John Esler Principal Technologist Unisys 30950 – Virtualization of Enterprise Applications
Jun 13, 2015
John Esler
Principal Technologist
Unisys
30950 – Virtualization of Enterprise Applications
Agenda
• Real-Time Infrastructure• Virtualization• VMware Workload Test• PeopleSoft Virtualization Test• Orchestration
What Is Real-Time Infrastructure?
And realize the vision of Business Driven IT
Consolidation• Reduction in physical footprint
Virtualization• Resource sharing for effective
infrastructure utilization
Utility• Models built for flexible resource
utilization
Infrastructure Management Suite (IMS)• Tools that enable Real Time
Infrastructure
Autonomics• Self-management of infrastructure
resource
These capabilities…
• Decentralized
• Complex
• Resource-intensive
• Costly
…allow you to transition from this…
• Centralized
• Simplified
• Maximum utilization
• Efficient
• Cost-effective
…to this
Future-state model delivering autonomic infrastructure capabilities to significantly reduce the overall cost of IT
Business agilityService-
level deliveryFlexibility,
reduce costsEconomies of scaleReduce
complexityReact
AgilityMonths to weeks
Weeks Weeks to days Weeks to minutes Minutes Minutes to seconds
EconomicsSubsidized Cost centre Static usage Flexible usage
costingVariable usage costing
Variable business investment
Quality of Service
No SLAs Basic SLAs Class-of-service SLAs
Flexible SLAs End-to-end SLAs
Business SLAs
Real-Time
Service-Based
Virtualized
Rationalized
Basic
Standardized
Source: Infrastructure Futures: Virtualized, Agile and Real-Time; Gartner 25th Annual Data Center Conference 2006, Thomas Bittman
IT Management Process Maturity
Chaotic Reactive Proactive Service Value
Journey to Real-Time Infrastructure
The Infrastructure Maturity Model
Are you Feeling the Pain?
• Glut of IT infrastructure – Most IT organizations buy a new
server every time they deploy a new application
– Decentralized IT resources (islands of control)
– Expensive to operate and maintain (more than 70 percent of IT budget spent “keeping the lights on”)
– Excessive power consumption cooling and real-estate
• Fixed, static resources – few virtualized resources
– Applications and data tied to hardware and network paths
– Rampant over-provisioning, unacceptable levels of stranded capacity
– CPU utilization – Utilization rate from 5 percent to 10 percent on x86 servers is the rule rather than the exception
IT does not align to business priorities and adapt rapidly to business change
020406080
100120140160180200
2006 2007 2008 2009
IT Operational Spending is Growing Disproportionately
Source: Unisys research
Customer requirements• High utilization of assets • Automation to reduce IT human capital• Proactive management of IT assets• Visibility and intelligent decision making• Dynamic and flexible infrastructure
$B
Cost of IT Operations
Capital Costs
Shift operational spending to new business priorities
Unisys Real-Time Infrastructure Solutions
Enterprise Virtualization
Provides mission-critical class virtualization, tools, services, and cost allocation software to reduce costs and improve quality of service
Enterprise Orchestration
Provides business-driven resource automation for combined physical and virtual environments to assure service level performance and maximum infrastructure efficiency
Business SLAsEnd-to-end SLAsFlexible SLAsClass-of-service SLAs
Basic SLAsNo SLAsQuality of Service
Variable business investment
Variable usage costing
Flexible usage costing
Static usageCost centerSubsidizedEconomics
Minutes to secondsMinutesWeeks to minutesWeeks to daysWeeksMonths to weeksAgility
Business agilityService-level
deliveryFlexibility, reduce costsEconomies of
scaleReduce complexity
React
Business SLAsEnd-to-end SLAsFlexible SLAsClass-of-service SLAs
Basic SLAsNo SLAsQuality of Service
Variable business investment
Variable usage costing
Flexible usage costing
Static usageCost centerSubsidizedEconomics
Minutes to secondsMinutesWeeks to minutesWeeks to daysWeeksMonths to weeksAgility
Business agilityService-level
deliveryFlexibility, reduce costsEconomies of
scaleReduce complexity
React
Service-Based
Virtualized
Rationalized
Basic
Standardized
Source: Infrastructure Futures: Virtualized, Agile and Real-Time; Gartner 25th Annual Data Center Conference 2006, Thomas Bittman
