2024-06 ICSA Industrial Keynote - Demystifying Hypes in Software Architecture

This is the presentation of the ICSA 2024 Industry Keynote speech by Davide Taibi

1. Demystifying Hypes in Cloud-
Native Technologies
Davide Taibi
Professor. University of Oulu

2. Software Evolution: Architecture

3. 2015s
Function as a Service
HighlyDecoupled
Software Evolution: Organization

4. New Technology Advantages

5. Software Architecture Evolution

6. Why Practitioners Adopt New
Technologies

7. MicroserviceMigrationMotivations
-Improve Maintainability
-Improve Scalability
-Reduce Costs
-Simplify Distributed Work
-Improve System understandability

8. Main Issues Identified
Davide Taibi -Colloquium Series -SPLAB ZHAW
Technical Issues
•Architectural Patterns are not clear
•Decoupling from the monolithic system
•Database migration and data splitting
Effort-Related issues
•Effort required for the DevOps infrastructure
•Effort required for library conversion
•Effort at least 20% higher

9. O’Reilly SW Architecture Conference 2018
Stop using microservices!
Move to serverless functions as soon as possible!

10. What is Serverless [3]
[3] BaldiniI. et al. (2017) Serverless Computing: Current Trends and Open Problems. In: Chaudhary S., SomaniG., Buyya
R. (eds) Research Advances in Cloud Computing. Springer, Singapore

11. FromMicroservicesto ServerlessFunctions
•Practitionersstartedmigratingfrommicroservicesto FaaS
•MixedApproach(microservices+ Functions)
•New applications100% serverless

12. Why practitioners are moving to
serverless

13. Migration Motivations
Companies Already moved to Microservices
•OPS Effort for Microservices
•Get rid of Kubernetes
•No OPS
Companies Migrating from Monolithic systems
•New (hype) technology
•Promising technology
•No initial infrastructural costs (pay as you use)
•Automatic scaling
•Lack of skilled OPS personnel

14. Migration Issues
•Developers are not used to the event-oriented programming
•Very hard to test
•Debug almost impossible
•Unknown Patterns and antipatterns
•Anomalies can generate unexpected costs!

15. Microservices or Serverless

16. Both! Serverless Microservices
Composing Serverless functions to Microservices

17. How About Micro-Frontends?

18. “
Micro-frontendsare the technical
representation of a business subdomain,
they allow independent implementationswith
same or different technology choices, finally
they avoid sharing logic with other subdomains
and they are own by a single team
21
From Domain
Driven Design
(DDD)

19. 2
2

20. How About Service Mesh?

22. Companies are rolling back
R.Su, X.Li, D.Taibi. Back to the Future: From
Microservice to Monolith. Microservices 2023, Pisa.

23. From Microservice to Monolithic
Microservices
Serverless
Micro-
Frontends
Cognitive and
Continuum Cloud

24. From Microservice to Monolithic
Microservices
Serverless
Micro-
Frontends
Cognitive and
Continuum Cloud
Reasons:
complexity, scalability, performance,
organizational reasons.

25. Our Demystification Process

26. Benefits and
Issues of new
Technologies
Patterns and
Anti-Patterns
Follow-Up
Case Studies
Technology
Assessment
Frameworks
Our Demystification Process

27. Demystifying New Hypes
Microservices
Serverless
Micro-
Frontends
Cognitive and
Continuum Cloud

28. Patterns and Anti-Patterns

29. Technology Assessment Feramework
•Microservice Assessment
Framework

30. Technology
Assessment
My favouriteexample (Blockchain)
M.E. PechDO YOU NEED A BLOCKCHAIN?
IEEE Spectrum. September 2017.
https://spectrum.ieee.org/do-you-need-a-blockchain

31. Architectural Drift Analysis and Degradation
Long-Term Architectural Choices
Shouldn’t be violated

32. Architectural Drift Analysis and Degradation
Is your architecture changed?

33. Architectural Drift Analysis and Degradation
Is your architecture changed?
Reverse Engineering –Architectural Reconstruction

34. Architectural Reconstruction
•Static
•From source code
•Dynamic
•From the execution of the system (e.g. log traces)

35. Mapping all the possible “paths” in your system
38
Static Analysis

36. 39
Credits: Google Maps
Static Analysis

37. •Source Code
•IaC
•Git Log
•Commit message
•PR
•Comments
•Issue Trackers.
•…
40
Static Analysis

38. •19Research-based tools for Architectural
reconstruction[Bakhtin et al]
•Arcan (Univ. Milano Bicocca)
•Prophet (Oulu and Baylor University)
•…
Static Analysis for Microservice-Based Applications
https://github.com/cloudhubs/prophet-web
Static Analysis: Tools
Alexander Bakhtin, XiaozhouLi, Jacopo Soldani, Antonio Brogi, Tomas Cerny, and Davide Taibi. 2023.
Tools Reconstructing Microservice Architecture: A Systematic Mapping Study. AMC@ECSA Workshop

39. 42
Dynamic Analysis
Credits: Science ABC

40. 43
Dynamic Analysis.
Street Traffic View Bicycle Traffic View

41. 18 tools
(10 commercial, 8 research-based)
[Bakhtin et al]
DynamcAnalysis: Tools
Alexander Bakhtin, XiaozhouLi, Jacopo Soldani, Antonio Brogi, Tomas Cerny, and Davide Taibi. 2023.
Tools Reconstructing Microservice Architecture: A Systematic Mapping Study. AMC@ECSA Workshop
Jaeger
Zipkin
AWS X-Ray
Netflix Interactive VIsualization

42. Existing tools shortfalls
•Reconstruction only
•Mainly for visualization
•High potential for SQA

43. What can we measure from Call
Graphs?

44. Architectural Quality
4
7
-Microservice Patterns and Anti-Patterns
-Identification
-Detection tools and methods
Anti-Patterns Detection and Visualization

45. 4
8
-Structural Coupling
-Logical Coupling
-Organizational/Architectural Coupling
Microservice Coupling Visualization Example Microservice Coupling and Anti-Pattern Visualization
Example
Architectural Quality: Coupling

46. A possible future… with generative AI
Refactoring and Quality Improvement
•Will it clean up our code?
•Will it refactor our SW Architecture?
•Will it enable to remove simple anti-patterns or prevent them?

47. Conclusion
•New Technologies are continuously introduced
•Silver Bullets do not exist
•Economical, and Technical motivations should be always considered
•Think about the issues in your system before migrating

48. Thanks
Davide Taibi
University of Oulu
[email protected]
https://m3s-cloud.github.io