Data Mesh Part 4 Monolith to Mesh

Data Mesh Part 4:
Data Monolith to Data Mesh
Future of Data Integration Tools and a
focus on Oracle GoldenGate and Stream Processing
Oracle Development
October 2020
Copyright © 2020, Oracle and/or its affiliates1
Channel: https://www.youtube.com/user/oraclegoldengate
Data Mesh Playlist:
https://www.youtube.com/playlist?list=PLbqmhpwYrlZJ-583p3KQGDAd6038i1ywe

for more than 30yrs, a “Hub
and Spoke” Architecture:
Is “Hub and Spoke” our destiny
forever…or could we be on a
journey to somewhere else?
ETL Tools…
Kimball EDWs…
Data Lakes…
Data
Hub Vendor DI Tools…
ETL
Hub
ODS
Hub
Big Data
Hub

The world around us will keep moving faster and faster…
…IT systems and the data that fuel them are going
to need to become faster and more agile…the
people processes that we follow for DevOps and
DataOps must also be faster and more agile
Our old ways of Data Integration are no longer
sufficient to meet the future.

Data Fabric | Stream Processing | Data Mesh

The future of Data Integration…is Mesh
…a new generation of Data Mesh capabilities will
leave behind the Monolithic Tools of the past to
interconnect modern, multi-cloud, data-driven
applications and create innovative, high-value
data products of all types

Data Mesh Series
6
https://www.youtube.com/playlist?list=PL
bqmhpwYrlZJ-583p3KQGDAd6038i1ywe
Part 1:
CDC and Distributed
Commit Logs
Best Practices for
Maintaining Transaction
Consistency with
Replication and Kafka
Managing Table to Topic
Mappings, Accounting for
Schema Evolution etc.
How to Handle the
Change Stream: Partial
Supplemental CDC,
Caching, Lookups etc.
Deployment Topologies:
Mid-tier, End-Point, Topic
Partitions etc.
Part 2:
Microservices Data
Architecture w/CDC
Microservices Design
Patterns for the Data Tier
Understanding the
GoldenGate Microservices
Architecture
Event Patterns for CDC:
• Transaction Outbox
• CQRS with CDC
• Event Sourcing with CDC
• Saga with CDC
Event Driven Processing,
with CEP, ESP Time Series,
& GoldenGate Stream
Analytics
Part 3:
Demo of Application &
Data Integration Mesh
What’s in a Name? Data
Fabric, Data Hub, Data
Mesh, Service Mesh
Purpose of a Data
Architecture: Operational
vs. Analytic Use Cases
Demonstration Video: Retail / Inventory Analysis
• Sources: Weather.com, Oracle DB, Retail Cloud, AWS S3, Salesforce
• Targets: Data Lake (Object Storage and Autonomous Data
Warehouse), Data Services (Mobile APIs), Stream Analytics
Part 4:
Monolith to Mesh, the
Future of DI Tools
Brief History of
(Monolithic) Integration
Tools
Future of Data Integration
is Mesh
DevOps and Deployment
of the Data Mesh
Business Value of a Data
Mesh (vs. the Monoliths)
Copyright © 2020, Oracle and/or its affiliates

Agenda
1
2
3
4
Brief History of Integration Tools
Data Mesh as a Next Step
GoldenGate Strategy for Data Mesh
Call to Action

Messaging &
Event Systems
Brief History of Enterprise-class Integration Tools
Biz Process
APIs
Data
Consistency
App
Integration
Data
Integration
Transaction
Processing Systems (TPS)
Enterprise Application
Integration (EAI)
Service-Oriented
Architecture (SOA)
Integration Platform
as a Service (iPaaS)
B2B
Business Process Management (BPM)
Enterprise Service Bus (ESB)
Robotic Process
Automation (RPA)
Message Queue (MQ)
Messaging: Kafka, Pulsar etc.
Extract Transform Load (ETL) Data Integration (DI)
Change Data Capture (CDC)
& Data Replication
Data Federation/Virtualization
Complex Event Processing (CEP)
Big Data Event Stream Processing (ESP)
Stream Integration
1980 1990 2000 2010 2020Historically, integration
tools have focused on
specific tiers of the
software architecture:
Focus on committed, reliable and
ACID-grade data, typically for both
Operational (OLTP) and Analytic
(OLAP) workloads…
Data
Governance
Catalog etc.
Data Quality (DQ)
Master Data (MDM)
Data Hubs etc.
8

