Cloud-Native Data
Extending Cloud-Native Beyond the App
Cornelia Davis
Sr. Director of Technology
Pivotal
@cdavisafc
@cdavisafc
Me?
Developer (wasn’t Ops)
Web architectures for >10 years
Cloud-native for 5 years
Cloud Foundry for 5 years
More recently
Discount code 40% off!: 40cloudnat
A Seattle book store
deploys code, on average,
every second
© 2015 Pivotal Software, Inc. All rights reserved. 4
Search µservice .
Image µservice .
Item Master µservice
Reviews µservice
Shopping
Cart
µservice
Other
dependen
t µservice
Other
dependen
t µservice
Other
dependen
t µservice
Why?
• Scale Applications
• Scale Teams
• Independent Development Cycles
• Experimentation
• Resilience
6
Continuous Delivery
of
Customer & Business
Value
Obstacles
• Silos: Dev, QA, Operations is
typical. No shared common goal
• Dissimilar Environments - “It
works on my machine”
• Risky Deployments: Manual
steps, done “off hours”
• Changes are treated as an
exception →Firefighting
• Processes designed around
these obstacles
Enabling Patterns
• Reinventing the Software
(Delivery) Value Chain
• Cloud-native Software
Architectures
• The Right Platform
• Devops
• Change is the Rule
(not the Exception)
© 2015 Pivotal Software, Inc. All rights reserved. 9
Search µservice .
Image µservice .
Item Master µservice
Reviews µservice
Shopping
Cart
µservice
Other
dependen
t µservice
Other
dependen
t µservice
Other
dependen
t µservice
10
Presentation
Frontend-integration
Order
Status
µService
Search
µService
Item

Master
µService
Item
MetaDatµS
ervice
Cart
µService
Image
µService
R12Ns
µService
Shipping
µService
µService
µService
µService
µService
µService
New LIGHTWEIGHT ARCHITECTURES are emerging
 Microservices addressing speed to market and cloud scale
Monolithic / Layered Microservices
12 http://highscalability.com/blog/2014/4/8/microservices-not-a-free-lunch.html
Spring Cloud Services 1.0.0
3
Spring Cloud Services
Config Server Service Registry Circuit Breaker
Dashboard
14
Operational Visibility: Distributed Tracing
• Latency visibility into a request’s end-to-end call graph
• Quickly identify a problematic service in a distributed system
• Zipkin is a open source distributed tracing system. It helps gather timing data
needed to troubleshoot latency problems in microservice architectures.
• Pivotal is investing in Zipkin to solve distributed tracing use cases
– Apache 2.0 License
– Created by Twitter in 2012.
– In 2015, OpenZipkin became the primary fork
Zipkin Tracing
• PCF Developers can redirect application traffic to a desired request
path in order to use logging, authentication or rate limiting systems
that exist outside of PCF
• PCF’s Service API will introduce a new field: route_service_url
• Developers will create a routing service instance and bind it to a
route (not an app)
– Service Instance can be created by a Service Broker or can
be a user-provided service instance
• Router is configured with and forwards requests to the URL
contained in the route_service_url field
• The route service is expected to forward the request back to the
route
• Knowing the request has already been forwarded to the route
service, the Router forwards to the associated applications
Route Services
client
load
balancer
CF router
CF app
route
service
1
2
3
4
5
6
New LIGHTWEIGHT ARCHITECTURES are emerging
 Microservices addressing speed to market and cloud scale
Monolithic / Layered Microservices
18
What is Cloud-native
Data?
It’s not this
At the Intersection of Cloud-native App Architecture & Data
What is Cloud-native Data?
Patterns
• Data-services Topology
• Interfacing to existing data sources
• Caching, but rethought
• ETL, but rethought
• Micro-databases
• Event Sourcing (and CQRS)
• Personas - what happens to the DBA, Data Architect, etc.?
Data-services topology
Goals/Needs:
• Resilience
• Tenancy
• Security - networks?
