HPE Network Virtualization for Performance Testing

Transcription

1 HPE Network Virtualization for Performance Testing Software Version: User Guide Document Release Date: May 2016 Software Release Date: May 2016

2 Legal Notices Warranty The only warranties for Hewlett Packard Enterprise Development LP products and services are set forth in the express warranty statements accompanying such products and services. Nothing herein should be construed as constituting an additional warranty. HPE shall not be liable for technical or editorial errors or omissions contained herein. The information contained herein is subject to change without notice. Restricted Rights Legend Confidential computer software. Valid license from HPE required for possession, use or copying. Consistent with FAR and , Commercial Computer Software, Computer Software Documentation, and Technical Data for Commercial Items are licensed to the U.S. Government under vendor's standard commercial license. Copyright Notice Copyright Hewlett Packard Enterprise Development LP Trademark Notices Adobe is a trademark of Adobe Systems Incorporated. Microsoft and Windows are U.S. registered trademarks of Microsoft Corporation. Linux is the registered trademark of Linus Torvalds in the U.S. and other countries. Java is a registered trademark of Oracle and/or its affiliates. Red Hat is a registered trademark of Red Hat, Inc. in the United States and other countries. VMware is a registered trademark of VMware, Inc. in the United States and other countries. Documentation Updates The title page of this document contains the following identifying information: Software Version number, which indicates the software version. Document Release Date, which changes each time the document is updated. Software Release Date, which indicates the release date of this version of the software. To check for recent updates or to verify that you are using the most recent edition of a document, go to: This site requires that you register for an HPE Passport and sign in. To register for an HPE Passport ID, go to and click Register. Support Visit the HPE Software Support Online web site at: This web site provides contact information and details about the products, services, and support that HPE Software offers. HPE Network Virtualization for Performance Testing (12.53) Page 2 of 137

3 HPE Network Virtualization for Performance Testing HPE Software online support provides customer self-solve capabilities. It provides a fast and efficient way to access interactive technical support tools needed to manage your business. As a valued support customer, you can benefit by using the support web site to: Search for knowledge documents of interest Submit and track support cases and enhancement requests Download software patches Manage support contracts Look up HPE support contacts Review information about available services Enter into discussions with other software customers Research and register for software training Most of the support areas require that you register as an HPE Passport user and sign in. Many also require a support contract. To register for an HPE Passport ID, go to: and click Register. To find more information about access levels, go to: HPE Software Solutions & Integrations and Best Practices Visit HPE Software Solutions Now at /facetsearch/document/km to explore how the products in the HPE Software catalog work together, exchange information, and solve business needs. Visit the Cross Portfolio Best Practices Library at to access a wide variety of best practice documents and materials. About this PDF Version of Online Help This document is a PDF version of the online help. This PDF file is provided so you can easily print multiple topics from the help information or read the online help in PDF format. Because this content was originally created to be viewed as online help in a web browser, some topics may not be formatted properly. Some interactive topics may not be present in this PDF version. Those topics can be successfully printed from within the online help. HPE Network Virtualization for Performance Testing (12.53) Page 3 of 137

4 HPE Network Virtualization for Performance Testing Contents HPE Network Virtualization for Performance Testing 1 What's new 8 NV Analytics Report 8 Updated VuGen Report 8 HPE AutoPass Licensing 9 Network Virtualization for Performance Testing Help Center is available on the Web 9 Microsoft Windows 9 Browser support 9 Chapter 1: Introducing Network Virtualization for Performance Testing 11 HPE Network Virtualization 11 Chapter 2: Network Virtualization for Performance Testing Installation 13 System requirements 14 Install HPE Network Virtualization for Performance Testing 14 Upgrading from a previous version 14 Installing Network Virtualization for Performance Testing 15 Manual installation 15 Setup wizard options 16 More NV applications 17 Installation logs 18 Silent installation of HPE Network Virtualization for Performance Testing 18 Silently install NV for Controller 19 Silently install NV for Performance Center 21 Silently install NV for Load Generator 22 Silently uninstall HPE Network Virtualization for Performance Testing 23 Post-installation tasks 24 NV log files 24 Chapter 3: Licensing 25 Community license 25 Purchased license 25 Install an NV Analytics license for VuGen 26 Chapter 4: Configure NV profiles with NV Network Editor 27 NV Network Editor overview 27 Open NV Network Editor 27 Create NV profiles 28 HPE Network Virtualization for Performance Testing (12.53) Page 4 of 137

5 HPE Network Virtualization for Performance Testing Configure gateway parameters 29 Configure the WAN parameters 32 Latency and packet loss [WAN Parameters] 33 Packet effects [WAN Parameters] 35 Link faults [WAN Parameters] 37 NV profile parameters 38 Chapter 5: Add and configure virtual locations 41 Enable network virtualization 43 Set the network virtualization mode 44 Chapter 5: Define virtual location test level settings 45 Packet Capture 45 IP filter 47 Select virtual location definition method 48 Configuring virtual locations 48 Define custom virtual location parameters 49 Extract virtual location parameters from the NV Global Library 50 Import virtual location parameters from an NV profile 51 Specify virtual location runtime settings 53 Allocate bandwidth to Vusers 54 Set packet capture for a virtual location 55 Create a local IP filter for a virtual location 56 Exclude machines from network virtualization 57 Chapter 6: Customize NV integration settings 59 Set the active adapter on NV for Load Generator 60 Change the NV for Controller port 60 Change the NV for Load Generator and NV Network Editor port 61 Set the default maximum packet list buffer size for each load generator 62 Enable the cleanup threshold on NV for Load Generator 63 Chapter 7: Analytics 64 NV Analytics / NV Insights Report 65 Dashboard 66 Summary data 66 Analysis coverage 66 Network impact 67 Optimization grades 68 Transaction distribution by script 69 Group by Transaction - Graph View 70 Group by Transaction - List View 73 Group by Virtual Location - Graph View 75 HPE Network Virtualization for Performance Testing (12.53) Page 5 of 137

