned from Building Machine Learning Systems
 

Semelhante a Challenges Encountered by Scaling Up Recommendation Services at Gravity R&D

Webinar Slides: High Volume MySQL HA: SaaS Continuous Operations with Terabyt...
Webinar Slides: High Volume MySQL HA: SaaS Continuous Operations with Terabyt...Webinar Slides: High Volume MySQL HA: SaaS Continuous Operations with Terabyt...
Webinar Slides: High Volume MySQL HA: SaaS Continuous Operations with Terabyt...
Continuent
 
Five Early Challenges Of Building Streaming Fast Data Applications
Five Early Challenges Of Building Streaming Fast Data ApplicationsFive Early Challenges Of Building Streaming Fast Data Applications
Five Early Challenges Of Building Streaming Fast Data Applications
Lightbend
 

Semelhante a Challenges Encountered by Scaling Up Recommendation Services at Gravity R&D (20)

There are 250 Database products, are you running the right one?
There are 250 Database products, are you running the right one?There are 250 Database products, are you running the right one?
There are 250 Database products, are you running the right one?
 
How to run Real Time processing on Big Data / Ron Zavner (GigaSpaces)
How to run Real Time processing on Big Data / Ron Zavner (GigaSpaces)How to run Real Time processing on Big Data / Ron Zavner (GigaSpaces)
How to run Real Time processing on Big Data / Ron Zavner (GigaSpaces)
 
The Cloud - What's different
The Cloud - What's differentThe Cloud - What's different
The Cloud - What's different
 
Serverless Computing: Driving Innovation and Business Value
Serverless Computing: Driving Innovation and Business ValueServerless Computing: Driving Innovation and Business Value
Serverless Computing: Driving Innovation and Business Value
 
Intro to AWS: Amazon EC2 and Compute Services
Intro to AWS: Amazon EC2 and Compute ServicesIntro to AWS: Amazon EC2 and Compute Services
Intro to AWS: Amazon EC2 and Compute Services
 
Webinar Slides: High Volume MySQL HA: SaaS Continuous Operations with Terabyt...
Webinar Slides: High Volume MySQL HA: SaaS Continuous Operations with Terabyt...Webinar Slides: High Volume MySQL HA: SaaS Continuous Operations with Terabyt...
Webinar Slides: High Volume MySQL HA: SaaS Continuous Operations with Terabyt...
 
Intro to AWS: Amazon EC2 and Compute Services
Intro to AWS: Amazon EC2 and Compute ServicesIntro to AWS: Amazon EC2 and Compute Services
Intro to AWS: Amazon EC2 and Compute Services
 
Scale Your Load Balancer from 0 to 1 million TPS on Azure
Scale Your Load Balancer from 0 to 1 million TPS on AzureScale Your Load Balancer from 0 to 1 million TPS on Azure
Scale Your Load Balancer from 0 to 1 million TPS on Azure
 
AWS Summit Kuala Lumpur Keynote with Stephen Orban - Head of Enterprise Strategy
AWS Summit Kuala Lumpur Keynote with Stephen Orban - Head of Enterprise StrategyAWS Summit Kuala Lumpur Keynote with Stephen Orban - Head of Enterprise Strategy
AWS Summit Kuala Lumpur Keynote with Stephen Orban - Head of Enterprise Strategy
 
Proact SYNC 2013 Breakout session - NetApp Clustered DataONTAP, dé storage hy...
Proact SYNC 2013 Breakout session - NetApp Clustered DataONTAP, dé storage hy...Proact SYNC 2013 Breakout session - NetApp Clustered DataONTAP, dé storage hy...
Proact SYNC 2013 Breakout session - NetApp Clustered DataONTAP, dé storage hy...
 
Intro to AWS: EC2 & Compute Services
Intro to AWS: EC2 & Compute ServicesIntro to AWS: EC2 & Compute Services
Intro to AWS: EC2 & Compute Services
 
Data Culture Series - Keynote - 3rd Dec
Data Culture Series - Keynote - 3rd DecData Culture Series - Keynote - 3rd Dec
Data Culture Series - Keynote - 3rd Dec
 
Leveraging Big Data with Hadoop, NoSQL and RDBMS
Leveraging Big Data with Hadoop, NoSQL and RDBMSLeveraging Big Data with Hadoop, NoSQL and RDBMS
Leveraging Big Data with Hadoop, NoSQL and RDBMS
 
Using real time big data analytics for competitive advantage
 Using real time big data analytics for competitive advantage Using real time big data analytics for competitive advantage
Using real time big data analytics for competitive advantage
 
Five Early Challenges Of Building Streaming Fast Data Applications
Five Early Challenges Of Building Streaming Fast Data ApplicationsFive Early Challenges Of Building Streaming Fast Data Applications
Five Early Challenges Of Building Streaming Fast Data Applications
 
Data Virtualization Journey: How to Grow from Single Project and to Enterpris...
Data Virtualization Journey: How to Grow from Single Project and to Enterpris...Data Virtualization Journey: How to Grow from Single Project and to Enterpris...
Data Virtualization Journey: How to Grow from Single Project and to Enterpris...
 
