Opportunities, Promises and Challenges of the New IT Model for Enterprises

International Conference 2013
1
V1.0
Frank Greco
Director of Technology
Kaazing Corporation
frank.greco@kaazing.com
@frankgreco
Opportunities, Promises
and Challenges of the New IT
Model for Enterprises

2
Topics
l  Case Study – Large Financial Services Company
l  Risk and the IT Manager
l  The New IT Technology Stack
l  Cloud Computing and Mobile
l  Value of Social Computing
l  Impact of HTML5 on the Enterprise
l  The Power of WebSocket and Cloud Services
l  Analytics and Big Data
l  Changing Role for Enterprise IT Architects
l  Trends

3
Opportunities, Promises and Challenges
Business Opportunities
•  Business requirements to be successful
•  Revenue increase
•  Decrease in cost
•  Competitive advantages
Technology Promises
•  Potential benefits of a technology
•  Tools to make things faster, cheaper, better…
IT Challenges
•  What are the possible negative impacts for the business
•  What can go wrong

4
Computing Models
Mainframe
Minicomputer
Personal
Computer
Desktop
Internet
Technology Cycles
Every 10-15 years, there
is a technology shift

5
Today’s IT Focus
Agility is now Mandatory for Enterprises

6
Risk and the IT Manager
l  Combination of technologies can improve revenue,
enhance business processes and create new business
models
l  Need to plan for these technologies
l  Need to outsource common IT services to focus on
strategic functions
l  Managed risk is how enterprises advance
l  Without risk, there is no reward

7
Technology Adoption Curve
See the strategic opportunity
of new technology and are
true change agents
True opportunists.
Looking to achieve
short-term impacts with
new technology
Appreciate the benefits of
new technology but remain
risk adverse. Utilize vendors
with well-known quality
reputation.
Extensive business and use cases
often required to justify new
investment. Companies prefer
“business as usual” approach and
look to references within the same
industry to gauge technology
adoption
Highly risk adverse. Require much
re-education of executives,
architects and engineers. Falling
behind competitors.

WebSocket
HTML5
Mobile
Big DataCloud
External
APIs
NoSQL
Social
Media
New Technology Tools

NoSQL
External
APIs+
Cloud Big Data WebSocket Mobile+ + +
Cloud
External
APIs+
Cloud HTML5 Mobile+ +
Cloud
Big
Data
WebSocket Mobile+ + + External
APIs +
Planned Technology Collaboration

New IT Stack
UI/UX
APIs
Mobile/Embedded
Cloud
Services
New Hardware
Smart, Fast
Networks
public private/hybrid
IT Manager
SDN

11
Cloud Computing

12
Cloud Misunderstandings
l  “We don’t use cloud computing because I don’t trust Amazon.”
u  Cloud computing ≠ Amazon Web Services (AWS)
l  “We switched to cloud computing because I don’t want to worry
about security.”
u  Security is always a concern wherever you do computing
l  “Cost is the primary reason why we use a public cloud.”
u  A heavily-used public cloud may not be cost-effective as an on-
premises private cloud
l  “We don’t need cloud computing. We have our own datacenter.”
u  Internal datacenters can run a cloud infrastructure quite effectively

13
What is Cloud Computing?
US National Institute of Standards and Tech:
•  On-demand self-service
•  Broad network access for thin/native client
•  Elastic scalability (on demand)
•  Monitorability and Metering
•  SaaS, IaaS, PaaS (simplistic…)
•  Public/Private/Hybrid, On/Off Premises
A Deployment Model usually implemented
with an Services (SOA) approach

14
Cloud Deployment Types
Internal ExternalHybrid
public/privateon/off premises

Industry View of Cloud Stack
Hardware
Infrastructure as a Service
IaaS
Platform as a Service
PaaS
Software as a Service
SaaS

More Accurate View - Cloud Services Stack
Hardware
Compute/Network/Storage
Development, Delivery,
Management, Security,
Messaging, Integration,
Testing, …
Applications

17
The Criticality of PaaS
l  PaaS is where the action is
l  Services for developers to build applications
u  UI frameworks, OS, database, messaging, web server, etc.
u  Frameworks, APIs, solution stacks, …
l  Beware of lock-in
u  Avoid proprietary APIs or create framework facades
u  Don’t always rely on one vendor to provide 100% of your services
u  Ask for SLAs