6 HPE Network Virtualization for Performance Testing Group by Virtual Location - List View 79 NV Analytics Report from VuGen 82 Shared Analytics Report Views 82 Summaries 83 Duration Breakdown 83 Metrics 84 HTTP Analysis 84 Optimization 86 Resources 88 Endpoint Latencies 88 Chapter 8: NV Analytics - Standalone 90 NV Analytics 90 System Requirements 90 Install NV Analytics 91 Log Files 94 Analyzing Results 94 Settings 94 Analyzing a Test Result File 96 Exporting NV Analytics Results 97 Viewing Reports 98 Response Time 99 Summaries 99 Client Network Server Breakdown 100 Parameters 100 General Analysis 101 Display Options 101 Subtransaction Parameters 102 Request/Response Details 102 Throughput 102 Endpoint Latencies 104 TCP/UDP Errors & Sessions 105 HTTP Analysis 106 Subtransaction Details 108 HTTP Parameters 109 HTTP Optimization 110 HTTP Resources and Responses 112 Resources Breakdown 112 HTTP Errors 112 HLS Errors 113 Response Summary 113 Secure Communication 113 NV Analytics API 114 HPE Network Virtualization for Performance Testing (12.53) Page 6 of 137

7 HPE Network Virtualization for Performance Testing AnalysisEngines 114 Code sample 115 Extract Packet Lists 115 Code sample 116 AnalysisRequest 117 AnalysisSummary 120 AnalysisArtifact 122 Structure of an Analysis Report 122 Structure of the HTTP Waterfall Analysis Report 123 Structure of the Best Practices Analysis Report 126 NV Analytics Protocols 127 Supported Protocols 127 Conversation Definition 127 Collecting Conversation Statistics 127 Classification of TCP, UDP, IP 128 Sub-Transaction Grouping 128 Understanding Protocol Association 128 Chapter 9: NV Predictor 129 NV Predictor overview 129 Install NV Predictor 129 Create reports 130 Analyze results 131 Modify locations 132 Update results with classifications 132 Update results with KPIs 133 Select transactions 133 Define global settings 134 Configure NV Predictor to run automatically after analysis 135 Send Us Feedback 136 HPE Network Virtualization for Performance Testing (12.53) Page 7 of 137

8 What's new This What's New provides an overview of the features that were introduced or enhanced in HPE Network Virtualization for Performance Testing for LoadRunner and Performance Center NV Analytics Report You can now generate a single report for HTTP/HTTPS traffic via the Controller. This report covers all of the selected scenarios in a Performance Center test. You choose scenarios in the Controller by selecting the new NV Insights check box. Learn more This comprehensive report includes: Dashboard showing summary information for the test and each of its scripts. The impact of different networks on the transaction duration is clearly illustrated, as are optimization grades that include links to optimization recommendations for your applications. Visual comparisons of transaction results time grouped by transaction or virtual location Advanced filtering capabilities Detailed results for each transaction and virtual location Highlighting for resources that require optimization Support for the following protocols: TruClient Web Web HTTP/HTML Flex SAP Web Siebel Web protocols Updated VuGen Report A separate report is available for each transaction run via VuGen. These reports include: New look and feel Advanced filtering capabilities Highlighting for resources that require optimization Optimization recommendations are now also available from HTTP Analysis page Support added for the TruClient Web protocol New NV-related runtime graphs Virtual location details are displayed when the test finishes, including: Average throughput (bytes/second) statistics Transaction summary details HPE Network Virtualization for Performance Testing (12.53) Page 8 of 137

9 What's new HPE AutoPass Licensing LoadRunner and Performance Center now directly manage AutoPass licensing for Network Virtualization for Performance Testing Vusers. Purchased license NV licenses are managed by the LoadRunner or Performance Center license utility NV Insights report - full report with optimization recommendations NV Analytics license for Vugen NV Global Library license Community license 2 free NV Vusers NV Insights report First 30 days Full report limited to 2 virtual locations After 30 days Partial report limited to 2 virtual locations NV Analytics report in VuGen First 30 days Full report After 30 days Partial report Network Virtualization for Performance Testing Help Center is available on the Web Documentation is now available both online and as a downloadable.zip file from the Network Virtualization for Performance Testing Help Center. To download an offline copy of the Network Virtualization for Performance Testing Help Center, click Download Help Center from the Explore footer at the bottom of the home page. Microsoft Windows Support was added for Windows bit support was added. Browser support Support was added for Microsoft Edge. HPE Network Virtualization for Performance Testing (12.53) Page 9 of 137

