© Copyright 2015 Vivit Worldwide© Copyright 2015 Vivit Worldwide
“Revolutionize and Transform your Application Delivery Lifecycle
with HP Service Virtualization”October 22, 2015
© Copyright 2015 Vivit Worldwide
Brought to you by
© Copyright 2015 Vivit Worldwide
Hosted by
Todd DeCapua
Champion for Vivit Performance Engineering SIG
HP Software, Senior Manager
Chief Technology Evangelist
© Copyright 2015 Vivit Worldwide
Today’s Speakers
Kimm Yeo
WW Senior Product Marketing Manager
HP Service Virtualization
Mirek Novotny
Senior Product Manager
HP Service Virtualization
© Copyright 2015 Vivit Worldwide
Housekeeping
• This “LIVE” session is being recorded
Recordings are available to all Vivit members
• Session Q&A:
Please type questions in the Questions Pane
© Copyright 2015 Vivit Worldwide
Webinar Control Panel
Questions
Toggle View Window between
Full screen/window mode.
Revolutionize and transform your App Delivery Lifecycle with HP Service Virtualization
Mirek Novotny, Sr Product Manager
Kimm Yeo, Sr Product Marketing Manger
Oct 22, 2015
7
Deliver amazing user experiences
Modern application development
Reduce costs
Increase customer attraction/retention
Increase the value of your brand
Get to market faster
Mobile
Cloud
Dev
Ops
Agile
Modern applications redefine application development
* Source: “Enterprise Mobile Facts You Need to Know in 2015” by App Data Room.
http://appdataroom.com/enterprise-mobile-facts-need-know-2015/
Velocity, quality and user experience
Need for
VELOCITY
Demand for
QUALITY
50% of consumers will
delete a mobile app if
they encounter a bugAPMdigest, Feb 5, 2014
30x increase in
application releasesEnterprise 20/20 Research, 2013
Access Anywhere
Composite Applications
Analytics
Proliferation of Tools
Shift “Left”Visibility
PredictiveElasticSecureScalable
Agile
User Experience
UX
How do you manage composite applications?Composite app software dev and test is often like “herding cats”
Search PayBuy Ship Track
Integrated to support the process
Traditional
●●
Public CloudTrack package
Payment
Composite
Product Customer Profiles Shipping
Managed Cloud
Each have SLAs and performance
dependencies
Components anywhere– in-house or
cloudShopping Process
ConstrainedDelayed
Severely RestrictedInhibit
Hold up
Limit Waiting
Deprive
Missed deadline
Reduced Testing
Development
Slowdown
11
ConsequencesYou can’t do anything until you have everything; and you never have everything
84%QA work delayed while waiting
81%Development work delayed
while waitingSource: voke Market SnapshotTM Report: Service Virtualization – January 2015
?
Compressing the software development lifecycle with service virtualization
Without service virtualization
With service virtualization
DevelopmentFunctional
TestIntegration
Performance Test
UAT
Development
Functional
Integration
Performance
UATTime saved
Wait timeBefore Service
Virtualization
On-demand (no wait) 0%
Seconds to minutes 0%
30 minutes to 2 hours 0%
4 to 7 hours 2%
1 day 1%
2 days 1%
3 days 9%
4 days 3%
1 week 8%
2 weeks 15%
3 weeks 27%
1 month 14%
2 months 10%
3 months 5%
4 to 6 months 3%
Never (no access
ever)2%
Source: voke Market SnapshotTM Report: Service Virtualization – January 2015
The consequence: crippling wait times
32 days
average wait time
Wait timeBefore Service
Virtualization
After Service
Virtualization
On-demand (no wait) 0% 27%
Seconds to minutes 0% 14%
30 minutes to 2 hours 0% 10%
4 to 7 hours 2% 17%
1 day 1% 11%
2 days 1% 10%
3 days 9% 8%
4 days 3% 1%
1 week 8% 1%
2 weeks 15% 1%
3 weeks 27% 0%
1 month 14% 0%
2 months 10% 0%
3 months 5% 0%
4 to 6 months 3% 0%
Never (no access
ever)2% 0%
From 32 days to 1 hour, by virtualizing services
1 hourmedian wait time after
service virtualization
Source: voke Market SnapshotTM Report: Service Virtualization – January 2015
ROI findings from latest Service Virtualization studyYears 2014-15
Release Cycle Time
40% achieved a decrease of 40% or greater in software release cycle
time
Test Cycle Time 34% achieved a decrease of 50% or greater in test cycle time
Defect Reproduction
Time
38% achieved a greater than 50% reduction in defect reproduction
time
# of Production Defects 36% achieved a greater than 41% reduction in production defects
# of Total Defects 46% achieved greater than 41% reduction in total defects
Test Coverage 20% achieved more than 2x the test coverage
Test Execution 26% achieved an increase of 2x or greater of test execution rates
Source: voke Market Snapshot Report Service Virtualization – Jan 2015 http://www.hp.com/go/sv/SVReport2015
The solution
Airlines simulate
18
NASA simulates
19
Why don’t developersand testers simulate?
