SlideShare a Scribd company logo
#APIdaysParis
☁-computing
We went through every steps of
#hypecycle
Shall we debrief ?
Adrien Blind, Ludovic Piot
starring…
Adrien Blind
@AdrienBlind
Devops Evangelist,
Docker Captain
Ludovic Piot
@lpiot
Head of
DevOps/Cloud unit
Introduction
Introduction
Cloud, bullshit bingo !
Every hopes!
Time-to-Market
accelerator
economic
Auto-magic
maintainance
security
performance
versatility
agile
automation
resiliency
With an even better offspring: #serverless
Every fears!
vendor
lock-in
Cloud-providers
spy your business
security
data privacy
deregulated
dev playground
death of
operation guys
A real market trend!
A real market trend!
Private cloud - an heterogeneous mkt
Many solutions…
… and as many definitions of what we
call private cloud computing
❏ virtualization,
❏ PaaS,
❏ Code to Prod integrated solution
❏ Infra as Code
2015 seems to be a loooong time ago.
Is CaaS changing the game?
Here comes the hype cycle
Technology trigger
Technology triggers
What do you really want?
On demand
Pay as you go
Elastic
Deliver a way we don’t know how to
in the on-premise legacy world
Time-to-Market
accelerator
security
performance
Available
everywhere
agile
resiliency
Challenge the legacy status quo?
Cloud computing is a
tremendous opportunity to
challenge how things exist in
the legacy world:
❏ Black vs. white wire
❏ Ops dpt is challenged
❏ Economically speaking
❏ operationnally speaking
❏ Green field
❏ On-demand provisionning
❏ Elasticity
First expectations
Wrong start with Google
App Engine…
… and as many definitions of
what we call private cloud
computing
❏ pricing model too mature for
consumers
❏ few runtimes available
❏ Vendor lock-in on data (BigTable)
❏ only Cloud-native apps
❏ No “pet” way for ops
Peak of inflated expectations
❏ Pay-per-Use / CAPEX vs. OPEX
= cheaper / no longer investment
Really?!
❏ Autoscaling
= app adapts itself to its workload
Automagically?!
❏ High availability
= app hosted in the Cloud are unbreakable
Without any change?!
Then marketing inflated expectations
Trough of disillusionment
❏ Pay-per-Use / CAPEX vs. OPEX
= you pay what you ask for NOT what you effectively use
Only advantage is the no-engagement thing
❏ Autoscaling
= you can script out/in-scaling following monitoring rules
Application has to be designed to behave correctly
❏ High availability
= multi-regions infrastructure have SLAs
Rest of HA is driven by your own software fail-over design
Desillusions!
Who’s responsible
On-premise Iaas Paas Caas
RESPONSIBILITIES
▪ Cloud provider
▪ Ops / hoster
▪ Dev / Client
PROPERTY
Cloud platforms may belong to
many client teams.
Platform governance is a
thing.
Hypervisor
VM
OS
Libs
conf.
Kernel
HDW
Middleware
conf.
Apps
conf.
Server Storage
Network
Logs / Metrology / Backups
Data
Runtime
conf.
Container
conf.
❏ Requires different skills
❏ Cloud vendors try to attract directly Dev
people: NoOps myth, shadow IT…
❏ Assumes consumers to be autonomous
❏ Here come DevOps ;)
From cases to APIs
Ride the raging bull
Cloud services are often
proprietary (no source code
available).
You have to master
❏ how they behave
❏ how to deal with their limitations
Often with try / fail / learn
methodology.
Patch management
Cloud providers offer
up-to-date assets.
But do not upgrade already
provisionned ones!
You have to use immutable
infrastructure management
patterns
Integration
Cloud providers offer many
services.
Building a platform means to
integrate 10s of services with
each others.
Cloud architect & devops
guys are deeply needed!
Price-driven governance
Usage changes, explosion of your VMs
counter
60% of time spent in operating Cloud
services is to optimize costs – Google
DevFinOps
“Tristar Medical Group Discovers
Solution That Reduced its AWS Costs
by 60% (~$20k per month)”
– May 9, 2018
Resiliency
Data gravitation / Reversibility
Cloud providers aim to store data.
❏ Weight of data
❏ Data format
❏ Data requesting technos
… lead to make reversibility
difficult.
High-end value services / Reversibility
Cloud providers deliver
high-end services that are
❏ quite easy to use
❏ very difficult to migrate from.
For example
❏ cognitive services
❏ authentication services
❏ Spanner
❏ Alexa chatbot
Plateau of productivity
Focus on value & better TTM
but support platform constraints
More flexibility, more tuning
capacity… but assume plumbing
Functions
ABSTRACT
FOCUSINFRA
Microservices
Monoliths
CaaS
IaaS
FaaS
& PaaS
Don’t put all one's eggs in one basket
Create an abstraction layer between your consumers and
cloud providers: let’s hybridate!
❏ Leverage on open-source / standards as much as you can. Notice how much
object storage vendors built up S3 (AWS) compliant APIs?
❏ Make an impact analysis (reversibility, etc.) before making your apps sticky with the
subsequent platform it relies on. (Mind the price upgrade of Google Maps)
❏ There’s no free lunch: possibly assume not to use everything you can?
Platform
AWS AZURE
Assume bi-modality
Containers & orchestrators
People & processes over technology
❏ Infra-as-code, software-defined-xxx: Ops progressively become Devs: you need to
engage on reskilling them!
❏ No longer be a robot, invent robots: it’s not just about hard skills!
❏ While some activities retire, new emerge: low level architecture, IT financing
metrology for instance
❏ Integrate your cloud into a broader devops or continuous delivery transformation
People > Processes > Tools
Conclusion
Conclusion
Architecture & cloud vendors
❏ App architecture & languages influenced cloud offers:
microservices, app resiliency & scalability contributed to
infrastructure commoditization
❏ Nowadays, the cloud influences you app: object storage,
serverless, cloud native apps
We created junkies, always searching for more abstraction. But
your dealer is not your friend, he as some interests in ($$$)
Subtle equilibrists choices
❏ Offload from the non valuable plumbing
❏ … but some infrastructure assets may be
finally part of your differentiators
❏ Get the control back may be valuable:
preserves autonomy, reduce operational
costs at scale...
❏ “Do as I say, not as I do”: refocusing on core-business is the exact opposite of what
cloud vendors does, being themselves fullstack
❏ Be an engineer, not a fashionista
(Re) Assess yourself
Time passed, the landscape changed
❏ Technologies are more mature
❏ Consulting/training/documentation... are more
developed
May be the time to get the control back ?
Questions ?