10 What's new HPE Network Virtualization for Performance Testing (12.53) Page 10 of 137

11 Chapter 1: Introducing Network Virtualization for Performance Testing HPE Network Virtualization for Performance Testing (NV) helps to improve the accuracy of your LoadRunner scenarios and Performance Center tests. NV achieves this by enabling you to emulate real-world network conditions in your testing environment. While the emulated network conditions typically include latency, packet loss, and bandwidth, various other network conditions can be emulated if required. By introducing these network conditions, a test can more accurately emulate the real-world network conditions. This, in turn, enables you to more accurately emulate the actual deployment of your system or application in the real world. You install Network Virtualization for Performance Testing during (or after) the LoadRunner or Performance Center installations. You can find more information on how to implement network virtualization in LoadRunner scenarios and Performance Center tests in the LoadRunner Help Center and the PC Help section in the ALM Help Center. HPE Network Virtualization How can we help? Use this Guide to get help on Network Virtualization that is integrated with LoadRunner and Performance Center. For information on the Network Virtualization standalone application, or any other Network Virtualization assistance, go to the HPE Network Virtualization Help Center ( HPE Network Virtualization - Standalone HPE Network Virtualization is a network emulation software solution that enables your organization to test application performance under a variety of current and potential network conditions. This is an ideal solution for your development, testing, and networking teams, as well as any users who need to HPE Network Virtualization for Performance Testing (12.53) Page 11 of 137

12 Chapter 1: Introducing Network Virtualization for Performance Testing quickly and accurately understand whether their code or application is ready for the network and remote end users. Network Virtualization is a standalone product for Windows and Linux systems, enabling you to: test both web and mobile applications automate testing by integrating with HPE Mobile Center and other HPE tools including Unified Functional Testing (UFT) and Service Virtualization (SV) use Network Virtualization in bridge/router mode to impair traffic between servers (vappliance) fully integrate Network Virtualization within a CI process using the NV Jenkins plugin HPE Network Virtualization - LR/PC Integration Version HPE Network Virtualization for Performance Testing integrates seamlessly into HPE LoadRunner and HPE Performance Center software to improve the accuracy of software testing by incorporating realworld network conditions into your load and performance testing processes. HPE Network Virtualization for Performance Testing (12.53) Page 12 of 137

13 Chapter 2: Network Virtualization for Performance Testing Installation The installation process includes these steps: This section includes: System requirements 14 Install HPE Network Virtualization for Performance Testing 14 Upgrading from a previous version 14 Installing Network Virtualization for Performance Testing 15 Manual installation 15 Setup wizard options 16 More NV applications 17 Installation logs 18 Silent installation of HPE Network Virtualization for Performance Testing 18 Silently install NV for Controller 19 Silently install NV for Performance Center 21 Silently install NV for Load Generator 22 Silently uninstall HPE Network Virtualization for Performance Testing 23 Post-installation tasks 24 NV log files 24 HPE Network Virtualization for Performance Testing (12.53) Page 13 of 137

14 Chapter 2: Network Virtualization for Performance Testing Installation System requirements To install Network Virtualization for Performance Testing components, LoadRunner or Performance Center version or later must be installed. Basic system requirements Same system requirements as LoadRunner or Performance Center, unless otherwise specified below. Network Virtualization for Performance Testing-specific system requirements The minimum requirements for running HPE NV components on HPE Load Generator are as follows: Processor Memory (RAM) Quad core processor or higher 4 GB RAM or more HPE Load Generator HPE Load Generator for LoadRunner or Performance Center Third-party component prerequisites Microsoft.NET Framework 4.0 Full or later LoadRunner. The following installations automatically install the minimum required version of a thirdparty component if not already installed on the host: LoadRunner full installation Standalone Load Generator VuGen Performance Center. The following is automatically installed during installation of the HPE NV for Performance Center Server component if not already installed on the host: Microsoft.NET Framework Full Install HPE Network Virtualization for Performance Testing To gain access to the HPE Network Virtualization for Performance Testing (NV) functionality in either LoadRunner or Performance Center, you must install the required NV components. Upgrading from a previous version If a previous version of HPE Network Virtualization for Performance Testing is installed on the machine, you must uninstall all NV components before you can install NV To keep existing results from your current version of NV, during the uninstallation select the option to save existing data. HPE Network Virtualization for Performance Testing (12.53) Page 14 of 137

15 Chapter 2: Network Virtualization for Performance Testing Installation To uninstall: From the Control Panel > Programs and Features, right-click the NV component and click Uninstall. Installing Network Virtualization for Performance Testing At the end of LoadRunner or Performance Center installation wizards, you are prompted to install Network Virtualization for Performance Testing. You can continue with the NV installation, or choose to install manually at a later time. If you install NV, follow the on-screen instructions. You can select one of the following installation modes: Typical. Installs the component with default options. Custom. Enables you to configure several options before installation. For details, see "Setup wizard options" on the next page. The following components are installed: NV component name HPE NV for Controller HPE NV for Performance Center Server HPE NV for Load Generator Installed with: LoadRunner full installation Performance Center Host installation Performance Center Server installation Load Generator (standalone) installation LoadRunner full installation Performance Center Host installation VuGen installation Manual installation If you do not install the NV components as part of the LoadRunner or Performance Center installations, you can run the NV installations manually at any time. Prerequisite: The appropriate LoadRunner or Performance Center components, version or later, must be installed before you can install the NV components. The installation files are located in the following folders: LoadRunner: <LoadRunner installation DVD>\Additional Components\HP NV\ Performance Center: <Performance Center installation DVD>\AdditionalComponents\HPNV\ HPE Network Virtualization for Performance Testing (12.53) Page 15 of 137