Meanwhile… Apps shift from Monoliths → Microservices → Mesh
shared frameworks
shared frameworks
host
App /
Component
App /
Component
App /
Component
host
shared frameworksframeworks
frameworks
App /
Component
App /
Component
App /
Component
App /
Component
App /
Component
App /
Component
host eg; serverless eg; container
Mesh
Controller
Distributed commit log
Classic Monolith “Minilith” / Client Server Microservice & Mesh
Very coarse grained, many components within
single App boundary, layers of shared
frameworks, dependencies on one/more
schema, single host is often mandatory
Coarse grained, some components may be
independently upgraded, but cross-component
dependencies still generally tightly-coupled.
Single host is preferred. Dependencies between
Apps and shared schema still exist.
Component isolation. Encapsulated App
schema, or Event Sourcing instead. All comms
via public APIs. Components may often be
stateless, to run in IT-managed service mesh
(eg; K8S) or in public cloud based serverless
runtimes (fully managed)
shared frameworks
sidecar
Note: a monolith with REST APIs is still a monolith, and putting a monolith in a container/K8S doesn’t make it a microservice!
Monoliths Mesh
9

Inflexion Point for DI Tools to Modernize…
Monolithic Data Hub
Classic Monolith
Data Hub
Client-server / Minilith
Data Mesh
Serverless / Event-driven Microservices
Batch ETL:
• Ab Initio, DataStage Grid
• PowerCenter (pre-10.x)
• Hadoop / Hive (CDH, HDP etc)
Streaming Data & Realtime Events:
• IBM Streams, Software AG Streams
• Lambda Big Data Architecture
(eg: Apache Hadoop + Apache Storm)
Batch ETL/ ELT and Cloud Native:
• PowerCenter (10.x and higher)
• Talend/Stitch, ODI, SAP, SAS etc
• Databricks “Data Lakehouse”
• Qlik/Attunity, IBM IIDR, etc.
• Kappa Big Data Architecture
(including: Confluent KSQL and Flink)
• GoldenGate, GoldenGate Stream Analytics
• AWS Kinesis, Lambda, Glue Streaming
• Event Sourcing Pattern with Domain
Aggregates (bespoke microservices)
Batch ET/ ELT and Cloud Native:
• OCI Data Integration, OCI Data Flow
• AWS Glue, Azure Data Factory
Compute + Storage
Compute
Storage
data data
data
Physical Site / Network Physical Site / Network
Monoliths Mesh
Hub Hub
10

Graph of Hubs != Data Mesh
Monoliths Distributed
Note: a monolith with REST APIs is still a monolith, and putting a monolith in a container/K8S doesn’t make it a microservice!
Hub
Hub
Hub
Compute
Storage
Compute
Shared Storage
Event Logs
Event Logs
VCN 1 VCN 2 VCN 3
data
Service Mesh Serverless
Runtime
Managed
Containers
ingest
prepare
pipe A
pipe B analyze
cleanse
sink
11

Agenda
1
2
3
4
Call to Action

What is a Data Mesh?
13
Microservice
Patterns
Log-based
Integrations
Polyglot
Replication
Data Mesh is a data-tier architecture to integrate and
govern enterprise data assets across distributed multi-cloud
environments – three defining characteristics are:
Data Product Oriented:
• Low code management of high-value data services that support
operational data stores, analytics, data lakes and data science
De-Centralized Processing:
• De-centralized data processing; no ETL/Hubs/Lake monoliths
• Microservices / Service Mesh and Serverless deployments,
utilization of “sidecar proxy” patterns, encapsulation, etc.
• Simplified continuous integration continuous delivery (CICD) and
lifecycle management (LCM) across public/private clouds
Event-Driven, Stream Centric:
• Real-time by default, batch patterns only when necessary
• Immutable event logs for messaging and data store events
• Trusted data semantics for consistent (ACID) and polyglot data
https://en.wikipedia.org/wiki/ACID
Data
Mesh
Event
Streaming
Immutable
Logs
Data
Replication
Polyglot
Persistence
Edge / 5G
Frameworks
Domain
Driven
Design
Service Mesh
“Sidecars”
Data
Mesh
Data Product Oriented
Eg: distributed
commit log in
Kafka
Eg: Kubernetes
controller +
kubelet
Eg: data
consistency
guarantees
Eg: low code,
data domain
centric

Data Monolith
Data as an IT artifact Data as a Product
Monolithic & Centralized Distributed & Decentralized
Waterfall Data/DevOps (dominant) Agile, CICD Data/DevOps
Batch Processing Centric Event-Driven Streaming by Default
OLTP vs. OLAP OLTP ∩ OLAP
Data Meshto…