More Related Content

PDF
APIdays Paris 2018 - From real-life challenges to industrial IoT solutions, i...
PDF
APIdays Paris 2018 - A little less conversation, a little more action, Alain ...
PDF
APIdays Paris 2018 - What a Mesh! Laurent Doguin, DevRel VP, Clever Cloud
PPTX
APIdays Paris 2018 - Europ Assistance, the transformation through Public APIs...
PDF
APIdays Paris 2018 - GraphQL Opening Talk: The State of GraphQL, Jonathan Jal...
PDF
APIdays Paris 2018 - API Management is Dead, Augusto Marietti, CEO and Co-Fou...
PDF
BUILD with Microsoft - Radu Stefan
PPTX
Red Hat Summit 2020 - Ask the experts: Hybrid multicloud pitfalls
APIdays Paris 2018 - From real-life challenges to industrial IoT solutions, i...
APIdays Paris 2018 - A little less conversation, a little more action, Alain ...
APIdays Paris 2018 - What a Mesh! Laurent Doguin, DevRel VP, Clever Cloud
APIdays Paris 2018 - Europ Assistance, the transformation through Public APIs...
APIdays Paris 2018 - GraphQL Opening Talk: The State of GraphQL, Jonathan Jal...
APIdays Paris 2018 - API Management is Dead, Augusto Marietti, CEO and Co-Fou...
BUILD with Microsoft - Radu Stefan
Red Hat Summit 2020 - Ask the experts: Hybrid multicloud pitfalls

What's hot (18)