16 Chapter 2: Network Virtualization for Performance Testing Installation Run the installation files, according to your needs: For this LoadRunner or Performance Center component... You need to install NV installation file LoadRunner full installation Load Generator (standalone) VuGen NV for Controller NV for Load Generator NV for Load Generator NV for Load Generator (Required because it contains NV Analytics reports, which are integrated with VuGen) NV4HPControllerSetup.exe NV4HPLGSetup.exe NV4HPLGSetup.exe NV4HPLGSetup.exe Performance Center Server NV for Performance Center Server NV4HPPCSetup.exe Performance Center Host NV for Controller NV for Load Generator NV4HPControllerSetup.exe NV4HPLGSetup.exe Note: You must restart the machine after installing or uninstalling NV for Load Generator. Setup wizard options If you run a custom installation during the setup wizard, or are installing the NV components manually, you can configure several options. Some options are enabled only when installing the first NV component on a machine. Option Destination Folder Data Folder NV component Enter a location on the file system to install the component or accept the default location. Default: C:\Program Files (x86)\hp\nv Stores temporary internal application data and user data, such as test run results. Default: C:\Data HPE Network Virtualization for Performance Testing (12.53) Page 16 of 137

17 Chapter 2: Network Virtualization for Performance Testing Installation Option Server Port Enable Remote Connections NV component The port that will be used for communication between various NV components. There is typically no need to change the default value. If you select to Enable Remote Connections, NV adds an exception to the Windows Firewall, enabling remote access to web-based NV components, using the specified server port. Default: 8182 Adds an exception to Windows Firewall. Required for accessing NV statistics and NV Network Editor from remote machines, or to access the NV License Manager remotely. More NV applications You can install the following optional HPE Network Virtualization for Performance Testing components on any machine that meets the system requirements. For installation file locations, see "Manual installation" on page 15. NV application Description NV installation file NV Analytics Provides in-depth analysis of your application s performance, based on test results from tests that use NV functionality. Helps to identify factors that negatively impact your application's performance. Analytics_setup.exe Note: You can now generate NV Insights and NV Analytics reports directly during load tests, and from VuGen. If you choose to continue using the standalone NV Analytics application, you can install the updated version, included with the NV installation files. Make sure to install it on the machine where the NV Analytics for VuGen license is installed. The standalone NV Analytics application uses the same license. HPE Network Virtualization for Performance Testing (12.53) Page 17 of 137

18 Chapter 2: Network Virtualization for Performance Testing Installation NV application Description NV installation file NV Predictor Analyzes test results from LoadRunner, and generates automated reports for SLO (service level objective) compliance. Evaluates results of one or more key metrics for SLO compliance by virtual location based on static values and/or baseline performance. The customizable reports produce a comprehensive analysis of application performance by virtual location. NVPredictorSetup.exe Installation logs Installation logs are located under C:\HP Log. Log files are named as follows: <Product name>_<date>_<time>.log Example: HPE NV for Load Generator_ _ log HPE NV for Controller_ _ log Silent installation of HPE Network Virtualization for Performance Testing You can silently install the HPE NV components for LoadRunner and Performance Center. Note: During a silent installation of LoadRunner or Performance Center, the appropriate NV components are installed by default. For more information, refer to the LoadRunner or Performance Center documentation. Consider before you begin: LoadRunner: When performing a silent installation of NV for Controller, you must be logged in as an Administrator. Make sure that all system requirements are met before beginning the installation. For details, see "System requirements" on page 14. You will not be notified if the silent installation fails. The installation log is located under C:\HP Log. Log files are named as follows: HPE Network Virtualization for Performance Testing (12.53) Page 18 of 137

19 Chapter 2: Network Virtualization for Performance Testing Installation <Product name>_<date>_<time>.log Example: HPE NV for Load Generator_ _ log HPE NV for Controller_ _ log The installation files are located in the following folders: LoadRunner: <LoadRunner installation DVD>\Additional Components\HP NV\ Performance Center: <Performance Center installation DVD>\AdditionalComponents\HPNV\ For this LoadRunner or Performance Center component... You need to install NV installation file LoadRunner full installation Load Generator (standalone) VuGen NV for Controller NV for Load Generator NV for Load Generator NV for Load Generator (Required because it contains NV Analytics reports, which are integrated with VuGen) NV4HPControllerSetup.exe NV4HPLGSetup.exe NV4HPLGSetup.exe NV4HPLGSetup.exe Performance Center Server NV for Performance Center Server NV4HPPCSetup.exe Performance Center Host NV for Controller NV for Load Generator NV4HPControllerSetup.exe NV4HPLGSetup.exe Note: You must restart the machine after installing or uninstalling NV for Load Generator. Silently install NV for Controller 1. Copy the NV4HPControllerSetup.exe file to a convenient location on the LoadRunner or Performance Center Host machine. 2. Run a command prompt with elevated privileges. From the Windows Start menu, click All Programs > Accessories, right-click Command Prompt, and then click Run as administrator. 3. In the Command Prompt window, navigate to the location of the file copied in step 1, and enter the following command together with the required command line options: HPE Network Virtualization for Performance Testing (12.53) Page 19 of 137

