A DevOps State of Mind with Microservices, Containers and Kubernetes Chris Van Tuin Red Hat - Chief Technologist, NA West

1. A DevOps State of Mind
with Microservices, Containers
and Kubernetes
Chris Van Tuin
Chief Technologist, NA West / Silicon Valley
cvantuin@redhat.com

2. “Only the paranoid survive”
- Andy Grove, 1996

3. Retail Finance Media
Transportation
?
?
SOFTWARE DISRUPTS BUSINESS

4. $4.3 billion in unsold inventory
H&M gets hit with the ‘Amazon effect’
https://www.marketwatch.com/story/hm-gets-hit-with-the-amazon-effect-2018-04-03

5. 6
months
SUPPLY CHAIN BOTTLENECK
in Era of Fast Fashion

6. 3 months
6
months
SUPPLY CHAIN BOTTLENECK
in Era of Fast Fashion

7. 1 to 8
weeks
3 months
6
months
SUPPLY CHAIN BOTTLENECK
in Era of Fast Fashion

8. SALES GROWTH VS LEAD TIMES

9. 2 year %
STOCK PERFORMANCE
+105%
-9%
-57%
(13 year low)

10. “H&M need to make
sure they’re innovating
ahead of the curve,
not just to catch up”
H&M’s position is
magnified by the fact
that they recognized the
problem later than their
peers
H&M investing in I.T. to …
Speed
Up
Innovation
Amplify &
Shorten
Feedback Loop

11. DISRUPT OR BE DISRUPTED

12. Innovation
Months
THE AVERAGE ENTERPRISE DOES
DEPLOYMENTS EVERY
6 TO 9 MONTHS.

13. DEV QA OPS
“THROW IT OVER THE WALL”
Challenges
1. Environment drift
2. Large change set =
more bugs, more testing
3. Extended feedback loop
Walled off people, walled off processes, walled off technologies

14. HOW DOES I.T. TRANSFORM FROM
A COST CENTER INTO AN INNOVATION CENTER?
Months
Innovation

15. THE DEVOPS FACTORY

16. DEVOPS
Speeding Up Time to Market
Leverage
tooling &
automation
Reduce
organizational
silos
Accept
failure
as normal
Implement
gradual
change
Measure
everything

17. DEV QA OPS
Linux + Containers
IaaS
Orchestration
CI/CD
Source Control Management
Collaboration
Build and Artifact Management
Testing
Frameworks
Culture Process Technology
+ +
Collaborative
Transparent
Open
Agile
Continuous
Iterative
Automate
Open Source
Hybrid
Dynamic
THE DEVOPS FACTORY

18. THE VALUE OF DEVOPS
Faster
Time to
Market
Security
More time
To
Innovate
Faster
Resolution
Of Problems
More Stable
Operating
Environment
Improved
Communication
& Collaboration
Less Complex
Less Risk

19. CULTURE

20. Collaboration Shared problems
are solved faster
+
OPEN SOURCE
CULTURE
Transparency
(both access and
the ability to act)
Working together
creates necessary
standardization
+
IT’S NOT JUST CODE; IT’S CULTURE

21. PROCESS

22. AUTOMATION: CI/CD PIPELINE
Continuous
Integration
Continuous
Build
Continuous
Deployment
Feedback Loop

23. TECHNOLOGY

24. KEY TECHNOLOGY TRENDS

25. Containers at Scale

26. 4
● Are there known vulnerabilities in
the application layer?
● Are the runtime and OS layers up
to date?
● How frequently will the container
be updated and how will I know
when it’s updated?
CONTENT: EACH LAYER MATTERS
CONTAINER
OS
RUNTIME
APPLICATION
CONTENT: EACH LAYER MATTERS
AYER MATTERS
CONTAINER
OS
RUNTIME
APPLICATION
JAR CONTAINER

27. docker.io
RegistryPrivate
RegistryRed Hat
Certified
FROM fedora:latest
CMD echo “Hello”
Build file
Physical, Virtual, Cloud
Image Container
Build RunShip
CONTAINERS: BUILD, SHIP, RUN

28. CI/CD

29. Scheduling Monitoring
Persistence
DiscoveryLifecycle & health
Scaling Aggregation Security
CONTAINERS AT SCALE
BARE METAL VIRTUAL PRIVATE CLOUD PUBLIC CLOUD

30. DEV QA OPS
Open organization +
cross-functional teams
Software factory
automation
CI/CD pipelines
with feedback
Culture Process Technology
+ +
THE DEVOPS FACTORY

31. Enabling Microservices

32. code config data
Kubernetes
configmaps
secrets
Container
image
Traditional
data services,
Kubernetes
persistent volumes
TREAT CONTAINERS AS IMMUTABLE

33. Web Database
replicas=1,
role=db
replicas=2,
role=web
ORCHESTRATION
Deployment, Declarative
Nodes
Controller
Manager
&
Data Store
(etcd)

34. role=web role=web
ORCHESTRATION
Schedule + Provision Pods (Compute/Storage/Network)
Image
Registry
Pods
Nodes
Web
replicas=2,
role=web
ReplicaSet