PDF
apidays LIVE Paris 2021 - Don’t put all of your APIs in one cloud by Adrien V...
PDF
Sergio Seabra - Red Hat - OSL19
PPTX
Red Hat Forum Poland 2019 - Red Hat Open Hybrid Cloud (keynote)
PPTX
Digital transformation buzzword or reality - Alon Fliess
PDF
apidays LIVE Singapore - Next-generation microservice architecture based on A...
PPTX
Cloud Native Demystified: Build Once, Run Anywhere!
PDF
An API-focused approach to Agile Integration
PDF
Ronen Levinson: Unified policy enforcement with opa - Architecture Next 20
PPTX
Building Modern Platforms on Microsoft Azure by Steef-Jan Wiggers
PDF
Engineering Manager, Edge Insights @Netflix
PPTX
Discover how the cloud will empower every organization on the planet to achie...
PPTX
Serverless Event-Driven Programming: Are We Ready for the Paradigm Shift?
PDF
Managin business process in microservice architecture
PPTX
Serverless operations for the iRobot fleet
PPTX
Flight Delay Compensation: How SwissRe is exploring new territories in Busine...
PDF
The role of integration in your cloud-native transformation (Richard Seroter ...
PPTX
Red Hat Forum Poland 2019 - 3 Pitfalls Everyone Should Avoid with Hybrid Mult...
PDF
An API-focused approach to Agile Integration
apidays LIVE Paris 2021 - Don’t put all of your APIs in one cloud by Adrien V...
Sergio Seabra - Red Hat - OSL19
Red Hat Forum Poland 2019 - Red Hat Open Hybrid Cloud (keynote)
Digital transformation buzzword or reality - Alon Fliess
apidays LIVE Singapore - Next-generation microservice architecture based on A...
Cloud Native Demystified: Build Once, Run Anywhere!
An API-focused approach to Agile Integration
Ronen Levinson: Unified policy enforcement with opa - Architecture Next 20
Building Modern Platforms on Microsoft Azure by Steef-Jan Wiggers
Engineering Manager, Edge Insights @Netflix
Discover how the cloud will empower every organization on the planet to achie...
Serverless Event-Driven Programming: Are We Ready for the Paradigm Shift?
Managin business process in microservice architecture
Serverless operations for the iRobot fleet
Flight Delay Compensation: How SwissRe is exploring new territories in Busine...
The role of integration in your cloud-native transformation (Richard Seroter ...
Red Hat Forum Poland 2019 - 3 Pitfalls Everyone Should Avoid with Hybrid Mult...
An API-focused approach to Agile Integration
Ad

Similar to APIdays Paris 2018 - Cloud computing - we went through every steps of the Gartner Hype cycle. Now it's time to debrief. Ludovic Piot, Lead of Customer Care, Clever Cloud (20)

PPTX
A perspective on cloud computing and enterprise saa s applications
PDF
Cloud Services Integration Automation-External
PPT
Cloud Computing
PPTX
Feet On The Ground Head In The Cloud
PPTX
Conversations in the Cloud
PPT
cloud computing
PPTX
The Cloud Is Rockin' and Rollin' In
PPT
Cloud Computing - Benefits and Challenges
PDF
Accion Labs - Rackspace - How can cloud help you?
PPTX
Cloud Computing Architecture Primer
PPT
Cloud Computing Realities - Getting past the hype and setting your cloud stra...
PPT
Pros and cons of Cloud Computing general.ppt
PPT
Cloud computing-2 (1)
PPT
Cloud Computing Impact On Small Business
PPTX
Cloud virtualization
PPT
The Cloud and Next Gen IT Gordon Haff - p camp-boston2012
PPT
cloud computng
PPTX
Evolvable architecture for hybrid multicloud with sdn
PPT
cloud computing
PPT
lect15_cloud.ppt
A perspective on cloud computing and enterprise saa s applications
Cloud Services Integration Automation-External
Cloud Computing
Feet On The Ground Head In The Cloud
Conversations in the Cloud
cloud computing
The Cloud Is Rockin' and Rollin' In
Cloud Computing - Benefits and Challenges
Accion Labs - Rackspace - How can cloud help you?
Cloud Computing Architecture Primer
Cloud Computing Realities - Getting past the hype and setting your cloud stra...
Pros and cons of Cloud Computing general.ppt
Cloud computing-2 (1)
Cloud Computing Impact On Small Business
Cloud virtualization
The Cloud and Next Gen IT Gordon Haff - p camp-boston2012
cloud computng
Evolvable architecture for hybrid multicloud with sdn
cloud computing
lect15_cloud.ppt
Ad