Business SLAsEnd-to-end SLAsFlexible SLAsClass-of-service SLAs
Basic SLAsNo SLAsQuality of Service
Variable business investment
Variable usage costing
Flexible usage costing
Static usageCost centerSubsidizedEconomics
Minutes to secondsMinutesWeeks to minutesWeeks to daysWeeksMonths to weeksAgility
Business agilityService-level
deliveryFlexibility, reduce costsEconomies of
scaleReduce complexity
React
Business SLAsEnd-to-end SLAsFlexible SLAsClass-of-service SLAs
Basic SLAsNo SLAsQuality of Service
Variable business investment
Variable usage costing
Flexible usage costing
Static usageCost centerSubsidizedEconomics
Minutes to secondsMinutesWeeks to minutesWeeks to daysWeeksMonths to weeksAgility
Business agilityService-level
deliveryFlexibility, reduce costsEconomies of
scaleReduce complexity
React
Service-Based
Virtualized
Rationalized
Basic
Standardized
The Infrastructure Maturity Model The Infrastructure Maturity Model
IT Management Process Maturity
Chaotic Reactive Proactive Service ValueIT Management Process Maturity
Chaotic Reactive Proactive Service ValueIT Management Process Maturity
Chaotic Reactive Proactive Service ValueIT Management Process Maturity
Chaotic Reactive Proactive Service Value
Real-TimeReal-Time
Unisys Real-Time Infrastructure Solutions
Real-Time Infrastructure Capabilities
Element Virtualization
Com
plia
nce/
Cha
nge
Man
agem
ent
Res
ourc
e U
tili
zati
on
IT Service Governance
Orchestration/Workflow
Provisioning and Re-Purposing
Server Virtualization Network Virtualization Storage Virtualization
Server Network Storage
Business agilityService-
level deliveryFlexibility,
reduce costsEconomies of scaleReduce
complexityReact
AgilityMonths to weeks
Weeks Weeks to days Weeks to minutes Minutes Minutes to seconds
EconomicsSubsidized Cost centre Static usage Flexible usage
costingVariable usage costing
Variable business investment
Quality of Service
No SLAs Basic SLAs Class-of-service SLAs
Flexible SLAs End-to-end SLAs
Business SLAs
Real-Time
Service-Based
Virtualized
Rationalized
Basic
Standardized
Source: Infrastructure Futures: Virtualized, Agile and Real-Time; Gartner 25th Annual Data Center Conference 2006, Thomas Bittman
IT Management Process Maturity
Chaotic Reactive Proactive Service Value
Let’s Talk Virtualization
The Infrastructure Maturity Model
What the Market is Saying
"Tools like virtual machines have clearly crossed the chasm and are being used by IT managers
in a variety of environments." - Tony Iams, VP & Sr Analyst, Ideas International
Virtualization is an example of how companies are squeezing every line item –spending less on server hardware, system administrators, and software licenses.
- InformationWeek, Sept 11, 2006
According to Computerworld's Vital Signs survey of 252 IT executives, server virtualization topped the list of technologies being tested.
- Computerworld, Jan 01, 2007
Server Virtualization Adoption is Exploding
User adoption grew 11% in 2006 and now exceeds 50% of firms- Forrester, December 2006
Firms are expanding virtualization into production and critical workloads- Forrester, December 2006
Firms are learning that infrastructure upgrades make server virtualization more useful - Forrester, December
2006
Problems Addressed by Virtualization
Under-utilized servers
Consolidation return
Slow response to business need
High cost of ownership in the data
center
Too many servers, not enough space
New applications take too long to test and
deploy
Lack of information on infrastructure
composition
Traditional consolidation difficult
to cost justify
Virtual Machine Technology Benefits
• Isolate workloads to prevent conflicts• More workloads per physical server/fewer servers• Increased utilization (CPU/memory/IO)• Ability to very rapidly provision new servers – minutes compared to
days or weeks for physical servers– “Real benefit lies in increased developer productivity, not reduced server
expense”
• Flexibility• Workload management – manual and automated • Automated deployment• Ease of backup and recovery• Enables on-demand and pay-as-you-go utility compute environment
busy teller
customer
Many Queues
idle teller
One Queue
busy tellers
Bank Teller Example
Which Model is Better?