• Day 2 Ops - Backup/Recovery
Data-services topology: Resilience
PCF
Data Center 1
PCF
Data Center 2
PCF PCF
Pipelines Distribute Deployments
Active/Active
Data Center 1 Data Center 2
PCF PCF
Pipelines Distribute Deployments
“Active/Active”
Data Center 1 Data Center 2
PCF PCF
Pipelines Distribute Deployments
“Active/Active”
Data Center 1 Data Center 2
PCF PCF
Pipelines Distribute Deployments
“Active/Active”
Data Center 1 Data Center 2
WAN Replication
Interfacing to Existing Data Sources
Goals/Needs:
• Cost reduction - offloading MIPS
• Agility
Pattern: Data API
• Microservices do not access data layer directly
• Except for the micro services that implement the data
API
• A surface area to:
• Implement access control
• (Instead of the likes of firewall rules)
• Implement throttling
• (Fair sharing of a resource)
• Perform logging
• Other policies…
Anti-pattern: Stateless Data APIs*
29
* We will maintain statelessness
at the app level
This is the architecture that dominated the
SOA era of the early 2000s
Culture tip: Data APIs needn't be
built by the database team
Pattern: Microservice Needs a Cache
30
We’ll have a lot more to discuss with respect to caching
… stay tuned
Pattern: Data API
• Microservices do not access data layer directly
• Except for the micro services that implement the data
API
• A surface area to:
• Implement access control
• (Instead of the likes of firewall rules)
• Implement throttling
• (Fair sharing of a resource)
• Perform logging
• Other policies…
Pattern: Data API
• Microservices do not access data layer directly
• Except for the micro services that implement the data
API
• A surface area to:
• Implement access control
• (Instead of the likes of firewall rules)
• Implement throttling
• (Fair sharing of a resource)
• Perform logging
• Other policies…
Pattern: Versioned Data API
• We are already familiar with versioned
micro services…
V1 V2
Possibly coupled with
Pattern: Parallel Deployments
Caching Rethought
Goals/Needs:
• Performance
• Interface to existing data sources
• Resilience
https://content.pivotal.io/white-papers/in-memory-data-caching-for-microservices-architectures
Caching Patterns
Look Aside
• Attempt retrieval from cache
• Client retrieves from source
• Write into cache
! ?
"
#
Advantages
• If cache is unavailable, data source
may still be
• Cache configuration is very simple
Disadvantages
• Developer may be responsible for
protocol implementation (Spring
Cache Abstractions do hide this from
the dev)
Caching Patterns
Read-through
• Attempt retrieval from cache
• Cache retrieves from source and stores
in cache
• Return value to client
! ?
"
#
Advantages
• Simpler client programming model
(though developer may be responsible
for code running in cache)
• Less processing load on the client
Disadvantages
• Cache must available
• Cache configuration, including code
deployment into cache, is more complex
Caching Patterns
Write-through
• Write to cache
• Cache writes to source
• ack sent to client
!
"
#
Advantages
• Simpler client programming model
• Consistent
Disadvantages
• Cache must available
• Cache configuration, including code deployment, is
more complex
• Depends on connectivity to cache and cache to source
• Higher latency
Caching Patterns
Write-behind
• Write to cache
• ack sent to client
• Cache writes to source asynchronously
!
"
#
Advantages
• Simpler client programming model
• Very low latency
Disadvantages
• Cache must available
• Cache configuration, including code deployment, is
more complex
• Depends on connectivity to cache and cache to source
• Eventual consistency
© Copyright 2014 Pivotal. All rights reserved.
40
Why Cloud-Native Apps Need an In-Memory Cache
As an integrated service on a world class platform
Fast, available
microservices
Legacy app
modernization
Performance
at scale
Auto-scaling High Availability Logging/Metrics Security Zero Downtime Updates …Multi-cloud
Pattern: Cache Warming
• Loading the
cache can be
expensive
• Spring Cloud
Data Flow for
modern ETL
Sources
Destination
Spring Boot
Apps
Filter
Microservice
Enrich
Microservice
Score
Microservice
Spring Boot
Apps
Spring Boot
Apps
IoT
Microdatabases & Event Sourcing
Goals/Needs:
• Agility ∴ Autonomy
• Resilience
44
Pattern: Database per Microservice
• Supports Polyglot
persistence
• Independent availability,
backup/restore, access
patterns, etc.
