Scaling N26 Technology Through Hypergrowth @folgerfonseca QCon - - PowerPoint PPT Presentation

scaling n26 technology through hypergrowth
SMART_READER_LITE
LIVE PREVIEW

Scaling N26 Technology Through Hypergrowth @folgerfonseca QCon - - PowerPoint PPT Presentation

Scaling N26 Technology Through Hypergrowth @folgerfonseca QCon London 03.2020 Building the bank world loves to use A new approach to banking Best digital user A betuer price experience +5M +$1.66Bn Customers Monthly transaction volume


slide-1
SLIDE 1 @folgerfonseca QCon London 03.2020

Scaling N26 Technology Through Hypergrowth

slide-2
SLIDE 2
slide-3
SLIDE 3
slide-4
SLIDE 4

Building the bank world loves to use

slide-5
SLIDE 5

A new approach to banking

A betuer price Best digital user experience
slide-6
SLIDE 6

+5M

Customers

26

Markets

+$680M

Funding

+$1.66Bn

Monthly transaction volume

4

Locations

+1500

Employees
slide-7
SLIDE 7 SPACES

Organise your finances.

slide-8
SLIDE 8
slide-9
SLIDE 9

N26 tech

+180 Microservices 100% In the cloud +60 Cross-functional teams +500 Deployments per week
slide-10
SLIDE 10

but

How?

slide-11
SLIDE 11

BACKGROUND

slide-12
SLIDE 12

+100K

Customers

6

Markets

+40M

Funding

+100

P&T Employees

3 Years ago

slide-13
SLIDE 13

Our challenges

Small team Relatively young Time critical
slide-14
SLIDE 14

Our challenges

Small team Relatively young Time critical
slide-15
SLIDE 15

Experimentation & Knowledge sharing 2 Days GSDD 1x every 6 weeks Lightning talks 1x Week

slide-16
SLIDE 16

HYPERGROWTH

slide-17
SLIDE 17

Hypergrowth

slide-18
SLIDE 18

More User → More Markets → More Security → More Stability → More Tech Hubs → More People → Our challenges

slide-19
SLIDE 19

PEOPLE RELEASES SERVICES RELIABILITY

slide-20
SLIDE 20 Our Goal

Bring in more teams to enable us to get more users and release the in new markets

slide-21
SLIDE 21 Backend Product QA Agile SRE Security

Company onboarding

slide-22
SLIDE 22

Team onboarding

slide-23
SLIDE 23

Buddy system

slide-24
SLIDE 24

Target operating model

Segments Group Group Team Team

slide-25
SLIDE 25

Target operating model 3 Different currencies 3 Different banking regulations +10 Payment schemes 4 Different locations

slide-26
SLIDE 26 Our Goal

Have consistency in technology in

  • rder to build up knowledge,

enable team mobility

slide-27
SLIDE 27

Microservices allows you to write each new service in a different language.

slide-28
SLIDE 28

Languages

slide-29
SLIDE 29
slide-30
SLIDE 30 Tech Radar

Decisions & Alignment

  • Based on Thoughtworks
  • Safe way to try things
  • Favor Consistency
  • Alignment across teams
slide-31
SLIDE 31

PEOPLE RELEASES SERVICES RELIABILITY

slide-32
SLIDE 32 Problem

How to avoid micro-monoliths

slide-33
SLIDE 33 Infrastructure as Code

IaC

slide-34
SLIDE 34 Infrastructure as code Servers Network Setuings
slide-35
SLIDE 35 Reduced human error Automation prevents deviation

Infra as Code Benefits

Scalable Small number of people-to-server ratio Strong separation AWS organisations enables very fine-grained segregation of services Agile Supporting change or new services easy → → → → Source controlled 4 eyes principle plus audit →
slide-36
SLIDE 36

Remember meltdown and spectre?

Example
slide-37
SLIDE 37 Meltdown Spectre
slide-38
SLIDE 38

Affected billions of systems globally

slide-39
SLIDE 39
slide-40
SLIDE 40

Vulnerabilities publicly announced

slide-41
SLIDE 41

Intel responds with initial statement

slide-42
SLIDE 42

Intel announces a fix “for the majority”

slide-43
SLIDE 43

Amazon announces updated Kernel

slide-44
SLIDE 44

Base image updated and tested

slide-45
SLIDE 45

Deployments with new image incrementally tested and rolled out

slide-46
SLIDE 46

Result? Done!

slide-47
SLIDE 47

the speed

  • f a startup and

the security of a bank

slide-48
SLIDE 48

PEOPLE RELEASES SERVICES RELIABILITY

slide-49
SLIDE 49 Continuous Delivery

CD

slide-50
SLIDE 50 Sign Off Continuous Delivery Code Review and commit Security and automated testing Artefact build Artefact Publish Dev Deploy Staging Deploy Live Canary Deploy Automated checks Automated checks Automated checks Live roll
  • ut
End
slide-51
SLIDE 51 Continuous Delivery - Live Blue Green LB Instance Instance Instance Instance Instance
slide-52
SLIDE 52 Instance Instance Instance Instance Instance Continuous Delivery - Live Blue Green LB +1
slide-53
SLIDE 53 Continuous Delivery - Live Blue Green LB +1 +1 +1 +1 +1 Instance Instance Instance Instance Instance
slide-54
SLIDE 54 Customisable Good machine selection for types of services Fine-grained control Implement fine-grained security and compliance controls

CD Benefits

Rollbacks are easy Historical AMIs enable rollback Immutable infrastructure No snowflakes in our infrastructure Scalable Multiple instances enable scaling control → → → → →
slide-55
SLIDE 55

PEOPLE RELEASES SERVICES RELIABILITY

slide-56
SLIDE 56 Problem

While having a high release rate, keep high availability and Minimize risks.

slide-57
SLIDE 57 Incidents >500 deployments per week

High Availability

slide-58
SLIDE 58

The best way to avoid incidents is by avoiding changes

slide-59
SLIDE 59

Detection Diagnostic Fix Delivery Minimize time

slide-60
SLIDE 60

Blameless Postmortem culture

High Availability

slide-61
SLIDE 61

What includes post mortem?

Preventing actions Actions to mitigate A written record The incident impact The root cause
slide-62
SLIDE 62

PEOPLE RELEASES SERVICES RELIABILITY

slide-63
SLIDE 63

Lessons learned

Keep a availability in mind Keep a close eye on service objectives Infra-as-code AWS makes it easy to implement Continuous Delivery Is enable trouble healthy automation Adjust to your journey conditions Trade off analysis
slide-64
SLIDE 64

Scaling N26 Technology Through Hypergrowth

Thank you! Any questions? @folgerfonseca