Enterprise Data
Producers:
ERP Apps, DBs,
Middleware etc.
IoT Data
Producers:
Devices &
Things
Raw Data
Prepared Data
Canonical Data
Data
Consumers
Data
Domain A
Data
Domain C
Raw Event Consumers
Automated Devices,
Edge Nodes (5G), Machine-to-
Machine (M2M)
Data
Domain B
Business
Data Product
Owners
APIs
M2M
Marts
Models
Analytics
Data Producers Data Mesh
Data Mesh Purpose is for Data Products

Raw data, Time Series & Alerting events are pushed
Direct to Database (high fidelity transaction semantics fully preserved)
Optimized for Data Product Owners
Enterprise Data
Producers
Detect
Event
Logical
Change
Records
(LCRs)
App
DB
committed!
CDC Replication
Data
Objects
Table
Data
Raw Data
/ Alerts
SQL
Consumers
Raw
Data
Prepared
Data
Canonical
Data
Raw Data (LCR)
Schema Events
(DDL)
Prepared
Data Topics
“Master”
Data Topics
JSON, XML,
Avro, Parquet,
CSV
Prepared data events are pushed
Canonical data events
Speed &
Fidelity
Trusted
Views
Ease of
Consumption
LCR/TFs
Applications,
Data Services
Biz Consumers
Analytics &
Data Marts
Data Science
& Streaming
Applications
DBAs for HA,
DR and OLTP
Data Product Owners /
Managers are responsible
to translate IT deliverables
into trusted data that
delivers business value
Data Model
Object Model
System
Of Record
(SoR)
User
Action
App APIs and
system log events
Data Product
Owners

Direct to Database (high fidelity transaction semantics fully preserved)
Decentralized by Design
Data Domain
Producers
Detect
Event
Logical
Change
Records
(LCRs)
App
DB
committed!
Data
Objects
Table
Data
Raw Data
/ Alerts
SQL
Consumers
Data Model
Object Model
System
Of Record
(SoR)
User
Action
CDC Replication
Microservices
Edge Compute
or Cloud for
Raw Data
Events
Prepare
Technical Data
Views
LCRs
Business
Data Views
Raw data, Time Series & Alerting events are pushed
Prepared data events are pushed
Canonical data
Events
(ephemeral or persisted)
Stream
Process
Events
(persisted)
Stream
Process
Events
(persisted)
Applications,
Data Services
Biz Consumers
Analytics &
Data Marts
Data Science
& Streaming
Applications
OLTP and
OLAP DBs
Data Products
Payloads

DevOps Attributes for a Data Integration Mesh
18
Physical
Deployments
• Runtime should be deployable in most infrastructure
• Also, mesh nodes may also be tightly-coupled to a single infrastructure
(eg; a serverless environment in a proprietary public cloud)
Mesh Controls • “Data mesh controller” should have Observability, Security and Routing controls
on node deployments in various infrastructures (of customer choice)
Data Latency • Event-driven, streaming pipelines by default (single digit seconds)
• Pipelines may execute as micro-batch or batches (eg; fixed windows)
• Large batch processing “by reference” for files or direct-path DB utilities
(not all data transfers are suitable for event protocols)
Enterprise Data
Semantics
• Polyglot data handling means both highly structured and semi-structured
• Must preserve ACID / full relational semantics into Targets
• Must handle non-relational document payloads
Data Governance • Must have comparable data governance features with mainstream data
integration tools (eg; Catalog, Lineage, Data Validation, etc)
Simple DevOps
/ CICD Lifecycle
Customer Managed:
• Runtime is microservices-based
(note: just a REST API is not sufficient)
• Runs in containers (eg; Docker), optional
service mesh: Kubernetes, OpenShift etc.
Cloud Vendor Managed:
• serverless execution

Example 1: Mesh of Data Integration Microservices
Edge Gateways
Edge
Multi-Cloud
Enterprise Applications
Analytics
single pane
of glass…
filter
capture
λ
dist.
ingest
xform
load
dist.
ingest
ingest
capture
dist.
capture
dist.
capture
capture
replicat
join
load
capture
dist.
capture
Exadata
Cloud@Customer
19

Data Mesh Workload Coexistence
Compute
Storage
Compute
Shared Storage
Event Logs
Event Logs
VCN 1 VCN 2 VCN 3
data
Service Mesh Serverless
Runtime
Managed
Containers
ingest
prepare
pipe A
pipe B analyze
cleanse
sink
Mixed Workloads: that can run in
different infrastructure & “engines”
Customer Managed:
runs “as a Service”
using containers and
service mesh
Vendor Managed:
runs as “Serverless” customer pays only for
the minutes that workloads are running
Data
Products
Hybrid Infrastructure:
runs within managed
cloud containers
20

