SlideShare a Scribd company logo
DevOps + BigData @ RealTime
DevOps+
(DevOps in a startup)
DevOps+
● Fixing the electrical outage
DevOps+
● Fixing the electrical outage
● Cycling to Maplin for an extension lead
● Dealing with the sewage from the blocked toilet
BigData
● Also known as - Data
● BigData is writing Doom or Elite for a 32k machine:
BigData
● Or just because you can’t quite fit it on a 39” rotating disk:
@RealTime
● Can you survive this ?
The Challenge - The 2014 (Football) World Cup
Analyse every Tweet issued during World Cup games in the 2014 competition
The Challenge - The 2014 (Football) World Cup
The Challenge - The 2014 (Football) World Cup
How busy was it expected to be ?
The Challenge - The 2014 (Football) World Cup
How busy was it expected to be ?
● Predicted 1.2M Tweets per hour
● Average time for our analysis is 100ms - lengthy texts up to 150ms
● Our analysis time increases logarithmically with sentence count and sentence length so needed to avoid pathological cases
● 150ms x 1.2M = 180k sec = 3000 minutes = between 33 and 50 hours
The Solution - AWS
● Lots of EC2 instances
● 3 Availability Zones
● ELB
The Solution - Unattended Build
● Cloud Formation
● Puppet
● Route53
● ELB
● 2 min
The Challenge - The 2014 (Football) World Cup
How busy was it ?
● Predicted peak 1.2M Tweets per hour
● Actual peak 2.5M Tweets per hour
● Whole tournament avg 17M per day for 30 days
● Peak 33M in one day
The Challenge - The 2014 (Football) World Cup
How busy was it ?
● Predicted peak 1.2M Tweets per hour
● Actual peak 2.5M Tweets per hour
● Whole tournament avg 17M per day for 30 days
● Peak 33M in one day
The Challenge - The 2014 (Football) World Cup
How busy was it ?
● Predicted peak 1.2M Tweets per hour
● Actual peak 2.5M Tweets per hour
● Whole tournament avg 17M per day for 30 days
● Peak 33M in one day
● A Long Time Ago In A Galaxy Far Far Away
● Carhire3000 began - we had 6 servers
● Building a new server took a week
● 8 years later:-
○ 2 acquisitions
○ Booking.com
○ $996 million
○ 600+ servers
○ DevOps team of 5
● Booking.com methodology - Unattended build in 5 minutes
● Control Room - Monitoring health of website
● Control Room - Columnar Databases
● Columnar Databases
● Fire & Forget UDP events
● What do you monitor ?
● Experiments & Z-score
● DB Queries < 2 seconds
● Hot event data in MySQL - Hands Off !
● 5 minutes later moved to columnar storage database
DevOps Oxford- DevOps + BigData @ RealTime

More Related Content

Viewers also liked

R.ADAMS
R.ADAMSR.ADAMS
R.ADAMS
Rhonda Adams
 
Proceso Tecnológico
Proceso TecnológicoProceso Tecnológico
Proceso Tecnológico
Faiver Andres choconta bolivar
 
Comunicación interactiva
Comunicación interactiva Comunicación interactiva
Comunicación interactiva
Nayeli Marquina
 
Actividad 2
Actividad 2Actividad 2
Actividad 2
IKER BILBAO BATIZ
 
La castañeda
La castañedaLa castañeda
La castañeda
EVA RAMÍREZ
 
La informacion
La informacionLa informacion
La informacion
Claudia1081
 
tabletop
tabletoptabletop
tabletop
Rhonda Adams
 
Presentation 1 1
Presentation 1 1Presentation 1 1
Presentation 1 1
jose hernandez villamizar
 
Caso carlos
Caso carlosCaso carlos
Caso carlos
leoruiz88
 
La Escuela y yo
La Escuela y yoLa Escuela y yo
La Escuela y yo
Elian López
 
Trabalho vilmar slide
Trabalho vilmar slideTrabalho vilmar slide
Trabalho vilmar slide
Renata Miranda
 
OAHSPRING2016_RENOVATION ARTICLE
OAHSPRING2016_RENOVATION ARTICLEOAHSPRING2016_RENOVATION ARTICLE
OAHSPRING2016_RENOVATION ARTICLE
Simon Melanson
 
KapustaBell
KapustaBellKapustaBell
KapustaBell
Cindi Rivera
 

Viewers also liked (15)

R.ADAMS
R.ADAMSR.ADAMS
R.ADAMS
 
Proceso Tecnológico
Proceso TecnológicoProceso Tecnológico
Proceso Tecnológico
 
Comunicación interactiva
Comunicación interactiva Comunicación interactiva
Comunicación interactiva
 