18
Amazon Cloud - Growth
Number of websites hosted
Sept 2012: 6.8M
May 2013: 11.6M (+71% in 8 months)
Externally-facing web sites @ Amazon
Total count of AWS servers: ~450,000
AWS Annual Revenue: > 1,5B € ($2B USD)

19
Cloud Standards
l  Amazon/Eucalyptus
l  OpenStack – openstack.org
u  Rackspace, Dell, Citrix, Intel, AMD, Nicira, Cisco,
Mellanox, NetApp, Docomo, HP, NEC, Akamai, Suse,
et al…
l  CloudStack – cloudstack.org (Apache 2 license)
u  Enstratus, NetApp, Citrix, Intel, Alcatel-Lucent,
Gigaspaces, Tata, Arista, Sungard, et al...
l  VMware – proprietary API
Standards are still evolving…

20
Adopting Cloud Computing
l  Assess Current IT and how it Relates to Cloud
l  Avoid PaaS Lock-in
l  Have HA/DR Strategy
l  Need Interoperable Strategy for Federations
l  Implement POC, then Pilot - internal cloud project
u  You will learn a lot!
l  Plan cross-department Hybrid Cloud strategy for
apps, data… etc.

21
•  Agility for application/services creation and revenue
Technology Promises
•  Many vendors
•  Strong industry trend
•  Unbelievably quick access to computing power
IT Challenges
•  Migration: why should you use cloud computing, define
an architecture, select an initial candidate application,
have regular milestones, be vigilant and learn
•  Cloud standards and portability
•  May create silos
•  Need services architecture for cloud computing
•  Lack of enterprise service level agreements (SLA)

22
Mobile Computing

New Computing Model
Clouds
A Mobile App is easier to port to Desktop
A Desktop App is a challenge to port to Mobile

24
Mobile Computing for the Enterprise
“Mobile innovation is a major focus for the
bank in 2013 and beyond”
“Customers will be able to do almost as
much on their mobile phone as they can
with our full online banking service,
wherever they are”
“Enel smart info can help customers
optimise their energy consumption… with
easy access to information on a
smartphone”

25
•  Conduct business anywhere
•  Increase user and employee (BYOD) satisfaction
•  Opportunity to modernize client-facing applications
Technology Promises
•  High speed wireless connectivity
•  Faster, low-power mobile devices
•  Higher quality apps
IT Challenges
•  Security issues
•  Erratic connectivity for transactions
•  Many UI toolkits to choose
•  Variety of mobile devices
•  Lack of experienced developers

26
Social Computing

27
Value of Social Computing
“Social computing is not a fad, it will
impact on almost every role, at every
kind of company”
“Enterprise social software will be the
biggest technology success story of
this decade”
“More than a quarter of businesses
surveyed had formal budgets for
social and collaboration software
above $250,000 (190.000 €)”

28
What is Social Computing
Social Media
Locked-in Environments
Set of tools to foster interaction
1-to-Many
Content
Social Networks
Connections
Relationship Structure
Environment Agnostic
Many-to-Many
Computing environments meets Social behavior

Advantages and Disadvantages
Collaboration platforms
are becoming prevalent
within corporations
Customers and Employees
have high expectations based
on personal use. “App Effect”
Email volume is overwhelming.
Losing effectiveness and agility
Ubiquity of mobile devices
and social apps
Pressure to improve efficiency and innovate
Enterprises jumping in
without understanding
value to business
Security, privacy, financial
disclosure, copyright… laws can
be violated without guidelines
Opinions can be dangerous
without a disclaimer
To reap full benefit, organizations
must transform their structures,
processes, and cultures

30
•  Immediate feedback from users
•  Inspire employee innovations
Technology Promises
•  Collaborative tools go beyond email
•  Access to customer sentiment
•  True experts become visible
IT Challenges
•  Security and privacy issues
•  Legal issues
•  Need to change corporate culture to be effective

31
The New Web

Welcome HTML5 (aka The New Web)
l  Users are Demanding more from Apps
l  UI/UX Requirements are more sophisticated
l  Browser Enhancements and Evolution
l  API Explosion
l  Web no longer just about Documents… even
partial or pseudo documents
The New Web is a Programmatic Foundation for Rich,
Reliable and Really useful Apps

33
Impact of HTML5 on the Enterprise
l  Enhanced user experience
l  High-performance 2D/3D graphics
l  No plugins – higher security
l  Offline capability
l  Standardized browser storage
l  Standard audio and video features
l  Efficient network capabilities with WebSocket
l  Device API – file, geolocation, accelerometer,
microphones, cameras, etc.