20
• 20-40% Developer time spent
maintaining stubs
• Restricted functionality
• No self-service for customization
“The core issue with stubbing was that the process was not repeatable or
scalable.”
Ovum Report on Service Virtualization, January 2015
Stubs?
21
STUBS
Existing Infrastructure
Third Party
Single sign on
JDBC
SOAP
MQ
LDAP
Mainframe
Existing database
Web service and
Legacy application
REST
JMS
Service virtualization becomes a key answerKeep developing and testing moving forward with virtualized services
• Virtualize for always available services
• Virtualize data scenarios
• Make available for Dev and Test
• Share services—lower infrastructure costsThird Party
JMSUnder
construction
Pay-per-transaction
Lack of Data Controls
Application Under Test
Mobile
App
Web
browser
Composite
Application
Application
services
Simulation
Service
Virtualization
SOAP
REST
JMS
MQ
Data
Perf.
Config
22
SV Usage in simple cases
Users Network Infrastructure AUT Configuration Provision Infrastructure
Provision Software
Provision Data
Browsers and
Devices
Customers
Employees
Major ISP
3rd Party
Cloud Services
Web
Servers
Load
Balancers MainframeDB
Servers
App
Servers
Web
Services
Storage
3rd party
Mobile
Components
Complexity of app dev and test environment
ClientCarriers
& ISPs
Content Delivery
Networks SecuredBackend and
3rd party Systems
Align Test Data
Application
Users
Browsers&
Devices
Customers
Employees
AUT ConfigurationNetwork Infrastructure
Major ISP
3rd Party
Cloud Services
Web
Servers
Load
BalancersApp
Servers
Mobile
Components
Simplify DevTest environments
ClientCarriers
& ISPs
Content Delivery
Networks SecuredBackend &
3rd party SystemsApplication
web
services
S
V
S
V
distributed
services
= Virtualized Services S
V
Users AUT Configuration
Browsers&
Devices
Customers
Employees Web
Servers
Load
BalancersApp
Servers
Mobile
Components
Simplify DevTest Environments
ClientCarriers
& ISPs
Content Delivery
Networks SecuredBackend &
3rd party SystemsApplication
web
services
S
V
S
V
distributed
services
N
VN
V
N
V
N
V= Virtualized Network = Virtualized Services
S
V
N
V
AUT Configuration
Browsers&
Devices
Customers
Employees Web
Servers
Load
BalancersApp
Servers
Mobile
Components
Simplify DevTest environments
ClientCarriers
& ISPs
Content Delivery
Networks SecuredBackend &
3rd party SystemsApplication
V
U
V
U
= Virtualized Users V
U
web
services
S
V
S
V
distributed
services
N
VN
V
N
V
N
V= Virtualized Network = Virtualized Services
S
V
Shorter test
cycles50%
N
V
V
U
Early App development and testing
Application
Backend
Mainframe
CRM
3rd Party
Payment
REST
SV
Virtual
Claim
ServiceUnder Construction
Simulated
Transactions
Development/Design
Documentation
Faster
software
cycle
40%
Performance Insights
What if? - performance validation3 challenges with 3 solutions in 1 simulated environment
Application
Backend
Mainframe
CRM
3rd Party
Payment
SV
Virtual
MF
Virtual
CRM
Virtual
Pay
Mobile App
Desktop App
Web App
NV
Network
Condit.
WiFi
3G4G
VPN
2.5G
21
3
Increased test
coverage2x
Setup What-if Conditions
Load Generators
Virtual
users
LR/PC/SR
Restricted
access
Speed Limit
High Fees
Demo
30
The tool
HP Service Virtualization 3
32
Quick and easy virtual service creation
• Easy and Intuitive IDE
• Embedded SV Runtime for local
use
• Predefined Virtualization starting
points
• Dialog based service creation
• Data oriented Functional
Modeling
• Learning, Data Driving, Manual
Authoring (Request-Responses)
• Performance and Scalability
Modeling
• Simulation Logging and Preview
• Pre-defined Technologies with
Extensibility SDK
Visual Modeling IDE
Dialog Based Wizards
Visual Data Modelling
Visual Performance
Modelling
Pre-packaged Technologies
33
Scalable virtual service infrastructure with shared mgmt.