Actividad 2
Actividad 2Actividad 2
Actividad 2
 
La castañeda
La castañedaLa castañeda
La castañeda
 
La informacion
La informacionLa informacion
La informacion
 
tabletop
tabletoptabletop
tabletop
 
feeling bue
feeling buefeeling bue
feeling bue
 
Presentation 1 1
Presentation 1 1Presentation 1 1
Presentation 1 1
 
Caso carlos
Caso carlosCaso carlos
Caso carlos
 
La Escuela y yo
La Escuela y yoLa Escuela y yo
La Escuela y yo
 
Trabalho vilmar slide
Trabalho vilmar slideTrabalho vilmar slide
Trabalho vilmar slide
 
Marcel bałdowski
Marcel bałdowskiMarcel bałdowski
Marcel bałdowski
 
OAHSPRING2016_RENOVATION ARTICLE
OAHSPRING2016_RENOVATION ARTICLEOAHSPRING2016_RENOVATION ARTICLE
OAHSPRING2016_RENOVATION ARTICLE
 
KapustaBell
KapustaBellKapustaBell
KapustaBell
 

Similar to DevOps Oxford- DevOps + BigData @ RealTime

THE RISE AND FALL OF SERVERLESS COSTS - TAMING THE (SERVERLESS) BEAST
THE RISE AND FALL OF SERVERLESS COSTS - TAMING THE (SERVERLESS) BEASTTHE RISE AND FALL OF SERVERLESS COSTS - TAMING THE (SERVERLESS) BEAST
THE RISE AND FALL OF SERVERLESS COSTS - TAMING THE (SERVERLESS) BEAST
Opher Dubrovsky
 
Big data @ Hootsuite analtyics
Big data @ Hootsuite analtyicsBig data @ Hootsuite analtyics
Big data @ Hootsuite analtyics
Claudiu Coman
 
Childhood Dream to Pi World Record by Emma Haruka, Developer Advocate @Google
Childhood Dream to Pi World Record by Emma Haruka, Developer Advocate @GoogleChildhood Dream to Pi World Record by Emma Haruka, Developer Advocate @Google
Childhood Dream to Pi World Record by Emma Haruka, Developer Advocate @Google
Paris Women in Machine Learning and Data Science
 
Playlist preprogramming, from (almost) nothing to something cool, ASKIP
Playlist preprogramming, from (almost) nothing to something cool, ASKIPPlaylist preprogramming, from (almost) nothing to something cool, ASKIP
Playlist preprogramming, from (almost) nothing to something cool, ASKIP
Pierre BERTRAND
 
Balogh gyorgy big_data
Balogh gyorgy big_dataBalogh gyorgy big_data
Balogh gyorgy big_data
LogDrill
 
stackconf 2021 | Help, My Datacenter is on Fire
stackconf 2021 | Help, My Datacenter is on Firestackconf 2021 | Help, My Datacenter is on Fire
stackconf 2021 | Help, My Datacenter is on Fire
NETWAYS
 
Help , My Datacenter is on fire
Help , My Datacenter is on fireHelp , My Datacenter is on fire
Help , My Datacenter is on fire
Kris Buytaert
 
Continuous Infrastructure First
Continuous Infrastructure FirstContinuous Infrastructure First
Continuous Infrastructure First
Kris Buytaert
 
OSMC 2023 | Large-scale logging made easy by Alexandr Valialkin
OSMC 2023 | Large-scale logging made easy by Alexandr ValialkinOSMC 2023 | Large-scale logging made easy by Alexandr Valialkin
OSMC 2023 | Large-scale logging made easy by Alexandr Valialkin
NETWAYS
 
Data Lessons Learned at Scale
Data Lessons Learned at ScaleData Lessons Learned at Scale
Data Lessons Learned at Scale
Charlie Reverte
 
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
VictoriaMetrics
 
testmon for Python
testmon for Pythontestmon for Python
testmon for Python
tib0r
 
Druid Summit 2023 : Changing Druid Ingestion from 3 hours to 5 minutes
Druid Summit 2023 : Changing Druid Ingestion from 3 hours to 5 minutesDruid Summit 2023 : Changing Druid Ingestion from 3 hours to 5 minutes
Druid Summit 2023 : Changing Druid Ingestion from 3 hours to 5 minutes
Shivji Kumar Jha
 
Prometheus the future
Prometheus   the futurePrometheus   the future
Prometheus the future
Goutham Veeramachaneni
 
Data Science in the Cloud @StitchFix
Data Science in the Cloud @StitchFixData Science in the Cloud @StitchFix
Data Science in the Cloud @StitchFix
C4Media
 