More from apidays (20)

PDF
apidays Munich 2025 - The Physics of Requirement Sciences Through Application...
PDF
apidays Munich 2025 - Developer Portals, API Catalogs, and Marketplaces, Miri...
PDF
apidays Munich 2025 - Making Sense of AI-Ready APIs in a Buzzword World, Andr...
PDF
apidays Munich 2025 - Integrate Your APIs into the New AI Marketplace, Senthi...
PDF
apidays Munich 2025 - The Double Life of the API Product Manager, Emmanuel Pa...
PDF
apidays Munich 2025 - Let’s build, debug and test a magic MCP server in Postm...
PDF
apidays Munich 2025 - The life-changing magic of great API docs, Jens Fischer...
PDF
apidays Munich 2025 - Automating Operations Without Reinventing the Wheel, Ma...
PDF
apidays Munich 2025 - Geospatial Artificial Intelligence (GeoAI) with OGC API...
PPTX
apidays Munich 2025 - GraphQL 101: I won't REST, until you GraphQL, Surbhi Si...
PPTX
apidays Munich 2025 - Effectively incorporating API Security into the overall...
PPTX
apidays Munich 2025 - Federated API Management and Governance, Vince Baker (D...
PPTX
apidays Munich 2025 - Agentic AI: A Friend or Foe?, Merja Kajava (Aavista Oy)
PPTX
apidays Munich 2025 - Streamline & Secure LLM Traffic with APISIX AI Gateway ...
PPTX
apidays Munich 2025 - Building Telco-Aware Apps with Open Gateway APIs, Subhr...
PPTX
apidays Munich 2025 - Building an AWS Serverless Application with Terraform, ...
PDF
apidays Helsinki & North 2025 - REST in Peace? Hunting the Dominant Design fo...
PDF
apidays Helsinki & North 2025 - Monetizing AI APIs: The New API Economy, Alla...
PDF
apidays Helsinki & North 2025 - How (not) to run a Graphql Stewardship Group,...
PDF
apidays Helsinki & North 2025 - APIs in the healthcare sector: hospitals inte...
apidays Munich 2025 - The Physics of Requirement Sciences Through Application...
apidays Munich 2025 - Developer Portals, API Catalogs, and Marketplaces, Miri...
apidays Munich 2025 - Making Sense of AI-Ready APIs in a Buzzword World, Andr...
apidays Munich 2025 - Integrate Your APIs into the New AI Marketplace, Senthi...
apidays Munich 2025 - The Double Life of the API Product Manager, Emmanuel Pa...
apidays Munich 2025 - Let’s build, debug and test a magic MCP server in Postm...
apidays Munich 2025 - The life-changing magic of great API docs, Jens Fischer...
apidays Munich 2025 - Automating Operations Without Reinventing the Wheel, Ma...
apidays Munich 2025 - Geospatial Artificial Intelligence (GeoAI) with OGC API...
apidays Munich 2025 - GraphQL 101: I won't REST, until you GraphQL, Surbhi Si...
apidays Munich 2025 - Effectively incorporating API Security into the overall...
apidays Munich 2025 - Federated API Management and Governance, Vince Baker (D...
apidays Munich 2025 - Agentic AI: A Friend or Foe?, Merja Kajava (Aavista Oy)
apidays Munich 2025 - Streamline & Secure LLM Traffic with APISIX AI Gateway ...
apidays Munich 2025 - Building Telco-Aware Apps with Open Gateway APIs, Subhr...
apidays Munich 2025 - Building an AWS Serverless Application with Terraform, ...
apidays Helsinki & North 2025 - REST in Peace? Hunting the Dominant Design fo...
apidays Helsinki & North 2025 - Monetizing AI APIs: The New API Economy, Alla...
apidays Helsinki & North 2025 - How (not) to run a Graphql Stewardship Group,...
apidays Helsinki & North 2025 - APIs in the healthcare sector: hospitals inte...