Who Supports HTML5?
34
All the browser vendors do!
http://paulirish.com/2010/high-res-browser-icons

Forms-based Applications
35

HTML5 Applications
36

37
•  Deploy enterprise-quality applications over the web
•  Transact business without a network (offline)
Technology Promises
•  High-performance graphics
•  Standardized audio/video for all browsers
•  Improved web security
IT Challenges
•  Inconsistent implementation among browsers
•  JavaScript is the dominant programming language
•  Desktop vs. mobile screen sizes
•  Retraining costs

38
Services in the Cloud

Web APIs
l  APIs from everywhere
l  Over 9,000 public APIs and even more Mashups
u  programmableweb.com/apis/directory
u  Amazon, Facebook, LinkedIn, AT&T, Google, Microsoft,
NYTimes, Orange, SalesForce, Telefonica, Twitter, Visa,
Vodafone, Yandex
l  Enterprise and B2B APIs
l  Over time, more will be event-based – NoREST?
l  Services… Services… Services…

Open APIs
Most Popular
Google Maps, Twitter, YouTube, Flickr, Amazon eCommerce, Facebook, Twilio, …

Disadvantages of Request/Response
41
l  Lack of resiliency and robustness
l  Enterprise events retrieved by resource intensive
polling techniques
u  Much bandwidth is wasted
u  Information can be delayed
l  Composite services brittle and lack
transactionality
l  Large enterprises learned advantages of ESB
over 10 years ago
l  See failures of CORBA, Sun RPC, etc.

42
Enter HTML5 WebSocket!

But Why WebSocket?
It’s the Most Important API in HTML5!
•  Facilitates other protocols
•  Puts the web in a better place…
“Connectedness” – Always On… not partially on.
Network Efficient

44
Legacy HTTP vs. WebSocket
http:// ws://
For Event-Driven Applications

45
WebSocket
TCP
JMS XMPP AMQP B2B FTP VNC mktdata etc
Browser and Native Applications
WebSocket
Server
Internet
WebSocket
Server
Reac%va%ng
applica%on
protocols
over
the
web

Protocol Layering

Messaging over the Web
46
l  Web Messaging growing in popularity
l  Leverages the Internet for economies of scale
l  More robust and resilient than traditional REST
interfaces
l  History repeats itself
u  ESBs were superior to client-server
u  ISBs will be superior to REST

Event-based XaaS
Monitoring as a
Service
Integration as a
Service
Enterprise Messaging
as a Service
WAN Optimization as a
Service
Governance as a
Service
Database as a Service
Analytics as a Service
EAI as a Service
CDN as a Service
Trade Clearance as a
Service
Windows Desktop
as a Service
Sentiment Analysis
as a Service
Auditing as a
Service
Telephony as
a Service
Data Center as a
Service
Network as a Service
Risk Analytics
as a Service
Backup
as a Service
Security as a Service
Notification as a
Service
Identity as a Service
Testing as a
Service

48
New Computing Model
CloudCloud
Unified
Communications
CPU
Cloud
Enterprise
Storage
Notifications
Transactions
DevPay
Email Docs
Cloud
CRM
Collaboration
Cloud Cloud Sentiment
AnalysisMaps
websocket
websocket
websocket
websocket
websocket

49
•  Extend more apps over the firewall to new customers
•  Offer real-time APIs to supply chain
Technology Promises
•  Event-driven, publish/subscribe over the web
•  XaaS will offer all types of services
•  WebSocket is the new Socket
IT Challenges
•  WebSocket not implemented by older browsers
•  API management
•  Security of external/internal services

50
Big Data

51
What is Big Data?
l  Tools, Techniques and Technologies for managing
growing datasets for computation and collaboration
l  Large datasets - difficult to capture and manage with
traditional IT tools
u  Data has been always been increasing
u  Storage costs are relatively cheap
u  Analysis of the information is increasingly complex
l  Gigabytes (109) è Terabytes (1012) èPetabytes (1015)
l  The Predictive Enterprise – faster than real-time!
l  Dramatically improve competitive advantage
l  Data warehousing and traditional BI not adequate

