mastering the microservices, fast data & hybrid cloud trifecta · apache kafka message queue...
TRANSCRIPT
© 2018 Mesosphere, Inc. All Rights Reserved.
Mastering the Microservices, Fast Data & Hybrid Cloud TrifectaEdward Hsu, VP Product2018.10.23
3
© 2018 Mesosphere, Inc. All Rights Reserved.
Operating ModelSourcing Model Architecture (Cloud-Native)
Three Definitions of Cloud
69%
OPTIMIZE
TRANSFORM
BOTH
Gartner: Digital Business Transformation: A Midsize Enterprise Perspective (July 2018, n=102)Gartner: Digital Business Transformation: A CIO Perspective (September 2018, n=372)
66% yet to deliver any success
14 out 15 fail focusing only on transformation
2X more likely to succeed by Optimizing first
Managing Disruption and Uncertainty
5
How CIOs are Tackling Digital Transformation
Most CIOs see risk to revenue due to new entrants
Overall 66% have yet to declare success
85%
59%
74%
Taking Control of Digital Transformation Economics
70%
30%
Run
Transform
Existing pressure to reduce costs
Gartner: Run, Grow and Transform the Business IT Spending: Approaches to Categorization and InterpretationGartner: 2017 CEO Survey: CIOs Must Scale Up Digital Business
MOST ENTERPRISES
6
30%
70%
Innovation Gap
New Digital Entrants
YOUR NEW COMPETITION
vs.
50%
50%
5-10% shift YoY
TYPICAL 5 YEAR GOAL
What Companies Need to Do
70%
30%
Run
Transform
Existing pressure to reduce costs
Gartner: Run, Grow and Transform the Business IT Spending: Approaches to Categorization and InterpretationGartner: 2017 CEO Survey: CIOs Must Scale Up Digital Business
MOST ENTERPRISES
7
Dramatically Reduce While Managing Risk
Find Force Multipliers to Maximize Impact
Source: 2017 IHS Markit estimates
TRADITIONAL APPLICATION
Users (Employees)
Monolith
Relational DB
DC or Cloud
MODERN DATA-DRIVEN APPLICATION
Containerized microservices(your code)
Data services & AI/ML tools(their code)
18 Billion connected users & devices
Always-connected customers & devices
Hybrid Cloud & Edge
Reality of Changing IT Landscape
9
TransportStore
Serve
Analyze / Streaming
Analyze
Analyze / ML
Apache Kafka message queue
Kubernetes container orchestrator
Store
Analyze
Transport
ServeApache Spark analytics engine
Apache Cassandra distributed DB
10
EXISTING APPLICATIONS NEW APPLICATIONS
+PRESENTATION
BUSINESS LOGIC
DATABASE
What this Means for Enterprise IT
Multi-Cloud is a Thing
Source: Mesosphere Cloud Native Ecosystem Survey, n~500
One out of four use more than one
cloud provider
15What Workloads Are They Running?General Classifications Named Workloads
Source: Mesosphere Cloud Native Ecosystem Survey, n~500; DC/OS Catalog Download Statistics July 2017 & 2018
Creating Automated Services on Hybrid Cloud and Edge
Mesosphere Approach:Optimize and Transform on a single platform
Data Analytics Cluster
Message Queue Cluster
Data Persistence
Cluster
Container Orchestration
Cluster
CI/CD Cluster
Traditional Approach: Slow, Expensive, Hard
Data Analytics
Message Queue
Data Persistence
Container Orchestration
Continuous Integration & Delivery
Distributed Cloud Operating System
1
2
3
● Manual & application-specific configurations
● Cluster sprawl and low utilization
● High risk with “snowflake” configurations in silos
Enterprise Applications
Enterprise Applications
18
Workload consolidation
Agile platform for digital innovation
Unified multi-cloud operations
20
© 2018 Mesosphere, Inc. All Rights Reserved.
Traditional Approach
Cluster: Dev Cluster: Prod
HW
OSVM
OSVM
HW
OSVM
OSVM
HW
OSVM
OSVM
Hypervisor Hypervisor Hypervisor
Mesosphere Approach
HW or VMOS
HW or VMOS
HW or VMOS
Cluster: DevA
Somebody else’s problem?
High-density pooling without virtualizationVirtualization required
Cluster: ProdB
● Virtualization & guest operating system cost & infrastructure overhead
● Secure resource isolation
Workload Consolidation - Kubernetes and Data Services1
Kafka Cluster:
Spark Cluster
Agile Platform for Digital InnovationAny Technology Become Automated Services
1. Prerequisites2. Installing the Client Tools3. Provisioning Compute Resources4. Provisioning the CA and Generating TLS Certificates5. Generating Kubernetes Configuration Files for Authentication6. Generating the Data Encryption Config and Key7. Bootstrapping the etcd Cluster… 3x for HA
10. Bootstrapping the Kubernetes Control Plane… 3x for HA13. Bootstrapping the Kubernetes Worker Nodes14. Configuring kubectl for Remote Access15. Provisioning Pod Network Routes16. Deploying the DNS Cluster Add-on… Deploying other Add-ons 20. Smoke Test21. Cleaning Up
Running on your own
$ dcos package install kubernetes
on
21
2
23
3
Stretched AMER / EMEA Cluster APJ Cluster
Unified Multi-Cloud Operations
● Easy migration of applications between clouds● Cross-cloud disaster recovery● Burst across clouds as needed
● Linked control plane for consistent operations across clusters
Unified Hybrid Cloud Operations:Live Migration of Workloads Between Clouds
24
3
“dcos-website” service live migration from AWS to Azure
Mastering the Microservices, Fast Data & Hybrid Cloud Trifecta
Don’t drive digital transformation projects in isolation
Consider Each Definition of Cloud Separately
32
Using a software layer for multi-cloud operations can be beneficial
Why Industry Leaders Choose Mesosphere
“I do not want to go down that path where I lock
myself into a provider. If I was ever told to migrate off
a cloud provider, it would only be two days worth of
work. And of course, DC/OS runs anywhere.”
Hybrid Cloud Portability
“With Kubernetes on DC/OS we are thrilled to offer our engineering teams their
preferred container orchestrator alongside big
data services with a common set of security,
maintenance, and management tools.”
Fast Data + Microservices
Save 50%+ on AWS Bill
Save on InfrastructureAccelerate Time to Value
Build and deploy time from
Weeks to
Minutes
Agility & Efficiency Strategic Platform
34