While microservices are hype, there are reasons to migrate to such an architecture. Also, benefits are to be evaluated in comparison to requirements. Are you ready to take a fresh look?

1. THE DARK SIDE
OF MICRO-
SERVICES
@NICOLAS_FRANKEL

2. #microservices
ME, MYSELF AND I
@nicolas_frankel
2
 Developer/Software -
/Solution Architect
• Java
• As consultant

3. #microservices
HYBRIS, AN SAP COMPANY
3
@nicolas_frankel #springboot

4. #microservices
MICRO-SERVICES ON PAPER
4
@nicolas_frankel

5. #microservices
QUOTE OF THE DAY
@nicolas_frankel
5
“[…] if people can't build
monoliths properly,
microservices won't help”
- Simon Brown
https://twitter.com/simonbrown/status/573072777147777024
https://genehughson.wordpress.com/2015/03/08/microservice-architectures-arent-for-everyone/

6. #microservices
QUOTE OF THE DAY
@nicolas_frankel
6
“I see you have a poorly
structured monolith. Would
you like me to convert it into a
poorly structured set of
microservices?”
- Architect Clippy
https://twitter.com/architectclippy/status/570025079825764352
https://genehughson.wordpress.com/2015/03/08/microservice-architectures-arent-for-everyone/

7. #microservices
MICRO-SERVICES IN REAL LIFE
7
@nicolas_frankel

8. #microservices
SUMMARY
What are micro-services?
What benefits to expect?
What requirements?
What’s in it for you now?
8
@nicolas_frankel

9. #microservices
A TENTATIVE
DEFINITION
THE DARK SIDE OF MICRO-SERVICES
9
@nicolas_frankel

10. #microservices
THE MONOLITH
“A monolithic application
puts all its functionality into a
single process…”
http://martinfowler.com/articles/microservices.html
10
@nicolas_frankel

11. #microservices
SCALING A MONOLITH
“… and scales by replicating
the monolith on different
servers”
11
@nicolas_frankel

12. #microservices
ASSUMPTION
All functionalities of your
application have the same
load
• Really?
12
@nicolas_frankel

13. #microservices
EXAMPLE: E-COMMERCE SHOP
Catalog
Cart
Login
Payment
etc.
13
@nicolas_frankel

14. #microservices
MICRO-SERVICES
“A microservices
architecture puts each
element of functionality into
a separate service…”
14
@nicolas_frankel

15. #microservices
SCALING MICRO-SERVICES
“… and scales by
distributing these services
across servers, scaling as
needed”
15
@nicolas_frankel

16. #microservices
SINGLE RESPONSIBILITY PRINCIPLE
 Component-based
 Wrapping a single business
capability
• Fine-grained
 Decoupled from one
another
 Communicating via simple
channels
16
@nicolas_frankel

17. #microservices
BENEFITS
THE DARK SIDE OF MICRO-SERVICES
17
@nicolas_frankel

18. #microservices
DIFFERENTIAL SCALING
Optimizing resources
18
@nicolas_frankel

19. #microservices
DECOUPLED LIFECYCLES
Enable continuous delivery
• No more release trains!
19
@nicolas_frankel

20. #microservices
CONTINUOUS/EVOLUTIONARY DESIGN
No necessary fixed upfront
design
• Allows for continuous changes
20
@nicolas_frankel

21. #microservices
POLYGLOT PERSISTENCE
Choose the best persistence
tier for the job
21
@nicolas_frankel

22. #microservices
POLYGLOT
Choose the best language
for the job
22
@nicolas_frankel

23. #microservices
THE FIRST AND FOREMOST BENEFIT
@nicolas_frankel
23
Scaling your development
team!