20 Chapter 2: Network Virtualization for Performance Testing Installation NV4HPControllerSetup.exe /s /v"/qn <command_line_options>" Command line options: Option Description Required/Optional Default value PORT=<port number> The port used to connect to NV for Controller. REQUIRED 8182 INSTALLDIR=\"<path to installation folder>\" The location where the application files will be installed. OPTIONAL C:\Program Files (x86) \HP\NV ENABLE_ REMOTE=<TRUE FALSE> Opens the port in the firewall. Required for enabling remote access to webbased NV components. OPTIONAL TRUE REBOOT_IF_ NEED=<TRUE FALSE> If a reboot is required, automatically reboots the computer after installation completes. OPTIONAL FALSE Even if the FALSE value is used, you will need to reboot the machine before NV for Controller is operational. DATA_ FOLDER="<path to data dir>" The location where temporary internal application data and user data is saved. OPTIONAL C:\%ProgramData%\HP\NV HPE Network Virtualization for Performance Testing (12.53) Page 20 of 137

21 Chapter 2: Network Virtualization for Performance Testing Installation Example: NV4HPControllerSetup.exe /s /v"/qn PORT=8182" Silently install NV for Performance Center 1. Copy the NV4HPPCSetup.exe file to a convenient location on the Performance Center Server machine. 2. Run a command prompt with elevated privileges. From the Windows Start menu, click All Programs > Accessories, right-click Command Prompt, and then click Run as administrator. 3. In the Command Prompt window, navigate to the location of the file copied in step 1, and enter the following command together with the required command line options: NV4HPPCSetup.exe /s /v"/qn <command_line_options>" Command line options: Option Description Required/Optional Default value PORT=<port number> The port used to connect to NV for Performance Center. REQUIRED 8182 INSTALLDIR=\"<path to installation folder>\" The location where the application files will be installed. OPTIONAL C:\Program Files (x86) \HP\NV ENABLE_ REMOTE=<TRUE FALSE> Opens the port in the firewall. Required for enabling remote access to webbased NV components. OPTIONAL TRUE HPE Network Virtualization for Performance Testing (12.53) Page 21 of 137

22 Chapter 2: Network Virtualization for Performance Testing Installation Option Description Required/Optional Default value REBOOT_IF_ NEED=<TRUE FALSE> If a reboot is required, automatically reboots the computer after installation completes. OPTIONAL FALSE Even if the FALSE value is used, you will need to reboot the machine before NV for Performance Center is operational. DATA_ FOLDER="<path to data dir>" The location where temporary internal application data and user data is saved. OPTIONAL C:\%ProgramData%\HP\NV Example: NV4HPPCSetup.exe /s /v"/qn PORT=8182" Silently install NV for Load Generator 1. Copy the NV4HPLGSetup.exe file to a convenient location on the Load Generator machine. 2. Run a command prompt with elevated privileges. From the Windows Start menu, click All Programs > Accessories, right-click Command Prompt, and then click Run as administrator. 3. In the Command Prompt window, navigate to the location of the file copied in step 1, and enter the following command together with the required command line options: NV4HPLGSetup.exe /s /v"/qn <command_line_options>" Command line options: Option Description Required/Optional Default value PORT=<port number> The port used to connect to NV for Load Generator. REQUIRED 8182 HPE Network Virtualization for Performance Testing (12.53) Page 22 of 137

23 Chapter 2: Network Virtualization for Performance Testing Installation Option Description Required/Optional Default value INSTALLDIR=\"<path to installation folder>\" The location where the application files will be installed. OPTIONAL C:\Program Files (x86) \HP\NV ENABLE_ REMOTE=<TRUE FALSE> Opens the port in the firewall. Required for remote access to the NV statistics and the NV Network Editor. OPTIONAL TRUE REBOOT_IF_ NEED=<TRUE FALSE> If a reboot is required, automatically reboots the computer after installation completes. OPTIONAL FALSE Even if the FALSE value is used, you will need to reboot the machine before NV for Load Generator is operational. DATA_ FOLDER="<path to data dir>" The location where temporary internal application data and user data is saved. OPTIONAL C:\%ProgramData%\HP\NV Example: NV4HPLGSetup.exe /s /v"/qn PORT=8182" Silently uninstall HPE Network Virtualization for Performance Testing To uninstall an component silently, you must be logged in as administrator. HPE Network Virtualization for Performance Testing (12.53) Page 23 of 137