Example 2: Maintain Data Consistency in Pipelines
SCN – System Change Number, is the Oracle DB clock – every time a transaction commits, the clock
increments. The SCN marks a consistent point in time in the database.
CSN – Commit Sequence Number, is the GoldenGate clock – GG uses CSN during apply to identify
the point in time at which the transaction is committed for maintaining transaction consistency and
data integrity. A CSN is available for all Source DB transactions captured via GoldenGate:
https://docs.oracle.com/en/middleware/goldengate/core/19.1/admin/commit-sequence-number.html
Kafka
Single Partition
A
A { “customer_id": “1" ,
“first_name": “Debra" ,
“last_name": “Burks" ,
“phone": “" , “email":
“debra.burks@yahoo.com" ,
“SCN”: “130” , “CSN” : “130”
}
B
B
{ “customer_id": “1" , “9273
Thome Ave." , “city":
“Orchard Park" , “state":
“NY" , “zip_code": “14127“ ,
“SCN”: “130” , “CSN” : “130”
}
Data
Consumer is
responsible to
maintain
transaction
boundaries
OLTP
Updates and
Deletes both show
up in Kafka as new
messages,
Consumers must
interpret the flags
correctly
21

Fast data architecture for customer
satisfaction (user activity stream,
trusted transactions)
Solution had to be fast, but most importantly the data had to be correct!
https://www.slideshare.net/r39132/big-data-fast-data-paypal-yow-2018

Data
Objects
Table
Data
Raw Data
/ Alerts
SQL
Consumers
Applications,
Data Services
Biz Consumers
Analytics &
Data Marts
Data Science
& Streaming
Applications
DBAs for HA,
DR and OLTP
Data Owners &
Data Products
Example 3: Continuous Transformation and Loading (CTL)
Enterprise Data
Producers:
ERP Apps, DBs,
Middleware etc.
IoT Data
Producers:
Devices &
Things
Queries
Data Patterns
Windowing
Data Policies
Business Rules
Filter
Aggregate
Correlate/Enrich
Thresholds
Joins
Time Series
Spatial Analytics
Anomalies
Classification
Scoring Models
23

Top Opportunities for a Data Mesh Solution
IT / tactical solutions:
1. 100% correct, Cloud-native Apps
• Sync of backend OLTP stores
• Multi-active, cross-region DBs
2. Distributed Data Lake
• Fast data from anywhere
3. Shift from Batch to Streaming DI
• ETL to Continuous-TL
4. More Agile, DataOps Lifecycle
• Microservices DevOps benefits
Business Transformation / strategic:
1. Next Best Action
• Realtime customer engagement
2. Smart Inventory Management
• Eliminate supply chain lag
3. Location Intelligence
• Correlate app + device events
4. Predictive Analytics
• Data monetization,
new services for sale

Agenda
1
2
3
4
Call to Action

Why GoldenGate for a Data
Mesh, what’s so special?
• Event detection for all popular
data stores, relational and NoSQL
• High speed data replication
• Trusted to never lose data –
availability is a core use case
• Transaction-safe (for dependable
analytics and applications)
• Event stream processing at Web-
scale, on open-source or Cloud
Data Products
26

GoldenGate Microservices
27
data replications
bi-directional
ms/sec updates
consistency guarantees
Cloud, Containers or Edge Devices
Extracts Replicats Client Libraries
Native GUIs and Full REST APIs
API Gateway or Proxy Service
GG GUI
GoldenGate is itself a set of microservices
that human users or other services may
interact with
Embedded User Interface:
• C-based microservices with embedded HTTP client for
native JavaScript based GUI
• Oracle Jet frameworks for intuitive interaction model
REST native APIs:
• Fully REST native
• Also available, a Command Line Interface (CLI) produces
REST calls to the native services
Full GoldenGate Replication Capabilities:
• 100% coverage for all traditional GG replication
patterns; fully capable of HA/DR use cases
GG
Admin
Service
GG
Distro
Service
GG
Metrics
Service
GG
Receiver
Service
GG
Service
Manager
Your
Services

GoldenGate Stream Analytics
Trusted, Transaction Outbox for the Whole Enterprise
DB2/z
Trusted
Replication of
Real-time Data
Transactions & Events
ETL
&ML
Object
Storage
Relational
Non-
Relational
Apps
https://www.oracle.com/middleware/technologies/goldengate.html
DBMS
Cloud
Big Data
NoSQL
Streams
28

