Transcript: #StandardsGoals for 2024: What’s new for BISAC - Tech Forum 2024
Keith Prabhu - Big Data Cloud Computing
1. Big Data Computing
Business Value
Keith Prabhu
Master of Business (Australia), CCSK, MBCI, CISSP, CISA
Executive Director, Confidis Advisory Services
Founder & Director, Cloud Security Alliance, Mumbai Chapter
November 21, 2013
Technology Consulting
www.confidis.co
3. Big Data Trends
•
•
•
•
Data is becoming more valuable
• No longer a question of “what to store?” but rather “what
can we do with the data?” – Implication increased data
volumes
• As per Gartner:
• Enterprise data will grow by 800% from 2011 – 2015
• 80% will be unstructured, 20% structured
• Need for combining diverse data sets
Data Analytics is moving from batch to real time
• Real time analytics technology is part hype stage and is in
early adoption
Emergence of predictive analytics
• Real time analytics provides organizations predictive
capabilities
Scope of Big Data continues to expand
• Machine data, Sensors etc.
• Ever increasing number of applications of Big Data for
Enterprises e.g. Customer churn analysis, flexible supply
chains, loyalty pricing, ad targeting, POS transaction
analysis etc..
www.confidis.co
BIG DATA
BIG Problem
or
BIG Opportunity
4. Cloud & Big Data – Perfect Allies!
Big Data Requirements
Cloud Computing Capabilities
Need for significant infrastructure investments
Pay as you go Cloud model is ideal for
enterprises “to think big but start small”
Rapid storage scalability
Easy storage scalability
Rapid provisioning of processing power
On the fly provisioning of additional processing
power
Interoperability with other data sources
(internal & external)
Provides a plethora of interfaces to connect to
other platforms
Quick deployment options to get quick insights
Availability of Analytics-As-A-Service (AaaS)
Big Data & Cloud Computing are complementary
www.confidis.co
5. Big Data Cloud - Technology
Service Models
Deployment Models
IaaS:
You get barebones IT infrastructure, you need
to install and manage the Big Data stack
Private:
Big Data stack is deployed on infrastructure
that is dedicated solely for your use
PaaS:
You get the development platform which can
be used to develop analytics applications
Public:
Big Data stack is deployed on a public cloud
that uses shared resources
SaaS (AaaS)
You get ready to run analytics application. Plug
in your data and switch on!
Hybrid:
Big Data stack is deployed across private and
public clouds
Big Data & Cloud Computing are complementary
www.confidis.co
6. Big Data Cloud –
Benefits & Challenges
Options
Public Cloud
Leverage Public
Data PaaS
Frameworks like
Amazon Elastic
MapReduce
Implications
•
•
Pros & Cons
Easy to get up and
running e.g. Amazon map
reduce jobs (EMR, EC2)
+ Easy and low cost
External data upload
requirement poses
security concerns
-
+ Can be used to study the map
reducibility problem
Limited Public Data PaaS options
available
- Vendor lock-in
•
Leverage Public
IaaS and setup Big
Data Solutions on
them
www.confidis.co
Significant size and
processing times impacts
latency
•
Requires on-premise Map + No procurement lead times
Reduce Cluster
+ No vendor lock-in
Data Transfer
+ May be cloud-neutral
requirement remains the
- Perimeter security concerns, data
same
transfer speed (more time to run)
and high latency limitations apply
•
- Perimeter security concerns, data
transfer speed (more time to run)
and high latency limitations apply
7. Big Data Cloud –
Benefits & Challenges
Options
Private Cloud
Build the Map Reduce
clusters on a set of virtual
machines
Implications
•
•
Provision VMs on shared
infrastructure
Setup Map Reduce
clusters on shared storage
infrastructure
Pros & Cons
+ Easier provisioning
+ Dynamic scale out of
infrastructure
- Shared infrastructure and
virtualization overheads may
impact latency
•
Build out the Map Reduce
clusters on its own
dedicated hardware
Overheads of virtualization
and limitations of
architecture
•
Provision dedicated
hardware
+ High performance, low
latency
•
Build out Map Reduce
clusters on a share
nothing setup
- Increased cost with lower
utilization of dedicated
hardware
- Procurement lead time
Source: http://www.infosys.com/cloud-services/resources/pages/big-data-spectrum.aspx
www.confidis.co
9. Cloud Security Alliance
•
•
•
•
Global, not-for-profit organization
Over 45,000+ individual members, 100+ corporate
members
Building best practices and a trusted cloud ecosystem
Agile philosophy, rapid development of applied
research
− GRC: Balance compliance with risk management
− Reference models: build using existing standards
− Identity: a key foundation of a functioning cloud
economy
− Champion interoperability
− Advocacy of prudent public policy
“To promote the use of best practices for providing
security assurance within Cloud Computing, and
provide education on the uses of Cloud Computing to
help secure all other forms of computing.”
Join: Cloud Security Alliance, Mumbai Chapter on LinkedIn
(http://www.linkedin.com/groups?gid=2963138)
www.confidis.co
11. Contact Us
For any further
information,
please contact:
www.confidis.co
Keith Prabhu
Executive Director
Confidis Advisory Services
Private Limited
Email: info @ confidis DOT co