• Unisys Whitepaper documenting Workloads on VMware
• On our Website
Commodity Server4-Socket System Test Results
4p 12 VMsPhysical Cpu (Total) % Processor Time
60% Average Utilization
0
20
40
60
80
100
0:28
0:31
0:33
0:35
0:38
0:40
0:42
0:45
0:47
0:49
0:52
0:54
0:56
0:59
1:01
1:04
1:06
1:08
Duration (h:mm)
%cp
u
3 VMs/
physical C
PU
Unisys Server Test Results
16p 96 VMsPhysical Cpu (Total) % Processor Time
73% Average Utilization
0
20
40
60
80
100
0:22
0:24
0:27
0:29
0:31
0:34
0:36
0:39
0:41
0:44
0:46
0:48
0:51
0:53
0:56
0:58
1:01
1:03
Duration (h:mm)
%cp
u
6 VMs/
physical C
PU
Scale-Up Economic Analysis
Scale-up model Scale-out Model
16 processors 32 processors
128 VMs 128 VMs
Base Platform (includes three-year maintenance and support cost)
204,715One 16X ES7000
178,544Eight 4X x86 servers
Power/Cooling (three-year cost) 26,910 48,941
VMware Cost 80,500 161,000
Connectivity Costs 8,700 24,800
Management Costs - ESX Host Mgmt (three-year cost) 4,680 12,480
Microsoft Windows OS Requirement, Datacenter Edition 62,021 156,422
Total Three-year Cost 387,527 582,207
Three-year TCO/VM 3,028 4,548
Annualized cost/VM 1,009 1,516
Percent TCO savings 33.4%
As of February 1, 2007
PeopleSoft Test
• Unisys and VMware published paper on HCM deployment
• Find it on VMware’s website
PeopleSoft and ESX 3.01
• Objectives
– Conduct Analysis around Virtualization of PeopleSoft Applications Tier
– Use VMware Virtual Center/VM Converter for Cloning and Creating Templates for PeopleSoft Application VMs
– Produce documentation for Installation, Configuration and Best Practices
Test Methodology
Configuration
Virtualization – Affinity 1-3 Core
Virtual and Bare Metal 8 Core
Virtual – Bare Metal 2 Core
Business agilityService-
level deliveryFlexibility,
reduce costsEconomies of scaleReduce
complexityReact
AgilityMonths to weeks
Weeks Weeks to days Weeks to minutes Minutes Minutes to seconds
EconomicsSubsidized Cost centre Static usage Flexible usage
costingVariable usage costing
Variable business investment
Quality of Service
No SLAs Basic SLAs Class-of-service SLAs
Flexible SLAs End-to-end SLAs
Business SLAs
Real-Time
Service-Based
Virtualized
Rationalized
Basic
Standardized
Source: Infrastructure Futures: Virtualized, Agile and Real-Time; Gartner 25th Annual Data Center Conference 2006, Thomas Bittman
IT Management Process Maturity
Chaotic Reactive Proactive Service Value
Let’s Talk Orchestration
The Infrastructure Maturity Model
Problems Addressed by Orchestration
Under-utilized servers
High cost of ownership in the data
center
Increased Complexity
Inability to adapt IT resources fast enough
Maintaining SLA Integrity
Chronic and costly over-provisioning of
server resourcesTime-to-market response for new
business requirements
Mixed physical and virtual environment
too complex to manage
IT spending not optimized
Provisioning and resource management
is inefficient IT not responsive to business
Enterprise Orchestration Solution
• Automated resource management of physical and virtual systems with business rules
• Full provisioning engine for Windows, Linux, and applications
Unisys Real-Time Orchestration and Provisioning
Real-Time Orchestration and Provisioning
• Provision physical and virtual servers
• Configure multiple servers
• Install operating systems remotely
• Optimize resources based on load
• Manage groups of servers
• Facilitate high availability failover
• Enable composite application resource management
• VMware, Zen, MS Interface
Key Features and Capabilities
Unisys Plan
Phase 1
Phase 3
Phase 4
Phase 2
Manage physical and virtual Servers
Support for business rules IT resources flexibly and
dynamically allocated between services based on changing business controlled via service
governance
Provision other vendors servers
Provision physical and virtual servers
Storage and network orchestration
Service governance
Composite apps policy
Enhanced service governance
Automation policy