24 Chapter 2: Network Virtualization for Performance Testing Installation At the command line, enter the command for the component you want to uninstall: NV component NV for Controller NV for Performance Center NV for Load Generator Uninstall command NV4HPControllerSetup.exe /s /removeonly /v"/qn PORT=<port number> REBOOT_IF_NEED=<TRUE FALSE>" PORT - optional; removes the port from the firewall NV4HPPCSetup.exe /s /removeonly /v"/qn PORT=<port number> REBOOT_IF_ NEED=<TRUE FALSE>" PORT - optional; removes the port from the firewall NV4HPLGSetup.exe /s /removeonly /v"/qn PORT=<port number> REBOOT_IF_ NEED=<TRUE FALSE>" PORT - optional; removes the port from the firewall Note: You must restart the machine after installation or uninstallation of the NV for Load Generator component. Post-installation tasks After installing NV components, you need to: install licenses configure the active adapter optionally configure addition NV customization For details, see the relevant sections of the HPE Network Virtualization for Performance Testing User Guide. Note: When you set up the LoadRunner/Performance Center testing environment, make sure that the LoadRunner Controller and the Performance Center Host can use port 8182 (default) to communicate with all load generators that will run NV. If this communication is not enabled, the LoadRunner Controller, Performance Center client, and Analysis will not be able to display the "Network Virtualization Charts". NV log files HPE Network Virtualization for Performance Testing products log files are located in the <installation directory>\logs. By default, in \Program Files\HP\NV\logs or \Program Files (x86)\hp\nv\logs. HPE Network Virtualization for Performance Testing (12.53) Page 24 of 137

25 Chapter 3: Licensing Network Virtualization for Performance Testing requires a license. Community license Community licenses are automatically included in a LoadRunner or Performance Center installation. The Community license includes: NV Vusers 2 free NV Vusers. Enables you to run tests that include 2 NV Vusers. NV Insights report Enables you to generate the NV Insights report during LoadRunner / Performance Center tests for up to 2 NV virtual locations. First 30 days Full report After 30 days Partial report The 30 day period begins the first time you use the feature. NV Analytics report in VuGen Enables you to generate the NV Analytics report in VuGen. First 30 days Full report After 30 days Partial report The 30 day period begins the first time you use the feature. Purchased license A purchased license provides access to the following functionality: NV Vusers Enables you to run tests using the number of purchased NV Vusers. NV Insights report Enables you to generate the full NV Insights report during LoadRunner / Performance Center tests. NV Analytics report in VuGen NV Global Library Enables you to generate the full NV Analytics report from VuGen. Enables you to import network profiles from the NV Global Library of world-wide network conditions. Note: You can now generate NV Insights and NV Analytics reports directly during load tests, and from VuGen. If for some reason you are using the standalone NV Analytics application, you can install the updated version, included with the NV installation files. Make sure to install it HPE Network Virtualization for Performance Testing (12.53) Page 25 of 137

26 Chapter 3: Licensing on the machine where the NV Analytics for VuGen license is installed. The standalone NV Analytics application uses the same license. Next steps: Learn how to install a license for NV Vusers, NV Analytics / NV Insights, and NV Global Library. (Done via the LoadRunner or Performance Center license utility) "Install an NV Analytics license for VuGen" below Install an NV Analytics license for VuGen 1. Open the NV License Manager on the computer on which VuGen is installed. From the Windows Start menu, select All Programs > HPE Software > HPE NV for Performance Testing > NV License Manager. In Windows 8.x or higher, you can access NV License Manager from the Start or Apps screen. 2. Click Update license. 3. Click Update via > File or key. 4. Copy the machine code displayed below. 5. Click HPE license portal to connect to the HPE licensing site, and activate your license. You will need to enter the machine code that you previously copied, and then generate a license file. 6. In NV License Manager, click the folder icon that appears to the right of the License File box to locate and upload the license file. 7. Click Update. The updated license details are displayed in the NV License Manager main page. HPE Network Virtualization for Performance Testing (12.53) Page 26 of 137

27 Chapter 4: Configure NV profiles with NV Network Editor NV Network Editor lets you create complex NV profiles. Thereafter, you import the parameters from the profile into virtual locations, inside LoadRunner or Performance Center. To learn more about NV profiles, see "NV Network Editor overview" below. NV Network Editor overview You use the NV Network Editor to create NV profiles that you can import to your LoadRunner scenarios and Performance Center tests. The Virtual Location Editor enables you to create virtual locations inside LoadRunner or Performance Center that include only latency, packet loss, and bandwidth parameters. Alternatively, you can import network emulation parameters from the NV Global Library to create a virtual location. If a virtual location requires additional emulation parameters, you can import these parameters from an NV profile using the Virtual Location Editor. You use NV Network Editor to create these NV profiles. This section describes how to use the Network Editor. For a list of NV profile parameters, see "NV profile parameters" on page 38. For details on using the Virtual Location Editor, see "Add and configure virtual locations" on page 41. Note: Each NV profile file that you create with NV Network Editor can include multiple flows. A flow is a set of parameters that defines a network. When you import parameters into LoadRunner or Performance Center, you specify from which flow to import the parameters. The NV Network Editor is installed as part of the HPE NV for Load Generator installation NV Network Editor is installed automatically each time HPE NV for Load Generator is installed. For details, see "Network Virtualization for Performance Testing Installation" on page 13. See also: "Open NV Network Editor" below "Create NV profiles" on the next page Open NV Network Editor Network Editor is a browser-based application that can be opened from an HPE NV for Load Generator host, or from a browser on any other computer. To open the Network Editor from an HPE NV for Load Generator host machine Select Start > All Programs > HP Software > HP Network Virtualization > HP NV for Load Generator > NV Network Editor. HPE Network Virtualization for Performance Testing (12.53) Page 27 of 137