Trend
52

53
Business Drivers for Big Data
Purchasing History
Customer Interactions
Web Logs
Dynamic Pricing
based on Page Views
Behavioral
Targeting
User Click Streams
RFID/Sensors for
Inventory Analysis
Sentiment Analysis
User Generated
Content
Geo-coordinates
for Selling
Global
Demographics
Business
Datafeeds
Video
streams
Regulatory Audits
Surveillance
Real-time Risk
Analysis
Identify High-value
Customers
Reputational Risk
Assessment

54
Enabling Technologies
l  Faster processors
l  Cheap storage
l  Cheap large memory capacities
l  Fast disk technologies (SSD)
l  High performance networking
l  Maturity of clustering and virtualization
l  Cloud computing – elastic resources

55
Vendor/Software Landscape

56
NoSQL
l  Many Big Data datastores are NoSQL
l  Non-Relational (rows and columns)
l  Store objects – Simple to program
l  Extremely scalable – web scale
l  Inherently distributed
l  Many tools available
l  Runs on commodity hardware
l  Not really new, just a refinement

57
What are Analytics?
l  Trying to duplicate human analyses
l  Make associations
l  Predict the probable future
l  Optimize business processes
l  Need to filter “noise” from Data deluge with New IT tools

58
Big Data Analytics
l  High volume data, High velocity information
u  “How fast can you process the data?”
u  Need real-time results
l  Analytics – extracting relevance for decision making
l  Google is the pioneer in this space
l  Big Data or Small Fast Data?
u  Analyze your requirements before jumping in
l  Great match for cloud computing

59
So What REALLY is “Big Data”?
High
Performance
Computing
Sophisticated
Analytics
“Big Data”
Business agility and enhanced flexibility with real-time answers

60
•  Specific and predictive insight into your users
•  Highly granular decision making for more revenue
opportunities
Technology Promises
•  Many Big Data and NoSQL solutions
•  Huge storage capacities and powerful machines
•  WebSocket events allow Big Data CEP
IT Challenges
•  Typical Big Data solutions are batch, not real-time
•  Determine if truly need Big Data or Small Data
•  Store all the data or subset? Which data is most important?
•  Lack of good data scientists
•  Bus. managers not trained in predictive modeling, forecasting

61
Changing Role for Enterprise IT Architects
l  Enterprise apps adopt more external services
l  Increasing use of cloud computing
u  Salesforce.com (CRM) is only the beginning
l  Need to extend services to partners/customers
l  Growing focus on security
l  New Programming Languages and tools
l  Less than 50% of IT Managers report they are investing
in increasing effectiveness of employees
l  Skills gap
l  What happened to training?

62
New Tools
500M page views per day,
15B page views per month
~3TB of new data per day
1TB+ per day into Hadoop cluster
1000+ servers
20 engineers
Acquired by Yahoo for €.85B ($1.1B)
PHP
Scala Ruby
Redis
HBase
Varnish
nginx
Memcache
Gearman
Kafka
Finagle
Thrift
Func
Git
Capistrano
Puppet
Jenkins
CentOS
How many of these tools
have you heard of?

63
Competitive technology staff
•  Reliable, robust architecture design for the business
•  Optimize costs
Technology Promises
•  Rapid development
•  Extremely scalable
IT Challenges
•  Skills gap with existing staff
•  Focus on regular training

64
Conclusions
l  New IT Tools are powerful and game-changing
l  Cloud, Mobile, Services, WebSocket, Big Data, etc.
l  Have a strategic plan
l  Have a well-understood strategy and implementation
l  With new technology, have many milestones
l  Negative results are useful as well as positive
l  Don’t use every technology all at once
l  Ensure staff is well-trained on New IT

Questions?
frank.greco@kaazing.com
@frankgreco

Opportunities, Promises and Challenges of the New IT Model for Enterprises

More Related Content

What's hot (20)

Similar to Opportunities, Promises and Challenges of the New IT Model for Enterprises (20)

More from Frank Greco (7)

Recently uploaded (20)

Opportunities, Promises and Challenges of the New IT Model for Enterprises