Single Pane of Glass
connect
DB2/z
Data
Objects
Table
Data
Raw Data
/ Alerts
SQL
Consumers
Applications,
Data Services
Biz Consumers
Analytics &
Data Marts
Data Science
& Streaming
Applications
DBAs for HA,
DR and OLTP
Real-Time Stream
Data Processing
Raw
Data
Data Owners &
Data Products
Deploys in a Mesh
Across Containers, Public Clouds and 5G Edge Devices

Microservices, Distributed Deployments
GoldenGate: Mesh Platform for Fast Data
Single Pane of Glass for Key Personas
OperationsContinuous
Transformation &
Loading (CTL)
Change Data
Capture (CDC)
Governance
Stream Analytics
(ML, Time-Series etc)
Data Replication
(source/target)
Oracle Cloud | 3rd Party Cloud | On-Prem Data Centers | Embedded Edge
Data Engineer Data AnalystData OpsCloud Admin
Sample Apps, Pre-Built Templates, Accelerators
Optional Containers, Service Mesh (K8S Pods) or Serverless
• Workspaces
• Catalog
• Data Verification
• Security
• Management
• Monitoring
• Metering (cloud)
• Administration
DataProducers
(Apps,DBs,IoT,etc)
Data Owners &
Data Products
Data
Objects
Table
Data
Raw Data
/ Alerts
SQL
Consumers
Applications,
Data Services
Biz Consumers
Analytics &
Data Marts
Data Science
& Streaming
Applications
DBAs for HA,
DR and OLTP

Data
Consumers
Data Product
Owners
Empower the Data Product Owners
The Age of Data Product Managers
https://medium.com/swlh/the-age-of-data-product-
managers-how-to-prepare-24c0fedc163f
More:
• https://airfocus.com/glossary/what-is-a-
data-product-manager/
• https://hbr.org/2018/10/how-to-build-
great-data-products
Data products include:
1. Analytics
• Reports and dashboards
• Historic and real-time
2. Models
• Data domain objects
• Data models / ML features
3. Algorithms (for Business Rules)
• ML models, AI/data science
• Pipeline policies and mappings
4. Data Services and APIs
• Data payloads
• REST APIs, Pub/Sub Topics etc.
Business

Compelling Data Products
Low Code Event-based Data Services Time Series Analytics Event Driven Dashboards
Streaming Data Patterns Geo-Spatial Analysis & Geo-Fencing Predictive Analytics / ML
32

Strong Governance
Workspace Management
• Low-code User Experience
• Role-based Access Controls
Data Catalog
• Asset Tagging, REST APIs
• Lineage Viewer for Pipelines
Security
• Certificates, Encryption
• Key Stores, Single Sign-On
• LDAP Integrations
• SSL, TLS 1.2/1.3
Conflict Detection
• Automatic CDR or User Defined
Data Verification
• Hash-based Digital Compare Tool
• Fast, 100% Certainty

Event-driven Data Mesh
Use Cases for Event Driven, Stream Data Processing
driven by business
demands for more flexible,
easier-to-change, loosely-
coupled applications […]
there has been a
widespread awakening to
the benefits of Event
Driven Architecture (EDA)
for increasing the
scalability and agility of
business systems.
W. Roy Schulte (of Gartner), March 2020:
EDA is Suddenly Popular Will Stream Analytics be Next?
Data & Microservice Events
Event/Data
Pipelines
Geospatial
Actions
Time-Series
Analysis
Real-time
AI/ML
Continuous
ETL
34

Agenda
1
2
3
4
Call to Action

More “Hub and Spoke”?
Which journey are you on?
Data
Hub
…or, are you ready to
move to the Mesh?

What Next?
Ask Oracle for a demo!
Oracle #1 in Data Fabric Strategy
GoldenGate YouTube | Data Mesh:
Free Trial of GoldenGate Streaming:
https://www.youtube.com/playlist?list=PL
bqmhpwYrlZJ-583p3KQGDAd6038i1ywe
https://cloudmarketplace.oracle.com/marke
tplace/en_US/listing/70961838
https://blogs.oracle.com/dataintegration/oracl
e_forresterwave_datafabric_2020?xd_co_f=66
bcf41f-e285-4ccc-a5b5-1c790cab0db0
Customer Success
37

Data Mesh Part 4 Monolith to Mesh

Our mission is to help people see
data in new ways, discover insights,
unlock endless possibilities.

Data Mesh Part 4 Monolith to Mesh

More Related Content

What's hot (20)

Similar to Data Mesh Part 4 Monolith to Mesh (20)

More from Jeffrey T. Pollock (20)

Recently uploaded (20)

Data Mesh Part 4 Monolith to Mesh