24. #microservices
REQUIREMENTS
THE DARK SIDE OF MICRO-SERVICES
24
@nicolas_frankel

25. #microservices
INFRASTRUCTURE AUTOMATION
To deliver continuously
25
@nicolas_frankel

26. #microservices
FORMALIZED DOCUMENTATION
To help service users
26
@nicolas_frankel

27. #microservices
MONITORING
Cross-components
27
@nicolas_frankel

28. #microservices
CORE ISSUES
THE DARK SIDE OF MICRO-SERVICES
29
@nicolas_frankel

29. #microservices
30
@nicolas_frankel

30. #microservices
THE NETWORK!
From a simple API call to a
network call…
31
@nicolas_frankel

31. #microservices
FALLACIES OF DISTRIBUTED
COMPUTING
1. The network is reliable.
2. Latency is zero.
3. Bandwidth is infinite.
4. The network is secure.
5. Topology doesn’t change.
6. There is one administrator.
7. Transport cost is zero.
8. The network is homogeneous.
@nicolas_frankel
32

32. #microservices
NEW PROBLEMS ARISE
Latency
• Asynchronicity
• Performance drop
33
@nicolas_frankel

33. #microservices
NEW PROBLEMS ARISE
Handling failures
• Different handling
• Unavailable service
34
@nicolas_frankel

34. #microservices
NEW PROBLEMS ARISE
Serialization/deserialization
• Performance drop
35
@nicolas_frankel

35. #microservices
STRIKE A BALANCE
Between scalability and
performance
36
@nicolas_frankel

36. #microservices
NOT ONLY TECHNICAL ISSUES
@nicolas_frankel
37

37. #microservices
AGILE?
Trying to do agile
Doing agile
Being agile
38
@nicolas_frankel

38. #microservices
ORGANIZING TEAMS
Limited size
• Fed by 2 pizzas
Autonomous
Self-organizing
39
@nicolas_frankel

39. #microservices
CONWAY’S LAW
“organizations which design systems ... are
constrained to produce designs which are copies
of the communication structures of these
organizations”
@nicolas_frankel
40

40. #microservices
ORGANIZATION AT NETFLIX
@nicolas_frankel
41

41. #microservices
THE REAL ISSUE OF MICRO-SERVICES
Makes middle managers
unnecessary
42
@nicolas_frankel

42. #microservices
DO YOU THINK…
@nicolas_frankel
43
…anybody would willingly
make themselves not
necessary?

43. #microservices
BETTER SUITED TO
@nicolas_frankel
44
Small companies
Rather flat organizations

44. #microservices
THE LIGHT SIDE
THE DARK SIDE OF MICRO-SERVICES
45
@nicolas_frankel

45. #microservices
AND YET…
46
@nicolas_frankel

46. #microservices
THERE ARE MICRO-SERVICES…
… and web-services
• RESTful
47
@nicolas_frankel

47. #microservices
MICRO-SERVICES ENABLERS
@nicolas_frankel
48
Hystrix
Ribbon
Eureka
Feign
Etc.

48. #microservices
FEATURES
@nicolas_frankel
49
Circuit breaker
• Fail-fast
Fallback
Cache
Collapse
Dashboard
And many more…
https://github.com/Netflix/Hystrix/wiki/How-To-Use

49. #microservices
CIRCUIT-BREAKER
@nicolas_frankel
50

50. #microservices
FALLBACK
@nicolas_frankel
51

51. #microservices
REQUEST
COLLAPSE
@nicolas_frankel
52

52. #microservices
REQUEST
COLLAPSE
@nicolas_frankel
53

53. #microservices
DASHBOARD
@nicolas_frankel
54

54. #microservices
API FLAVORS
@nicolas_frankel
55
Synchronous
Asynchronous
Reactive

55. #microservices
RIBBON
@nicolas_frankel
56
Client-side load-balancer
Pluggable rules

56. #microservices
COMMONS RULES
@nicolas_frankel
57
Round robbin
Availability filtering
Weighted response time
<insert your own there>

57. #microservices
SERVER LIST
@nicolas_frankel
58
Ad hoc
Configuration-based
Discovery enabled
• Through Eureka

58. #microservices
EUREKA
@nicolas_frankel
59
Registry for middle tier load
balancing
• Server
• Client

59. #microservices
FEIGN
@nicolas_frankel
60
Annotation based REST
client

60. #microservices
SPRING CLOUD*
@nicolas_frankel
61

61. #microservices
SUMMARY
THE DARK SIDE OF MICRO-SERVICES
62
@nicolas_frankel

62. #microservices
ADOPT AN ENGINEER MINDSET
@nicolas_frankel
63
Evaluate benefits
Evaluate costs
Evaluate feasibility
• Organizational
• (Technical)

63. #microservices
Q&A
@nicolas_frankel
64
http://blog.frankel.ch/
@nicolas_frankel
http://frankel.in/