Recently uploaded (20)

PDF
KodekX | Application Modernization Development
PDF
solutions_manual_-_materials___processing_in_manufacturing__demargo_.pdf
PPTX
20250228 LYD VKU AI Blended-Learning.pptx
PDF
Unlocking AI with Model Context Protocol (MCP)
PDF
Per capita expenditure prediction using model stacking based on satellite ima...
PDF
Advanced IT Governance
PDF
cuic standard and advanced reporting.pdf
PDF
Dropbox Q2 2025 Financial Results & Investor Presentation
PPTX
Detection-First SIEM: Rule Types, Dashboards, and Threat-Informed Strategy
PDF
NewMind AI Weekly Chronicles - August'25 Week I
PDF
The Rise and Fall of 3GPP – Time for a Sabbatical?
PPT
Teaching material agriculture food technology
PDF
Electronic commerce courselecture one. Pdf
PDF
Review of recent advances in non-invasive hemoglobin estimation
PDF
Bridging biosciences and deep learning for revolutionary discoveries: a compr...
PDF
Advanced methodologies resolving dimensionality complications for autism neur...
PPTX
Cloud computing and distributed systems.
PDF
Spectral efficient network and resource selection model in 5G networks
PDF
Shreyas Phanse Resume: Experienced Backend Engineer | Java • Spring Boot • Ka...
PDF
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows
KodekX | Application Modernization Development
solutions_manual_-_materials___processing_in_manufacturing__demargo_.pdf
20250228 LYD VKU AI Blended-Learning.pptx
Unlocking AI with Model Context Protocol (MCP)
Per capita expenditure prediction using model stacking based on satellite ima...
Advanced IT Governance
cuic standard and advanced reporting.pdf
Dropbox Q2 2025 Financial Results & Investor Presentation
Detection-First SIEM: Rule Types, Dashboards, and Threat-Informed Strategy
NewMind AI Weekly Chronicles - August'25 Week I
The Rise and Fall of 3GPP – Time for a Sabbatical?
Teaching material agriculture food technology
Electronic commerce courselecture one. Pdf
Review of recent advances in non-invasive hemoglobin estimation
Bridging biosciences and deep learning for revolutionary discoveries: a compr...
Advanced methodologies resolving dimensionality complications for autism neur...
Cloud computing and distributed systems.
Spectral efficient network and resource selection model in 5G networks
Shreyas Phanse Resume: Experienced Backend Engineer | Java • Spring Boot • Ka...
Peak of Data & AI Encore- AI for Metadata and Smarter Workflows

APIdays Paris 2018 - Cloud computing - we went through every steps of the Gartner Hype cycle. Now it's time to debrief. Ludovic Piot, Lead of Customer Care, Clever Cloud

