Slide 1
Slide 1 text
Anti-Fragile
Principles +
DevOps =
Success
@KeithResar
Slide 2
Slide 2 text
Keith Resar: Bio
Wear many hats...
Recovering Developer
Open Source Contributor and Advocate
Cloud Automation Engineer
[email protected]
@KeithResar
Slide 3
Slide 3 text
Fragile Software
DISORDER
Stress Failure
@KeithResar
Slide 4
Slide 4 text
Anti-Fragile Software
THRIVES
Stress, Failure and Change
@KeithResar
Slide 5
Slide 5 text
No content
Slide 6
Slide 6 text
dsdsa
https://azure.microsoft.com/en-us/blog/inside-azure-search-chaos-engineering/
Slide 7
Slide 7 text
dsdsa
Slide 8
Slide 8 text
No content
Slide 9
Slide 9 text
No content
Slide 10
Slide 10 text
No content
Slide 11
Slide 11 text
Meltdown breaks the most
fundamental isolation between user
applications and the operating
system, allowing a program to
access the memory, and thus also
the secrets, of other programs.
Spectre breaks the isolation
between different applications. It
allows an attacker to trick error-free
programs, which follow best
practices, into leaking their secrets.
Black Swan
Events
@KeithResar
https://meltdownattack.com/
Slide 12
Slide 12 text
Fragile
● Difficult to modify
● Cannot cope with change
● Some value when stable,
liability with stress
Many organisations have applications
(mainframes for example) which are
impossible to change, very expensive to
maintain, but still running on high cost
as they are very critical to the business.
@KeithResar
Slide 13
Slide 13 text
Fragile
● Withstands stress/change to a point
● Does not adapt
● Eventually fails and becomes a
liability
Robust
@KeithResar
Slide 14
Slide 14 text
Fragile
● Designed with stress in mind
● Adaptability features
● Best engineered solution
Robust Resilient
@KeithResar
Slide 15
Slide 15 text
Fragile
● Created with change in mind
● Improves with stress
● Not just technical
Creating an Anti-Fragile system is much
harder, but once it’s in place it drives
the business based on change.
Anti-Fragile
Robust Resilient
@KeithResar
Slide 16
Slide 16 text
No content
Slide 17
Slide 17 text
Individual
just one pretty plane
@KeithResar
Slide 18
Slide 18 text
Population
the entire system
@KeithResar
Slide 19
Slide 19 text
Mistake
affects individual
@KeithResar
Slide 20
Slide 20 text
@KeithResar
Slide 21
Slide 21 text
Good Mistake
IMPROVES
Population
@KeithResar
Slide 22
Slide 22 text
Bad Mistake
No introspection
Doesn’t exploit
Gives excuses
@KeithResar
Slide 23
Slide 23 text
He who has never
sinned is less reliable
than he who has only
sinned once
@KeithResar
Slide 24
Slide 24 text
Anti-Fragile?
RANDOM EVENTS
⬆ or ⬇
UPSIDE
@KeithResar
Slide 25
Slide 25 text
No content
Slide 26
Slide 26 text
Specific steps to
introduce
anti-fragility
Slide 27
Slide 27 text
Code
...and we’ve already failed
Test Deploy Run Scale
Identify Fragile Points
@KeithResar
Slide 28
Slide 28 text
Code Test Deploy Run Scale
Identify Fragile Points
@KeithResar
Slide 29
Slide 29 text
Code Test Deploy Run Scale
Identify Fragile Points
Smaller code drops
@KeithResar
Slide 30
Slide 30 text
Code Test Deploy Run Scale
Identify Fragile Points
Incorporate feedback loops, including
data from prod ops
@KeithResar
Slide 31
Slide 31 text
Code Test Deploy Run Scale
Identify Fragile Points
Improved visibility
@KeithResar
Slide 32
Slide 32 text
Code Test Deploy Run Scale
Identify Fragile Points
More frequent deployments
More repeatable deployments
@KeithResar
Slide 33
Slide 33 text
Platform Encourages Systems Thinking
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
DEVELOPER GIT SERVER ARTIFACT REPOSITORY
OPENSHIFT
CI/CD PIPELINE
(JENKINS)
IMAGE BUILD
& DEPLOY
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
NON-PROD PROD
DEV
@KeithResar
https://www.openshift.com
Slide 34
Slide 34 text
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
DEVELOPER GIT SERVER ARTIFACT REPOSITORY
OPENSHIFT
CI/CD PIPELINE
(JENKINS)
IMAGE BUILD
& DEPLOY
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
PROMOTE
TO TEST
NON-PROD PROD
DEV TEST
Platform Encourages Systems Thinking
@KeithResar
https://www.openshift.com
Slide 35
Slide 35 text
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
DEVELOPER GIT SERVER ARTIFACT REPOSITORY
OPENSHIFT
CI/CD PIPELINE
(JENKINS)
IMAGE BUILD
& DEPLOY
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
PROMOTE
TO TEST
PROMOTE
TO UAT
NON-PROD PROD
DEV TEST UAT
Platform Encourages Systems Thinking
@KeithResar
https://www.openshift.com
Slide 36
Slide 36 text
ServiceNow
JIRA Service Desk
Zendesk
BMC Remedy
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
DEVELOPER GIT SERVER ARTIFACT REPOSITORY
OPENSHIFT
CI/CD PIPELINE
(JENKINS)
IMAGE BUILD
& DEPLOY
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
GO
LIVE?
PROMOTE
TO TEST
PROMOTE
TO UAT
RELEASE MANAGER
NON-PROD PROD
☒
☑
DEV TEST UAT
Platform Encourages Systems Thinking
@KeithResar
https://www.openshift.com
Slide 37
Slide 37 text
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
DEVELOPER GIT SERVER ARTIFACT REPOSITORY
OPENSHIFT
CI/CD PIPELINE
(JENKINS)
IMAGE BUILD
& DEPLOY
INTEGRATED
IMAGE
REGISTRY
OPENSHIFT
CLUSTER
GO
LIVE?
PROMOTE
TO TEST
PROMOTE
TO UAT
PROMOTE
TO PROD
RELEASE MANAGER
NON-PROD PROD
DEV TEST UAT
☒
☑
Platform Encourages Systems Thinking
@KeithResar
https://www.openshift.com
Slide 38
Slide 38 text
Code Test Deploy Run Scale
Identify Fragile Points
Deliberately exercising systems
with good mistakes
@KeithResar
Slide 39
Slide 39 text
@KeithResar
https://azure.microsoft.com/en-us/blog/inside-azure-search-chaos-engineering/
Slide 40
Slide 40 text
@KeithResar
1. Focus on population , not individuals. Force
the systems view
2. Set yourself up for success, identify common
fragile points - usually at integration spots
and not always technological, then…
3. Stress the system on your terms to improve
Anti-Fragile Takeaways
Slide 41
Slide 41 text
Thank you!
@KeithResar