Introducing Amazon Kinesis: Real-time Processing of Streaming Big Data (BDT10...
Introducing Amazon Kinesis: Real-time Processing of Streaming Big Data (BDT10...Introducing Amazon Kinesis: Real-time Processing of Streaming Big Data (BDT10...
Introducing Amazon Kinesis: Real-time Processing of Streaming Big Data (BDT10...
 
Neha Narkhede | Kafka Summit London 2019 Keynote | Event Streaming: Our Cloud...
Neha Narkhede | Kafka Summit London 2019 Keynote | Event Streaming: Our Cloud...Neha Narkhede | Kafka Summit London 2019 Keynote | Event Streaming: Our Cloud...
Neha Narkhede | Kafka Summit London 2019 Keynote | Event Streaming: Our Cloud...
 
Vertica Analytics Database general overview
Vertica Analytics Database general overviewVertica Analytics Database general overview
Vertica Analytics Database general overview
 
Migrating from Oracle to Postgres
Migrating from Oracle to PostgresMigrating from Oracle to Postgres
Migrating from Oracle to Postgres
 

Mais de Domonkos Tikk

General factorization framework for context-aware recommendations
General factorization framework for context-aware recommendationsGeneral factorization framework for context-aware recommendations
General factorization framework for context-aware recommendations
Domonkos Tikk
 
Tartalomgazdagítás (content enrichment)
Tartalomgazdagítás (content enrichment) Tartalomgazdagítás (content enrichment)
Tartalomgazdagítás (content enrichment)
Domonkos Tikk
 
Context-aware similarities within the factorization framework - presented at ...
Context-aware similarities within the factorization framework - presented at ...Context-aware similarities within the factorization framework - presented at ...
Context-aware similarities within the factorization framework - presented at ...
Domonkos Tikk
 
Slides from CARR 2012 WS - Enhancing Matrix Factorization Through Initializat...
Slides from CARR 2012 WS - Enhancing Matrix Factorization Through Initializat...Slides from CARR 2012 WS - Enhancing Matrix Factorization Through Initializat...
Slides from CARR 2012 WS - Enhancing Matrix Factorization Through Initializat...
Domonkos Tikk
 
Fast ALS-Based Tensor Factorization for Context-Aware Recommendation from Imp...
Fast ALS-Based Tensor Factorization for Context-Aware Recommendation from Imp...Fast ALS-Based Tensor Factorization for Context-Aware Recommendation from Imp...
Fast ALS-Based Tensor Factorization for Context-Aware Recommendation from Imp...
Domonkos Tikk
 

Mais de Domonkos Tikk (9)

Neighbor methods vs matrix factorization - case studies of real-life recommen...
Neighbor methods vs matrix factorization - case studies of real-life recommen...Neighbor methods vs matrix factorization - case studies of real-life recommen...
Neighbor methods vs matrix factorization - case studies of real-life recommen...
 
General factorization framework for context-aware recommendations
General factorization framework for context-aware recommendationsGeneral factorization framework for context-aware recommendations
General factorization framework for context-aware recommendations
 
Tartalomgazdagítás (content enrichment)
Tartalomgazdagítás (content enrichment) Tartalomgazdagítás (content enrichment)
Tartalomgazdagítás (content enrichment)
 
Idomaar crowd rec_reference_fw
Idomaar crowd rec_reference_fwIdomaar crowd rec_reference_fw
Idomaar crowd rec_reference_fw
 
Big Data in Online Classifieds
Big Data in Online ClassifiedsBig Data in Online Classifieds
Big Data in Online Classifieds
 
Context-aware similarities within the factorization framework - presented at ...
Context-aware similarities within the factorization framework - presented at ...Context-aware similarities within the factorization framework - presented at ...
Context-aware similarities within the factorization framework - presented at ...
 