• Shared, Scalable and Secured
Simulation Infrastructure
• Web Based Virtual Service
Management Interface and
Dashboard
• Unified VS management across
multiple SV Server Nodes
• Parameterized filtering and search
• Provisioning and Control of Virtual
Environments
• ACL management – Users/groups
• Integrated to ALM/QC, VCS and
other repositories
• Integrated with Enterprise Identity
System (LDAP)
Management and
Administration
Virtual Service Provisioning and Control
Virtual Service and Server
Metrics
SV Server Dashboard
34
Designed for use with HP Testing Toolset UFT/MC/LR/PC/SR/NV
Virtual Services Real-
time Performance
Virtual Services Real-time
Performance
Test reports with Virtual
Service metrics stored
in ALM
Control VirtualizationFrom Inside UFT/LR/PC
35
Network ModelIntegrated Shunra Network Virtualization and HP Service Virtualization
Existing Infrastructure
Single sign on
JDBC
SOAP
MQ
LDAP
Third Party
Application Under Test
Mobile
App
Web
browser
Composite
Application
Application
services
Mainframe
Existing database
Web service
and
Legacy
application
RESTPay-per-
transaction
Under
construction
Service
Virtualization
SOAP
JDBC
JMS
MQ
REST
CICS
Simulation
Network
Virtualization
Virtualized!
• Define and Control network parameters with
SV Virtual Services
• New Network Model definitions
• Integrated to Shunra Global Library for
quick network parameters import
Data
Perf.
Config
Integrated with HP ALM/QC, SVN, Jenkins
37
• Store Virtual Services with
Development or Testing
Assets
• Version Control
• Collaboration with Check-
out/Check-in support
• Complete traceability
• ALM integrated from SV
Designer, Server and SV
Integration in Automation
Tools
Virtual Service Projects
Versions, Revisions and Dependencies
ALM & VCS Repositories in SV Designer
SV Management Integrated with
ALM/QC Repository
Quarterly Expanding Virtualization Protocols
TRANSPORT
HT
TP
(S)
Gate
way
HT
TP
(S)
Pro
xy
1
IBM
WS
MQ
1
JM
S
JD
BC
1
IMS
Co
nn
ec
t
CIC
S T
S
TIB
CO
A
ctive
Ma
trix
/
EM
S1
SA
PN
etW
eaver
XI/P
I, A
BA
P
Ora
cle
AQ
Mic
ros
oft
MQ
We
b-
Me
tho
ds
IS1
TC
P/I
P
JD
K
(Beta
)
MESSAGE
WS/SOAP
XML2
REST (XML, JSON,
Bin)
Cobol
SQL
RFC/IDOC
Fix Length
Java Objects
Text
Binary
Protocol supported1 Non-intrusive2 All XML-based protocols supported
SV Protocol Extensibility SDK Available
Virtual service management automationIntegrate 3rd party development tools, ides, build and Continuous integration systems
• Control Virtual Services from command line and external scripts
using SVConfigurator tool
• Supports full Virtual Service lifecycle over API/Command Line
• Java based, multi OS, supporting ANT tasks
• Deploy/un-deploy Virtual Services to/from any HP SV Server
• Change mode of a Virtual Service (including Learning)
• View Virtual Service details and metrics;
• List/Export/Update deployed Virtual Services and Projects
• Unlock Virtual Service locked by another user
• JavaDoc like documentation
39
Extensibility and integration
• Integration and Automation APIs
• REST Based Management API and Command line automation interface
• Protocol Extensibility SDK
• Transport & Message Level Extensibility
• Binary TCP Protocol SDK
• TCP based protocol Accelerator framework
• Simulation Extensibility
• Simulation Functions
• Data Masking Algorithms
• Service Description
SV SDK Updates (3.62)
• Easily create new protocols from
predefined Protocol Templates
• SOAP over HTTP, XML over HTTP, Text based
TCP protocol
• Easily introduce new custom protocols
based on HTTP by leveraging HTTP based
protocol templates
40
HTTP Virtualized Services
Invocation
R&D Laptop6GB
Windows 7 x64 Windows Server 2008 R2 x64Intel® Xeon® X5660 2.8 GHz
2x6 Cores, Hyper-threading OFF32GB
DB Storage
Windows Server 2008 R2 x64Intel® Xeon® 5160 2.66 GHz
2x2 Cores16GB
4 x LoadRunner GeneratorWindows Server 2008 R2 x64
2 x Intel® Xeon® 5150 2.66GHz8GB
Virtual Service Management
Co
nn
ect
ion
s
Storage Access
SV ServerSV Designer
Service Clients
SQL Server 2008 R2
IBM MQ 7.0.1.3Windows Server 2003 R2 x64Intel® Xeon® 5150 2.66 GHz
2x2 Cores8GB
Queue
Send./R
ec.