OSMC 2018 | Learnings, patterns and Uber’s metrics platform M3, open sourced ...
OSMC 2018 | Learnings, patterns and Uber’s metrics platform M3, open sourced ...OSMC 2018 | Learnings, patterns and Uber’s metrics platform M3, open sourced ...
OSMC 2018 | Learnings, patterns and Uber’s metrics platform M3, open sourced ...
NETWAYS
 
AWS Techniques and lessons writing low cost autoscaling GitLab runners
AWS Techniques and lessons writing low cost autoscaling GitLab runnersAWS Techniques and lessons writing low cost autoscaling GitLab runners
AWS Techniques and lessons writing low cost autoscaling GitLab runners
Anthony Scata
 
Serverless for High Performance Computing
Serverless for High Performance ComputingServerless for High Performance Computing
Serverless for High Performance Computing
Luciano Mammino
 
What is Software Development by Thesys Tech Head of Development
What is Software Development by Thesys Tech Head of DevelopmentWhat is Software Development by Thesys Tech Head of Development
What is Software Development by Thesys Tech Head of Development
Product School
 
Embedded Recipes 2017 - Understanding SCHED_DEADLINE - Steven Rostedt
Embedded Recipes 2017 - Understanding SCHED_DEADLINE -  Steven RostedtEmbedded Recipes 2017 - Understanding SCHED_DEADLINE -  Steven Rostedt
Embedded Recipes 2017 - Understanding SCHED_DEADLINE - Steven Rostedt
Anne Nicolas
 

Similar to DevOps Oxford- DevOps + BigData @ RealTime (20)

THE RISE AND FALL OF SERVERLESS COSTS - TAMING THE (SERVERLESS) BEAST
THE RISE AND FALL OF SERVERLESS COSTS - TAMING THE (SERVERLESS) BEASTTHE RISE AND FALL OF SERVERLESS COSTS - TAMING THE (SERVERLESS) BEAST
THE RISE AND FALL OF SERVERLESS COSTS - TAMING THE (SERVERLESS) BEAST
 
Big data @ Hootsuite analtyics
Big data @ Hootsuite analtyicsBig data @ Hootsuite analtyics
Big data @ Hootsuite analtyics
 
Childhood Dream to Pi World Record by Emma Haruka, Developer Advocate @Google
Childhood Dream to Pi World Record by Emma Haruka, Developer Advocate @GoogleChildhood Dream to Pi World Record by Emma Haruka, Developer Advocate @Google
Childhood Dream to Pi World Record by Emma Haruka, Developer Advocate @Google
 
Playlist preprogramming, from (almost) nothing to something cool, ASKIP
Playlist preprogramming, from (almost) nothing to something cool, ASKIPPlaylist preprogramming, from (almost) nothing to something cool, ASKIP
Playlist preprogramming, from (almost) nothing to something cool, ASKIP
 
Balogh gyorgy big_data
Balogh gyorgy big_dataBalogh gyorgy big_data
Balogh gyorgy big_data
 
stackconf 2021 | Help, My Datacenter is on Fire
stackconf 2021 | Help, My Datacenter is on Firestackconf 2021 | Help, My Datacenter is on Fire
stackconf 2021 | Help, My Datacenter is on Fire
 
Help , My Datacenter is on fire
Help , My Datacenter is on fireHelp , My Datacenter is on fire
Help , My Datacenter is on fire
 
Continuous Infrastructure First
Continuous Infrastructure FirstContinuous Infrastructure First
Continuous Infrastructure First
 
OSMC 2023 | Large-scale logging made easy by Alexandr Valialkin
OSMC 2023 | Large-scale logging made easy by Alexandr ValialkinOSMC 2023 | Large-scale logging made easy by Alexandr Valialkin
OSMC 2023 | Large-scale logging made easy by Alexandr Valialkin
 
Data Lessons Learned at Scale
Data Lessons Learned at ScaleData Lessons Learned at Scale
Data Lessons Learned at Scale
 
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
Large-scale Logging Made Easy: Meetup at Deutsche Bank 2024
 
testmon for Python
testmon for Pythontestmon for Python
testmon for Python
 
Druid Summit 2023 : Changing Druid Ingestion from 3 hours to 5 minutes
Druid Summit 2023 : Changing Druid Ingestion from 3 hours to 5 minutesDruid Summit 2023 : Changing Druid Ingestion from 3 hours to 5 minutes
Druid Summit 2023 : Changing Druid Ingestion from 3 hours to 5 minutes
 
Prometheus the future
Prometheus   the futurePrometheus   the future
Prometheus the future
 
Data Science in the Cloud @StitchFix
Data Science in the Cloud @StitchFixData Science in the Cloud @StitchFix
Data Science in the Cloud @StitchFix
 
OSMC 2018 | Learnings, patterns and Uber’s metrics platform M3, open sourced ...
OSMC 2018 | Learnings, patterns and Uber’s metrics platform M3, open sourced ...OSMC 2018 | Learnings, patterns and Uber’s metrics platform M3, open sourced ...
OSMC 2018 | Learnings, patterns and Uber’s metrics platform M3, open sourced ...
 