Slides from CARR 2012 WS - Enhancing Matrix Factorization Through Initializat...
Slides from CARR 2012 WS - Enhancing Matrix Factorization Through Initializat...Slides from CARR 2012 WS - Enhancing Matrix Factorization Through Initializat...
Slides from CARR 2012 WS - Enhancing Matrix Factorization Through Initializat...
 
Fast ALS-Based Tensor Factorization for Context-Aware Recommendation from Imp...
Fast ALS-Based Tensor Factorization for Context-Aware Recommendation from Imp...Fast ALS-Based Tensor Factorization for Context-Aware Recommendation from Imp...
Fast ALS-Based Tensor Factorization for Context-Aware Recommendation from Imp...
 
Recommender Systems Evaluation: A 3D Benchmark - presented at RUE 2012 worksh...
Recommender Systems Evaluation: A 3D Benchmark - presented at RUE 2012 worksh...Recommender Systems Evaluation: A 3D Benchmark - presented at RUE 2012 worksh...
Recommender Systems Evaluation: A 3D Benchmark - presented at RUE 2012 worksh...
 

Último

6.High Profile Call Girls In Punjab +919053900678 Punjab Call GirlHigh Profil...
6.High Profile Call Girls In Punjab +919053900678 Punjab Call GirlHigh Profil...6.High Profile Call Girls In Punjab +919053900678 Punjab Call GirlHigh Profil...
6.High Profile Call Girls In Punjab +919053900678 Punjab Call GirlHigh Profil...
@Chandigarh #call #Girls 9053900678 @Call #Girls in @Punjab 9053900678
 
valsad Escorts Service ☎️ 6378878445 ( Sakshi Sinha ) High Profile Call Girls...
valsad Escorts Service ☎️ 6378878445 ( Sakshi Sinha ) High Profile Call Girls...valsad Escorts Service ☎️ 6378878445 ( Sakshi Sinha ) High Profile Call Girls...
valsad Escorts Service ☎️ 6378878445 ( Sakshi Sinha ) High Profile Call Girls...
Call Girls In Delhi Whatsup 9873940964 Enjoy Unlimited Pleasure
 
( Pune ) VIP Baner Call Girls 🎗️ 9352988975 Sizzling | Escorts | Girls Are Re...
( Pune ) VIP Baner Call Girls 🎗️ 9352988975 Sizzling | Escorts | Girls Are Re...( Pune ) VIP Baner Call Girls 🎗️ 9352988975 Sizzling | Escorts | Girls Are Re...
( Pune ) VIP Baner Call Girls 🎗️ 9352988975 Sizzling | Escorts | Girls Are Re...
nilamkumrai
 
VIP Call Girls Pollachi 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Pollachi 7001035870 Whatsapp Number, 24/07 BookingVIP Call Girls Pollachi 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Pollachi 7001035870 Whatsapp Number, 24/07 Booking
dharasingh5698
 
₹5.5k {Cash Payment}New Friends Colony Call Girls In [Delhi NIHARIKA] 🔝|97111...
₹5.5k {Cash Payment}New Friends Colony Call Girls In [Delhi NIHARIKA] 🔝|97111...₹5.5k {Cash Payment}New Friends Colony Call Girls In [Delhi NIHARIKA] 🔝|97111...
₹5.5k {Cash Payment}New Friends Colony Call Girls In [Delhi NIHARIKA] 🔝|97111...
Diya Sharma
 

Último (20)

Shikrapur - Call Girls in Pune Neha 8005736733 | 100% Gennuine High Class Ind...
Shikrapur - Call Girls in Pune Neha 8005736733 | 100% Gennuine High Class Ind...Shikrapur - Call Girls in Pune Neha 8005736733 | 100% Gennuine High Class Ind...
Shikrapur - Call Girls in Pune Neha 8005736733 | 100% Gennuine High Class Ind...
 
Call Girls Sangvi Call Me 7737669865 Budget Friendly No Advance BookingCall G...
Call Girls Sangvi Call Me 7737669865 Budget Friendly No Advance BookingCall G...Call Girls Sangvi Call Me 7737669865 Budget Friendly No Advance BookingCall G...
Call Girls Sangvi Call Me 7737669865 Budget Friendly No Advance BookingCall G...
 
Top Rated Pune Call Girls Daund ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...
Top Rated  Pune Call Girls Daund ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...Top Rated  Pune Call Girls Daund ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...
Top Rated Pune Call Girls Daund ⟟ 6297143586 ⟟ Call Me For Genuine Sex Servi...
 