35. role=web role=db role=web
Pods
Nodes
Image
Registry
ORCHESTRATION
Schedule + Provision Pods (Compute/Storage/Network)
Web
replicas=2,
role=web
ReplicaSet
Database
replicas=1,
role=db
StatefulSet

36. Web Database
role=web role=db role=web
replicas=1,
role=db
replicas=2,
role=web
DISCOVERY
Service (Load Balancer)
Pods
Nodes
Services
Controller
Manager
&
Data Store
(etcd)

37. HEALTH CHECK
Monitoring &
Logging
Pods
Nodes
Services
Web Database
role=web role=db role=web
replicas=1,
role=db
replicas=2,
role=web

38. HEALTH CHECK
Pods
Nodes
Services
Web Database
role=web role=db role=web
replicas=1,
role=db
replicas=2,
role=web
role=web
Controller
Manager
&
Data Store
(etcd)

39. Web Database
replicas=1,
role=db
replicas=2,
role=web
HEALTH CHECK
Pods
Nodes
Services
role=web role=db role=web
Controller
Manager
&
Data Store
(etcd)

40. Web Database
replicas=1,
role=db
replicas=2,
role=web
AUTO-SCALE
Monitoring &
Logging
80% CPU
Pods
Nodes
Services
role=web role=db role=web

41. Web Database
replicas=1,
role=dbreplicas=3
role=web
AUTO-SCALE
80% CPU
Pods
Nodes
Services
role=web role=db role=web role=web
Controller
Manager
&
Data Store
(etcd)

42. Pods
Nodes
Services
Web Database
replicas=1,
role=dbreplicas=3
role=web
AUTO-SCALE
50% CPU
role=web role=db role=web role=web
Controller
Manager
&
Data Store
(etcd)

43. Recreate

44. Version 1 Version 1Version 1
Version 1.2
`
Tests / CI
RECREATE WITH DOWNTIME

45. Version 1 Version 1Version 1
Version 1.2
`
Tests / CI
RECREATE WITH DOWNTIME

46. Version 1.2 Version 1.2Version 1.2
RECREATE WITH DOWNTIME
Use Case
• Non-mission critical services
Cons
• Downtime
Pros
• Simple, clean
• No Schema incompatibilities
• No API versioning

47. Rolling Updates

48. Version 1 Version 1Version 1
Version 1.2
`
Tests / CI
ROLLING UPDATES with ZERO DOWNTIME

49. Deploy new version and wait until it’s ready…
Version 1 Version 1 V1.2
Health Check:
readiness probe
e.g. tcp, http, script
V1

50. Each container/pod is updated one by one
Version 1.2
50%
Version 1 V1 V1.2

51. Each container/pod is updated one by one
Version 1.2Version 1.2Version 1.2
100%
Use Case
• Horizontally scaled
• Backward compatible
API/data
• Microservices
Cons
• Require backward
compatible APIs/data
• Resource overhead
Pros
• Zero downtime
• Reduced risk, gradual
rollout w/health checks
• Ready for rollback

52. Blue / Green Deployment

53. Version 1
BLUE / GREEN DEPLOYMENT
Route
BLUE

54. Version 1
BLUE / GREEN DEPLOYMENT
Version 1.2
BLUE GREEN

55. Version 1 Tests / CI
BLUE / GREEN DEPLOYMENT
Version 1.2
BLUE GREEN

56. Version 1 Version 1.2
BLUE / GREEN DEPLOYMENT
Route
Version 1.2
BLUE GREEN

57. Version 1
BLUE / GREEN DEPLOYMENT
Rollback
Route
Version 1.2
BLUE GREEN
Use Case
• Self-contained micro
services (data)
Cons
• Resource overhead
• Data synchronization
Pros
• Low risk, never
change production
• No downtime
• Production like testing
• Rollback

58. RAPID INNOVATION &
EXPERIMENTATION

59. ”only about 1/3 of ideas improve the metrics
they were designed to improve.”
Ronny Kohavi, Microsoft (Amazon)
MICROSERVICES
RAPID INNNOVATION & EXPERIMENTATION

60. CONTINUOUS FEEDBACK LOOP

61. A/B TESTING USING CANARY DEPLOYMENTS

62. Version 1.2Version 1
100%
Tests / CI
Version 1.2
Route
25% Conversion Rate ?! Conversion Rate
CANARY DEPLOYMENTS

63. 50% 50%
Version 1.2Version 1
Route
Version 1.2
25% Conversion Rate 30% Conversion Rate
CANARY DEPLOYMENTS

64. 25% Conversion Rate
100%
Version 1 Version 1.2
Route
Version 1.2
30% Conversion Rate
CANARY DEPLOYMENTS

65. Version 1.2Version 1
100%
Route
Rollback
25% Conversion Rate 20% Conversion Rate
CANARY DEPLOYMENTS

66. THANK YOU
linkedin: Chris Van Tuin
email: cvantuin@redhat.com
twitter: @chrisvantuin