• In PCF, on-demand,
dedicated clusters
My
Connections
API
Posts
API
New from
Network
API
Client Side Join
47
My
Connections
API
Posts
API
New from
Network
API
Event-driven Systems
48
My
Connections
API
Posts
API
New from
Network
API
Independent Databases - Shared Entities
• We’ve started to break up the
data monolith
• BUT our data integration
“strategy” is rather brittle and
bespoke
• How are changes to data in
one bounded context
reflected in the other?
Sales
Support
?
My
Connections
API
Posts
API
New from
Network
API
My
Connections
API
Posts
API
New from
Network
API
My
Connections
API
Posts
API
New from
Network
API
Unified Log
CQRS is what allows
the unified log to be
the source of truth
54
My
Connections
API
Posts
API
New from
Network
API
Unified Log
But then, what about people?
Info Sec
Srv Build
Cap PlanNetwork
OpsMid. Eng.
SW Arch
SW Dev
Client SW Dev
Svc Govern
CUSTOMER FACING APP TEAM
Ops
Cap Plan
DCTM Eng
DCTM
Cap Plan
Ops
SW Arch
SW Dev
Client SW Dev
CUSTOMER FACING APP TEAM
Ops
Cap Plan
ENTERPRISE
ARCH
Ent Arch
Proj Mgmt
Biz An
Prod MgmtData Arch
DBA
Biz An
Prod MgmtData Arch
SW Arch
SW Dev
Client SW Dev
LEGACY SERVICE TEAM
Ops
Cap Plan
Biz An
Prod MgmtData Arch
CSO INFRA
MID/
DEV
BIZ
ENT
APPS
DATA
Change Control
PLATFORM TEAM
Ent Arch
Prod Mgmt
57
Cloud-native Patterns
Summary
Legacy Data
Access
Service APIs
Data APIs
Shared DB
Database Per
Service
Data Integration
Client-side “Joins”
Event Sourcing
CQRSData Replication
Parallel
Deployments
Caching
Cache Provisioning
and Management
Look Aside
Read-through
Write-through/
behind
Warming
Transforming The Way the World Builds Software
59
And Software Needs Data
Thank you!
Cornelia Davis
Sr. Director of Technology
Pivotal
@cdavisafc
Discount code 40% off!: 40cloudnat
Cloud-native Data

Cloud-native Data

  • 1.
    Cloud-Native Data Extending Cloud-NativeBeyond the App Cornelia Davis Sr. Director of Technology Pivotal @cdavisafc
  • 2.
    @cdavisafc Me? Developer (wasn’t Ops) Webarchitectures for >10 years Cloud-native for 5 years Cloud Foundry for 5 years More recently Discount code 40% off!: 40cloudnat
  • 3.
    A Seattle bookstore deploys code, on average, every second
  • 4.
    © 2015 PivotalSoftware, Inc. All rights reserved. 4 Search µservice . Image µservice . Item Master µservice Reviews µservice Shopping Cart µservice Other dependen t µservice Other dependen t µservice Other dependen t µservice
  • 5.
    Why? • Scale Applications •Scale Teams • Independent Development Cycles • Experimentation • Resilience
  • 6.
  • 7.
    Obstacles • Silos: Dev,QA, Operations is typical. No shared common goal • Dissimilar Environments - “It works on my machine” • Risky Deployments: Manual steps, done “off hours” • Changes are treated as an exception →Firefighting • Processes designed around these obstacles
  • 8.
    Enabling Patterns • Reinventingthe Software (Delivery) Value Chain • Cloud-native Software Architectures • The Right Platform • Devops • Change is the Rule (not the Exception)
  • 9.
    © 2015 PivotalSoftware, Inc. All rights reserved. 9 Search µservice . Image µservice . Item Master µservice Reviews µservice Shopping Cart µservice Other dependen t µservice Other dependen t µservice Other dependen t µservice
  • 10.