Trump Diapers Over Dems t shirts Sweatshirt
Trump Diapers Over Dems t shirts SweatshirtTrump Diapers Over Dems t shirts Sweatshirt
Trump Diapers Over Dems t shirts Sweatshirt
 
Yerawada ] Independent Escorts in Pune - Book 8005736733 Call Girls Available...
Yerawada ] Independent Escorts in Pune - Book 8005736733 Call Girls Available...Yerawada ] Independent Escorts in Pune - Book 8005736733 Call Girls Available...
Yerawada ] Independent Escorts in Pune - Book 8005736733 Call Girls Available...
 
Dubai Call Girls Milky O525547819 Call Girls Dubai Soft Dating
Dubai Call Girls Milky O525547819 Call Girls Dubai Soft DatingDubai Call Girls Milky O525547819 Call Girls Dubai Soft Dating
Dubai Call Girls Milky O525547819 Call Girls Dubai Soft Dating
 
Real Men Wear Diapers T Shirts sweatshirt
Real Men Wear Diapers T Shirts sweatshirtReal Men Wear Diapers T Shirts sweatshirt
Real Men Wear Diapers T Shirts sweatshirt
 
6.High Profile Call Girls In Punjab +919053900678 Punjab Call GirlHigh Profil...
6.High Profile Call Girls In Punjab +919053900678 Punjab Call GirlHigh Profil...6.High Profile Call Girls In Punjab +919053900678 Punjab Call GirlHigh Profil...
6.High Profile Call Girls In Punjab +919053900678 Punjab Call GirlHigh Profil...
 
Real Escorts in Al Nahda +971524965298 Dubai Escorts Service
Real Escorts in Al Nahda +971524965298 Dubai Escorts ServiceReal Escorts in Al Nahda +971524965298 Dubai Escorts Service
Real Escorts in Al Nahda +971524965298 Dubai Escorts Service
 
Pune Airport ( Call Girls ) Pune 6297143586 Hot Model With Sexy Bhabi Ready...
Pune Airport ( Call Girls ) Pune  6297143586  Hot Model With Sexy Bhabi Ready...Pune Airport ( Call Girls ) Pune  6297143586  Hot Model With Sexy Bhabi Ready...
Pune Airport ( Call Girls ) Pune 6297143586 Hot Model With Sexy Bhabi Ready...
 
Sarola * Female Escorts Service in Pune | 8005736733 Independent Escorts & Da...
Sarola * Female Escorts Service in Pune | 8005736733 Independent Escorts & Da...Sarola * Female Escorts Service in Pune | 8005736733 Independent Escorts & Da...
Sarola * Female Escorts Service in Pune | 8005736733 Independent Escorts & Da...
 
Katraj ( Call Girls ) Pune 6297143586 Hot Model With Sexy Bhabi Ready For S...
Katraj ( Call Girls ) Pune  6297143586  Hot Model With Sexy Bhabi Ready For S...Katraj ( Call Girls ) Pune  6297143586  Hot Model With Sexy Bhabi Ready For S...
Katraj ( Call Girls ) Pune 6297143586 Hot Model With Sexy Bhabi Ready For S...
 
valsad Escorts Service ☎️ 6378878445 ( Sakshi Sinha ) High Profile Call Girls...
valsad Escorts Service ☎️ 6378878445 ( Sakshi Sinha ) High Profile Call Girls...valsad Escorts Service ☎️ 6378878445 ( Sakshi Sinha ) High Profile Call Girls...
valsad Escorts Service ☎️ 6378878445 ( Sakshi Sinha ) High Profile Call Girls...
 
( Pune ) VIP Baner Call Girls 🎗️ 9352988975 Sizzling | Escorts | Girls Are Re...
( Pune ) VIP Baner Call Girls 🎗️ 9352988975 Sizzling | Escorts | Girls Are Re...( Pune ) VIP Baner Call Girls 🎗️ 9352988975 Sizzling | Escorts | Girls Are Re...
( Pune ) VIP Baner Call Girls 🎗️ 9352988975 Sizzling | Escorts | Girls Are Re...
 
VIP Call Girls Pollachi 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Pollachi 7001035870 Whatsapp Number, 24/07 BookingVIP Call Girls Pollachi 7001035870 Whatsapp Number, 24/07 Booking
VIP Call Girls Pollachi 7001035870 Whatsapp Number, 24/07 Booking
 
₹5.5k {Cash Payment}New Friends Colony Call Girls In [Delhi NIHARIKA] 🔝|97111...
₹5.5k {Cash Payment}New Friends Colony Call Girls In [Delhi NIHARIKA] 🔝|97111...₹5.5k {Cash Payment}New Friends Colony Call Girls In [Delhi NIHARIKA] 🔝|97111...
₹5.5k {Cash Payment}New Friends Colony Call Girls In [Delhi NIHARIKA] 🔝|97111...
 