  • 1. #APIdaysParis ☁-computing We went through every steps of #hypecycle Shall we debrief ? Adrien Blind, Ludovic Piot
  • 2. starring… Adrien Blind @AdrienBlind Devops Evangelist, Docker Captain Ludovic Piot @lpiot Head of DevOps/Cloud unit
  • 5. Every fears! vendor lock-in Cloud-providers spy your business security data privacy deregulated dev playground death of operation guys
  • 6. A real market trend!
  • 7. A real market trend!
  • 8. Private cloud - an heterogeneous mkt Many solutions… … and as many definitions of what we call private cloud computing ❏ virtualization, ❏ PaaS, ❏ Code to Prod integrated solution ❏ Infra as Code 2015 seems to be a loooong time ago. Is CaaS changing the game?
  • 9. Here comes the hype cycle
  • 12. What do you really want? On demand Pay as you go Elastic Deliver a way we don’t know how to in the on-premise legacy world Time-to-Market accelerator security performance Available everywhere agile resiliency
  • 13. Challenge the legacy status quo? Cloud computing is a tremendous opportunity to challenge how things exist in the legacy world: ❏ Black vs. white wire ❏ Ops dpt is challenged ❏ Economically speaking ❏ operationnally speaking ❏ Green field ❏ On-demand provisionning ❏ Elasticity
  • 14. First expectations Wrong start with Google App Engine… … and as many definitions of what we call private cloud computing ❏ pricing model too mature for consumers ❏ few runtimes available ❏ Vendor lock-in on data (BigTable) ❏ only Cloud-native apps ❏ No “pet” way for ops
  • 15. Peak of inflated expectations
  • 16. ❏ Pay-per-Use / CAPEX vs. OPEX = cheaper / no longer investment Really?! ❏ Autoscaling = app adapts itself to its workload Automagically?! ❏ High availability = app hosted in the Cloud are unbreakable Without any change?! Then marketing inflated expectations
  • 18. ❏ Pay-per-Use / CAPEX vs. OPEX = you pay what you ask for NOT what you effectively use Only advantage is the no-engagement thing ❏ Autoscaling = you can script out/in-scaling following monitoring rules Application has to be designed to behave correctly ❏ High availability = multi-regions infrastructure have SLAs Rest of HA is driven by your own software fail-over design Desillusions!
  • 19. Who’s responsible On-premise Iaas Paas Caas RESPONSIBILITIES ▪ Cloud provider ▪ Ops / hoster ▪ Dev / Client PROPERTY Cloud platforms may belong to many client teams. Platform governance is a thing. Hypervisor VM OS Libs conf. Kernel HDW Middleware conf. Apps conf. Server Storage Network Logs / Metrology / Backups Data Runtime conf. Container conf.
  • 20. ❏ Requires different skills ❏ Cloud vendors try to attract directly Dev people: NoOps myth, shadow IT… ❏ Assumes consumers to be autonomous ❏ Here come DevOps ;) From cases to APIs
  • 21. Ride the raging bull Cloud services are often proprietary (no source code available). You have to master ❏ how they behave ❏ how to deal with their limitations Often with try / fail / learn methodology.
  • 22. Patch management Cloud providers offer up-to-date assets. But do not upgrade already provisionned ones! You have to use immutable infrastructure management patterns
  • 23. Integration Cloud providers offer many services. Building a platform means to integrate 10s of services with each others. Cloud architect & devops guys are deeply needed!
  • 24. Price-driven governance Usage changes, explosion of your VMs counter 60% of time spent in operating Cloud services is to optimize costs – Google DevFinOps “Tristar Medical Group Discovers Solution That Reduced its AWS Costs by 60% (~$20k per month)” – May 9, 2018
  • 26. Data gravitation / Reversibility Cloud providers aim to store data. ❏ Weight of data ❏ Data format ❏ Data requesting technos … lead to make reversibility difficult.
  • 27. High-end value services / Reversibility Cloud providers deliver high-end services that are ❏ quite easy to use ❏ very difficult to migrate from. For example ❏ cognitive services ❏ authentication services ❏ Spanner ❏ Alexa chatbot
  • 29. Focus on value & better TTM but support platform constraints More flexibility, more tuning capacity… but assume plumbing Functions ABSTRACT FOCUSINFRA Microservices Monoliths CaaS IaaS FaaS & PaaS
  • 30. Don’t put all one's eggs in one basket Create an abstraction layer between your consumers and cloud providers: let’s hybridate! ❏ Leverage on open-source / standards as much as you can. Notice how much object storage vendors built up S3 (AWS) compliant APIs? ❏ Make an impact analysis (reversibility, etc.) before making your apps sticky with the subsequent platform it relies on. (Mind the price upgrade of Google Maps) ❏ There’s no free lunch: possibly assume not to use everything you can? Platform AWS AZURE
  • 33. People & processes over technology ❏ Infra-as-code, software-defined-xxx: Ops progressively become Devs: you need to engage on reskilling them! ❏ No longer be a robot, invent robots: it’s not just about hard skills! ❏ While some activities retire, new emerge: low level architecture, IT financing metrology for instance ❏ Integrate your cloud into a broader devops or continuous delivery transformation People > Processes > Tools
  • 35. Architecture & cloud vendors ❏ App architecture & languages influenced cloud offers: microservices, app resiliency & scalability contributed to infrastructure commoditization ❏ Nowadays, the cloud influences you app: object storage, serverless, cloud native apps We created junkies, always searching for more abstraction. But your dealer is not your friend, he as some interests in ($$$)
  • 36. Subtle equilibrists choices ❏ Offload from the non valuable plumbing ❏ … but some infrastructure assets may be finally part of your differentiators ❏ Get the control back may be valuable: preserves autonomy, reduce operational costs at scale... ❏ “Do as I say, not as I do”: refocusing on core-business is the exact opposite of what cloud vendors does, being themselves fullstack ❏ Be an engineer, not a fashionista
  • 37. (Re) Assess yourself Time passed, the landscape changed ❏ Technologies are more mature ❏ Consulting/training/documentation... are more developed May be the time to get the control back ?