  • 11.
    New LIGHTWEIGHT ARCHITECTURESare emerging
 Microservices addressing speed to market and cloud scale Monolithic / Layered Microservices
  • 12.
  • 13.
    Spring Cloud Services1.0.0 3 Spring Cloud Services Config Server Service Registry Circuit Breaker Dashboard
  • 14.
    14 Operational Visibility: DistributedTracing • Latency visibility into a request’s end-to-end call graph • Quickly identify a problematic service in a distributed system • Zipkin is a open source distributed tracing system. It helps gather timing data needed to troubleshoot latency problems in microservice architectures. • Pivotal is investing in Zipkin to solve distributed tracing use cases – Apache 2.0 License – Created by Twitter in 2012. – In 2015, OpenZipkin became the primary fork Zipkin Tracing
  • 15.
    • PCF Developerscan redirect application traffic to a desired request path in order to use logging, authentication or rate limiting systems that exist outside of PCF • PCF’s Service API will introduce a new field: route_service_url • Developers will create a routing service instance and bind it to a route (not an app) – Service Instance can be created by a Service Broker or can be a user-provided service instance • Router is configured with and forwards requests to the URL contained in the route_service_url field • The route service is expected to forward the request back to the route • Knowing the request has already been forwarded to the route service, the Router forwards to the associated applications Route Services client load balancer CF router CF app route service 1 2 3 4 5 6
  • 16.
    New LIGHTWEIGHT ARCHITECTURESare emerging
 Microservices addressing speed to market and cloud scale Monolithic / Layered Microservices
  • 18.
  • 19.
  • 20.
    At the Intersectionof Cloud-native App Architecture & Data What is Cloud-native Data? Patterns • Data-services Topology • Interfacing to existing data sources • Caching, but rethought • ETL, but rethought • Micro-databases • Event Sourcing (and CQRS) • Personas - what happens to the DBA, Data Architect, etc.?
  • 21.
    Data-services topology Goals/Needs: • Resilience •Tenancy • Security - networks? • Day 2 Ops - Backup/Recovery
  • 22.
  • 23.
    PCF PCF Pipelines DistributeDeployments Active/Active Data Center 1 Data Center 2
  • 24.
    PCF PCF Pipelines DistributeDeployments “Active/Active” Data Center 1 Data Center 2
  • 25.
    PCF PCF Pipelines DistributeDeployments “Active/Active” Data Center 1 Data Center 2
  • 26.
    PCF PCF Pipelines DistributeDeployments “Active/Active” Data Center 1 Data Center 2 WAN Replication
  • 27.
    Interfacing to ExistingData Sources Goals/Needs: • Cost reduction - offloading MIPS • Agility
  • 28.
    Pattern: Data API •Microservices do not access data layer directly • Except for the micro services that implement the data API • A surface area to: • Implement access control • (Instead of the likes of firewall rules) • Implement throttling • (Fair sharing of a resource) • Perform logging • Other policies…
  • 29.
    Anti-pattern: Stateless DataAPIs* 29 * We will maintain statelessness at the app level This is the architecture that dominated the SOA era of the early 2000s Culture tip: Data APIs needn't be built by the database team
  • 30.
    Pattern: Microservice Needsa Cache 30 We’ll have a lot more to discuss with respect to caching … stay tuned
  • 31.
    Pattern: Data API •Microservices do not access data layer directly • Except for the micro services that implement the data API • A surface area to: • Implement access control • (Instead of the likes of firewall rules) • Implement throttling • (Fair sharing of a resource) • Perform logging • Other policies…
  • 32.
    Pattern: Data API •Microservices do not access data layer directly • Except for the micro services that implement the data API • A surface area to: • Implement access control • (Instead of the likes of firewall rules) • Implement throttling • (Fair sharing of a resource) • Perform logging • Other policies…
  • 33.
    Pattern: Versioned DataAPI • We are already familiar with versioned micro services… V1 V2 Possibly coupled with Pattern: Parallel Deployments
  • 34.