28 Chapter 4: Configure NV profiles with NV Network Editor In Windows 8.x or higher, you can access Network Editor from the Start or Apps screen. To open the Network Editor from a browser Type the following: NV for Load Generator host IP>:<HPE NV for Load Generator Port>/Shunra/networkeditor/ Example: The HPE NV for Load Generator Port is set initially when HPE NV for Load Generator is installed. Port 8182 is the default port. Create NV profiles 1. Open NV Network Editor, and on the home page, click New. 2. Click Add new flow. 3. In the New Flow page, specify the required parameters as described below: a. Setting Client and Server IP Ranges: In order to be able to save the NV profile, it is necessary to enter at least a single client IP address range and a single server IP address range. Note: These IP addresses are not used by LoadRunner or Performance Center, so it is not necessary to specify actual addresses. b. Configuring Gateway Parameters: For details, see "Configure gateway parameters" on the next page. c. Configuring the WAN Shape: For details, see "Configure the WAN parameters" on page To enable packet capture, click Packet List, and select Enable packet list capture. HPE Network Virtualization for Performance Testing (12.53) Page 28 of 137

29 Chapter 4: Configure NV profiles with NV Network Editor 5. After selecting the required parameters, click OK. The Flow Summary displays all the flows in the NV profiles. To edit a flow, click the pencil icon; to delete a flow, click the "x"; to delete all flows, click the table icon. 6. Click Save to save the NV profile to a file called (by default) "HP Emulation Profile.ntxx," located in the browser's default download folder. What's next? You can now import the NV profile file into LoadRunner or Performance Center. For details, see "Import virtual location parameters from an NV profile" on page 51. Configure gateway parameters When you use NV Network Editor to create an NV profile, you can specify gateway parameters, as described in this section, and WAN parameters, as described in "Configure the WAN parameters" on page 32. When network virtualization is applied, gateway parameters are used to emulate the behavior of access gateways on the real networks. For example, in the NV profile, you can specify the bandwidth or queue limitation that a specific gateway imposes on the actual network. HPE Network Virtualization for Performance Testing (12.53) Page 29 of 137

30 Chapter 4: Configure NV profiles with NV Network Editor Note: Entering gateway parameters is optional when you configure an NV profile. The definitions of the gateway parameters are the same for both the client and the server gateways. The gateway parameters can be configured separately for incoming and outgoing traffic. Click the Client Gateway and Server Gateway tabs to configure the gateway parameter settings. The gateway settings include the following: Bandwidth Settings Queue Settings Bandwidth Settings Bandwidth values can be either defined or imported from a pre-recorded file. You can specify different bandwidth values for upstream network traffic and downstream network traffic. Note: The following procedures are valid for both the upstream and downstream settings. Define Bandwidth Settings You can select pre-defined bandwidth settings, or specify custom bandwidth settings. 1. In the Client Gateway or the Server Gateway tab, select one of the following options: a. Predefined: Select one of the available values. (The available values correspond to common bandwidth settings, such as T1, etc.) b. Custom: Specify a specific bandwidth. (Valid values are between 2.4 and 10,000,000 Kbps.) 2. Click OK. Import Recorded Bandwidth Settings You can import the bandwidth values from a recorded file. Use HPE Network Capture or HPE Network Capture Express to record the bandwidth of a network over a period of time. Network Capture and Network Capture Express measure and record the bandwidth, and save the values to an.ntx file. Note: Importing recorded bandwidth settings is supported only for Vuser groups that share the specified bandwidths not by Vuser groups that allocate the specified bandwidth to each Vuser. For details, see "Allocate bandwidth to Vusers" on page 54. HPE Network Virtualization for Performance Testing (12.53) Page 30 of 137

31 Chapter 4: Configure NV profiles with NV Network Editor You can import recorded parameters from a.ntx file directly into a virtual location, as described in "Import virtual location parameters from an NV profile" on page 51. Import the recorded parameters into an NV profile (using NV Network Editor) only if you want to edit the recorded parameters, before importing them into a virtual location. 1. In the Client Gateway or the Server Gateway tab, click Recorded then click Choose File. 2. Locate and select the required file, and then click Open. Queue Settings The Queue settings define the queue limitations and packet overhead that you want to use in your emulations. Enable Queue Limitation This group contains parameters that emulate limitations imposed on the maximum size of IP packet queues at the gateway NIC. If you specify a queue size, HPE NV emulates network behavior by dropping data packets when the queue is full. 1. In the Client Gateway or the Server Gateway tab, select Queue Limitation. 2. Select from the following parameters: Parameter Queue Size Description Select the amount of memory (KB) that HPE NV allocates to the queue [16 to 4,096 MB]. Drop Mode Drop Tail Random Early Detectio n (RED) Select this option for HPE NV to drop newer data packets when the queue is full. This option provides a more sophisticated queue management method. The RED algorithm keeps track of the average increase in queue occupancy. If it detects an increase, it signals to the packet source that the queue may soon be full, by randomly dropping very small amounts of data packets. Keep Queue size between: Indicate the minimum and maximum average occupancy. The system issues signals (by dropping data packets) when the average occupancy fluctuates between the two values. It drops all packets if the average occupancy exceeds the maximum threshold. Fill mode Use the fixed size: 1,500 bytes for each packet. Use the actual packet size. HPE Network Virtualization for Performance Testing (12.53) Page 31 of 137