AWS Techniques and lessons writing low cost autoscaling GitLab runners
AWS Techniques and lessons writing low cost autoscaling GitLab runnersAWS Techniques and lessons writing low cost autoscaling GitLab runners
AWS Techniques and lessons writing low cost autoscaling GitLab runners
 
Serverless for High Performance Computing
Serverless for High Performance ComputingServerless for High Performance Computing
Serverless for High Performance Computing
 
What is Software Development by Thesys Tech Head of Development
What is Software Development by Thesys Tech Head of DevelopmentWhat is Software Development by Thesys Tech Head of Development
What is Software Development by Thesys Tech Head of Development
 
Embedded Recipes 2017 - Understanding SCHED_DEADLINE - Steven Rostedt
Embedded Recipes 2017 - Understanding SCHED_DEADLINE -  Steven RostedtEmbedded Recipes 2017 - Understanding SCHED_DEADLINE -  Steven Rostedt
Embedded Recipes 2017 - Understanding SCHED_DEADLINE - Steven Rostedt
 

DevOps Oxford- DevOps + BigData @ RealTime

  • 1. DevOps + BigData @ RealTime
  • 3. DevOps+ ● Fixing the electrical outage
  • 4. DevOps+ ● Fixing the electrical outage ● Cycling to Maplin for an extension lead ● Dealing with the sewage from the blocked toilet
  • 5. BigData ● Also known as - Data ● BigData is writing Doom or Elite for a 32k machine:
  • 6. BigData ● Or just because you can’t quite fit it on a 39” rotating disk:
  • 7. @RealTime ● Can you survive this ?
  • 8. The Challenge - The 2014 (Football) World Cup Analyse every Tweet issued during World Cup games in the 2014 competition
  • 9. The Challenge - The 2014 (Football) World Cup
  • 10. The Challenge - The 2014 (Football) World Cup How busy was it expected to be ?
  • 11.
  • 12. The Challenge - The 2014 (Football) World Cup How busy was it expected to be ? ● Predicted 1.2M Tweets per hour ● Average time for our analysis is 100ms - lengthy texts up to 150ms ● Our analysis time increases logarithmically with sentence count and sentence length so needed to avoid pathological cases ● 150ms x 1.2M = 180k sec = 3000 minutes = between 33 and 50 hours
  • 13. The Solution - AWS ● Lots of EC2 instances ● 3 Availability Zones ● ELB
  • 14. The Solution - Unattended Build ● Cloud Formation ● Puppet ● Route53 ● ELB ● 2 min
  • 15. The Challenge - The 2014 (Football) World Cup How busy was it ? ● Predicted peak 1.2M Tweets per hour ● Actual peak 2.5M Tweets per hour ● Whole tournament avg 17M per day for 30 days ● Peak 33M in one day
  • 16. The Challenge - The 2014 (Football) World Cup How busy was it ? ● Predicted peak 1.2M Tweets per hour ● Actual peak 2.5M Tweets per hour ● Whole tournament avg 17M per day for 30 days ● Peak 33M in one day
  • 17. The Challenge - The 2014 (Football) World Cup How busy was it ? ● Predicted peak 1.2M Tweets per hour ● Actual peak 2.5M Tweets per hour ● Whole tournament avg 17M per day for 30 days ● Peak 33M in one day
  • 18.
  • 19.
  • 20. ● A Long Time Ago In A Galaxy Far Far Away ● Carhire3000 began - we had 6 servers ● Building a new server took a week ● 8 years later:- ○ 2 acquisitions ○ Booking.com ○ $996 million ○ 600+ servers ○ DevOps team of 5
  • 21. ● Booking.com methodology - Unattended build in 5 minutes ● Control Room - Monitoring health of website
  • 22. ● Control Room - Columnar Databases
  • 23. ● Columnar Databases ● Fire & Forget UDP events ● What do you monitor ? ● Experiments & Z-score ● DB Queries < 2 seconds ● Hot event data in MySQL - Hands Off ! ● 5 minutes later moved to columnar storage database