VVIP Pune Call Girls Mohammadwadi WhatSapp Number 8005736733 With Elite Staff...
VVIP Pune Call Girls Mohammadwadi WhatSapp Number 8005736733 With Elite Staff...VVIP Pune Call Girls Mohammadwadi WhatSapp Number 8005736733 With Elite Staff...
VVIP Pune Call Girls Mohammadwadi WhatSapp Number 8005736733 With Elite Staff...
 
Call Now ☎ 8264348440 !! Call Girls in Sarai Rohilla Escort Service Delhi N.C.R.
Call Now ☎ 8264348440 !! Call Girls in Sarai Rohilla Escort Service Delhi N.C.R.Call Now ☎ 8264348440 !! Call Girls in Sarai Rohilla Escort Service Delhi N.C.R.
Call Now ☎ 8264348440 !! Call Girls in Sarai Rohilla Escort Service Delhi N.C.R.
 
(INDIRA) Call Girl Pune Call Now 8250077686 Pune Escorts 24x7
(INDIRA) Call Girl Pune Call Now 8250077686 Pune Escorts 24x7(INDIRA) Call Girl Pune Call Now 8250077686 Pune Escorts 24x7
(INDIRA) Call Girl Pune Call Now 8250077686 Pune Escorts 24x7
 
Call Girls Ludhiana Just Call 98765-12871 Top Class Call Girl Service Available
Call Girls Ludhiana Just Call 98765-12871 Top Class Call Girl Service AvailableCall Girls Ludhiana Just Call 98765-12871 Top Class Call Girl Service Available
Call Girls Ludhiana Just Call 98765-12871 Top Class Call Girl Service Available
 

Challenges Encountered by Scaling Up Recommendation Services at Gravity R&D

  • 1. Scaling UP Challenges Encountered Scaling Up Recommendation Services @Gravity R&D Bottyán Németh
  • 2. Who we are and what we do Gravity R&D is a recommender system vendor company. We provide recommendation as a service since 2009 for our customers all around the globe. 2
  • 3. How we imagine growth? 3 ?
  • 4. How we imagine growth? 4
  • 5. How it actually happens? 5 ?
  • 6. How it actually happens? 6
  • 7. # of requests 7 Vatera.hu largest online marketplace in Hungary served by one “server” Alexa TOP100 video chat webpage (~40M recommendation requests / day):  Served by 5 application servers and 1 DB  Too many events to store in MySQL  using Cassandra (v0.6)  Training time for IALS too long  speedup by IALS1  Max. 5 sec latency in “product” availability
  • 8. Using new/beta technologies 8 Cassandra (v0.6) Nginx (v0.5) (22% of top 1M sites) Kafka (v0.8) MySQL auto. failover
  • 9. Reaching the limits 9 Even if the technology is widely used if you reach it’s limits the optimization is very costly / time consuming. Java GC – service collapsed because increased minor GC times due to a JVM bug (26th of January 2013) Maintaining MySQL with lots of data (optimize table, slave replication lag, faster storage device)
  • 10. Complexity increases 10 There is always a business request or an algorithmic development which requires more resources.
  • 12. Infrastructure 12 Currently 200+ hosts and 3500+ services monitored 0 50 100 150 200 250 2008 2009 2010 2011 2012 2013 2014 2015 2016 Number of servers
  • 13. # of items 13 How to store item model / metadata in memory to serve requests fast?
  • 14. # of items 14 How to store item model / metadata in memory to serve requests fast? VS. Auto increment IDs for the items? 231 not enough
  • 15. Preconceptions 15 More data better results. If the CTR of a new algorithm is low than the old algorithm is better. Daily retrain is enough.
  • 17. 100+ Algorithms 17 0 10 20 30 40 50 60 0 20 40 60 80 100 120 Number of times an algorithm is used
  • 18. Now 18 • Performance: Gravity’s performance oriented architecture enables real-time response to the always changing environment and user behavior • Algorithms: more than 100 different recommendation algorithm enables true personalization and to reach the highest KPIs in different domains • Infrastructure: fast response times all around the globe and data security thanks to the private cloud infrastructure located in 4 different data centers • Flexibility: the advanced business rule engine with intuitive user interface allows to satisfy various business requirements Performance 140M requests served daily Algorithms 30 man-years invested Infrastructure 4 data centers globally Flexibility 100s of logics configurable
  • 19. Cross the river when you come to it 19