    Caching Rethought Goals/Needs: • Performance •Interface to existing data sources • Resilience https://content.pivotal.io/white-papers/in-memory-data-caching-for-microservices-architectures
  • 35.
    Caching Patterns Look Aside •Attempt retrieval from cache • Client retrieves from source • Write into cache ! ? " # Advantages • If cache is unavailable, data source may still be • Cache configuration is very simple Disadvantages • Developer may be responsible for protocol implementation (Spring Cache Abstractions do hide this from the dev)
  • 36.
    Caching Patterns Read-through • Attemptretrieval from cache • Cache retrieves from source and stores in cache • Return value to client ! ? " # Advantages • Simpler client programming model (though developer may be responsible for code running in cache) • Less processing load on the client Disadvantages • Cache must available • Cache configuration, including code deployment into cache, is more complex
  • 37.
    Caching Patterns Write-through • Writeto cache • Cache writes to source • ack sent to client ! " # Advantages • Simpler client programming model • Consistent Disadvantages • Cache must available • Cache configuration, including code deployment, is more complex • Depends on connectivity to cache and cache to source • Higher latency
  • 38.
    Caching Patterns Write-behind • Writeto cache • ack sent to client • Cache writes to source asynchronously ! " # Advantages • Simpler client programming model • Very low latency Disadvantages • Cache must available • Cache configuration, including code deployment, is more complex • Depends on connectivity to cache and cache to source • Eventual consistency
  • 39.
    © Copyright 2014Pivotal. All rights reserved.
  • 40.
    40 Why Cloud-Native AppsNeed an In-Memory Cache As an integrated service on a world class platform Fast, available microservices Legacy app modernization Performance at scale Auto-scaling High Availability Logging/Metrics Security Zero Downtime Updates …Multi-cloud
  • 41.
    Pattern: Cache Warming •Loading the cache can be expensive • Spring Cloud Data Flow for modern ETL Sources Destination Spring Boot Apps Filter Microservice Enrich Microservice Score Microservice Spring Boot Apps Spring Boot Apps IoT
  • 42.
    Microdatabases & EventSourcing Goals/Needs: • Agility ∴ Autonomy • Resilience
  • 44.
  • 45.
    Pattern: Database perMicroservice • Supports Polyglot persistence • Independent availability, backup/restore, access patterns, etc. • In PCF, on-demand, dedicated clusters
  • 46.
  • 47.
  • 48.
  • 49.
    Independent Databases -Shared Entities • We’ve started to break up the data monolith • BUT our data integration “strategy” is rather brittle and bespoke • How are changes to data in one bounded context reflected in the other? Sales Support ?
  • 50.
  • 52.
  • 53.
  • 54.
    CQRS is whatallows the unified log to be the source of truth 54 My Connections API Posts API New from Network API Unified Log
  • 55.
    But then, whatabout people?
  • 56.
    Info Sec Srv Build CapPlanNetwork OpsMid. Eng. SW Arch SW Dev Client SW Dev Svc Govern CUSTOMER FACING APP TEAM Ops Cap Plan DCTM Eng DCTM Cap Plan Ops SW Arch SW Dev Client SW Dev CUSTOMER FACING APP TEAM Ops Cap Plan ENTERPRISE ARCH Ent Arch Proj Mgmt Biz An Prod MgmtData Arch DBA Biz An Prod MgmtData Arch SW Arch SW Dev Client SW Dev LEGACY SERVICE TEAM Ops Cap Plan Biz An Prod MgmtData Arch CSO INFRA MID/ DEV BIZ ENT APPS DATA Change Control PLATFORM TEAM Ent Arch Prod Mgmt
  • 57.
  • 58.
    Legacy Data Access Service APIs DataAPIs Shared DB Database Per Service Data Integration Client-side “Joins” Event Sourcing CQRSData Replication Parallel Deployments Caching Cache Provisioning and Management Look Aside Read-through Write-through/ behind Warming
  • 59.
    Transforming The Waythe World Builds Software 59 And Software Needs Data
  • 60.
    Thank you! Cornelia Davis Sr.Director of Technology Pivotal @cdavisafc Discount code 40% off!: 40cloudnat