Queue Send./Rec.
HP Service Virtualization 3
Benchmark Objective
HP Service Virtualization performance benchmark studies
maximum number of transactions (requests) per second of HP
Service Virtualization Server v3.00
In this test 4*N clients are invoking two randomly chosen services
from 200 deployed services in SV Server - where N is number of
SV Server CPU cores.
Figure 1. Deployment diagram.
Table 1. Benchmark characteristics.
Deployment
HP Service Virtualization Server is installed on Intel Xeon X5660
with 2x6 CPU Cores, 32GB memory machine. Database was
installed on Xeon 5160 with 2x2 CPU Cores, fixed 300GB 15k disk
as a storage. To load virtualized services we used four HP Load
Runner generators.
Parameter Value
Message structure 30 elements (average)
Deployed services 200
Concurrent clients 48
Service model size 1000 unique messages
Protocol HTTP/SOAP
Physical Server 1
Physical Server 2
4 x Physical Server
Laptop
Maximum Transactions Benchmark
Conclusion
Maximum transactions benchmark test shows scalability of HP
Service Virtualization Server. SV Server is capable of handling
2900 TPS and 3700 TPS with in-memory simulation on 12 core
CPU Intel machine with linear grows of response time after full
CPU utilization.
Results
HP Service Virtualization Server version 3.00 simulates 2900 TPS
(transactions per second) or 10,440.000 TPH (transactions per hour)
in steady state with 48 parallel users (Figure 2)
In-memory and pre-loading of simulation data into memory enhances
performance by up to 25%, depending on scenario.
HP Service Virtualization Server performance improves to 3700 TPS
or 13,320.000 TPH with in-memory simulation enabled.
Figure 2. Number of transactions per second.
Figure 3. Virtual Service response time dependence on
number of parallel users. Each color represents different
service operation.
Figure 4. server CPU (red) and database CPU (green)
utilization.
Response time is linearly dependent on number of parallel users
(Figure 3) It goes down after a short warm up and linearly up from
about 20 users. This linear growths is consequence of high CPU
utilization This slowdown is typical for today’s server based
applications.
CPU utilization grows linearly until its 100% capacity at which point
slowdown of response time starts growing (Figure 4)
SV Server CPU utilization is more significant than database CPU
utilization, thus limiting factor for SV Server scalability is number of
server CPU Cores.
2900 TPS with
Database
3700 TPS with In-memory
Simulation
41
Summary
Portfolio-wideintegration
SAP certification
Why HP Service Virtualization stands out
Easy set-upand use
Complete lifecycle virtualization solution
How to Get Started:
• Download latest 2015 full detailed report on Voke service virtualization study at http://www.hp.com/go/sv/SVReport2015
• Download HP SV Release 3.7 for a free 30day trial at http://www.hp.com/go/sv/freetrial
• HP SV Release datasheet available on handout panel and at http://www.hp.com/go/sv
Limited Time only :
• Participate in HP Service Virtualization survey & be placed in a drawing for Bluetooth speaker with tablet stand !
Go, virtualize!hp.com/go/SV
Appendix
© Copyright 2015 Vivit Worldwide
HP Discover 2015 London
• December 1-3, 2015 London
• Register Now via the unique Vivit link:
http://hpsw.co/y9T3Bzj
• If you use our unique link when you
register and enter your Vivit ID you can
still receive the Early Bird discount!
• Check out Vivit Breakout Sessions!
Details coming soon!
© Copyright 2015 Vivit Worldwide
Thank you
• Complete the short survey and opt-in for more information from HP
Software and you will be entered in a drawing for a Bluetooth
speaker tablet stand
www.hp.com
www.vivit-worldwide.org