32 Chapter 4: Configure NV profiles with NV Network Editor Enable Packet Overhead If your gateway adds overhead bytes to the IP data that flows through the gateway NIC, use this parameter to specify the number of additional bytes, in order to emulate the bandwidth consumption they impose on the network. 1. Select Packet Overhead. 2. Select either Ethernet (adds 18 bytes to each packet) or PPP (adds 9 bytes to each packet). Configure the WAN parameters When you use NV Network Editor to create NV profiles, you can define the following WAN parameters: Latency and Packet Loss: The latency defines the time it takes an IP packet to cross the WAN (oneway); the packet loss determines how many packets are dropped. For details, see "Latency and packet loss [WAN Parameters]" on the next page. Packet Effects: Dynamic IP packet routing effects, including out-of-order packets (reordering), duplicated packets, and packet fragmentation. For details, see "Packet effects [WAN Parameters]" on page 35. Link Faults: Possible damage to bit streams, network disconnections, and network congestion. For details, see "Link faults [WAN Parameters]" on page 37. Note: You can manually specify the latency and packet loss parameters, or you can import the parameters from an.ntx file. Packet Effects and Link Faults are always set manually, i.e. they cannot be imported from a file. For details on how to set the client and server gateway parameters, see "Configure gateway parameters" on page 29. HPE Network Virtualization for Performance Testing (12.53) Page 32 of 137

33 Chapter 4: Configure NV profiles with NV Network Editor Latency and packet loss [WAN Parameters] Latency is the time it takes an IP packet to cross the WAN. You can define a fixed latency or a statistically distributed latency. Set latency and packet loss parameters 1. Click the WAN tab, and then select Latency and Packet Loss. 2. Select: Custom: To set the parameters manually. For details, see Specifying Custom Latency and Packet Loss. Recorded: To import a recorded file (.ntx) produced by HP Network Capture or HP Network Capture Express. For details, see Importing network conditions from a recorded file below. Specify Custom Latency and Packet Loss 1. To define the settings manually, click Custom. 2. To set the latency values, select one of the following options: Fixed: Enter the number of milliseconds, from 0 to 8,000. Uniform distribution: Specifies a changing latency over time (also referred to as jitter). Type the minimum and maximum latency values in the Minimum and Maximum boxes. This causes NV to randomly change the latency between the minimum and maximum values (from 0 to 8,000 ms). You can limit the change in latency between each two consecutive packets to a specified number. To use this option, select Limit latency change and enter the maximum allowed change, in milliseconds. For example, if you enter 45, NV will change the latency by a maximum of 45 milliseconds between two consecutive packets. Normal distribution: Causes latency to fluctuate randomly and non-uniformly around an average value (milliseconds) that you specify in the Average box. This option can be used to emulate jitter conditions. The Standard Deviation parameter allows you to exert some control over the random change in latency. Entering a value in this parameter sets a range for two thirds of the random values. For example, if the average is 600 milliseconds and the standard deviation is 100 milliseconds, two thirds of random latency values will be between 500 and 700. Linear Latency: Allows you to set a range of latency values and to indicate the time it takes latency to increase from the minimum to the maximum value in the range. HPE Network Virtualization for Performance Testing (12.53) Page 33 of 137

34 Chapter 4: Configure NV profiles with NV Network Editor Enter the range in Minimum and Maximum (from 0 to 8,000 ms), and the cycle duration in Graph duration (from 1 to 65,535 seconds). For example, if you define a range of 100 to 200 milliseconds and a cycle duration of 100 seconds, NV will increase latency by one millisecond each second. When the NV reaches maximum latency, it cycles back to the minimum value. 3. To add packet loss parameters into an NV profile, select a Packet Loss option: Note: Packet loss can also be imported in a recorded file. No packet loss: WAN behavior is not affected by packet loss. In this case, HPE NV does not lose any packets and the impairment is not applied to network traffic. Periodic loss: HPE NV discards one packet for every xth packet that passes through the WAN. Enter a number (from 2 to 65,535) in the Lose one packet every x packets box. For example, if you enter 8, NV discards every 8th packet that enters the WAN. HPE NV counts packets separately in both directions. It loses the incoming xth packet and the outgoing xth packet. Random Loss: Allows you to set the probability for losing each packet (enter a percentage from 0.01 to 90). For example, entering 2 means that each packet has a 2 percent chance of being dropped while passing through the emulated WAN cloud. Burst Loss: Allows you to introduce a "bursty" packet loss model in the emulated WAN cloud, by setting a number of packets to lose (burst size) every time a loss event occurs. You can specify the probability of the loss event (Burst Probability), and the Minimum and Maximum parameters (from 1 to 65,535 packets) that specify a range for the random size of each packet loss burst. Gilbert-Eliot Loss: Allows you to emulate a good state and a bad state in the network by specifying an average packet loss percentage in the Lose parameter of the two states (from 0.01% to 100%). In Change state, you specify the chances that the network will leave the specific state and move to the other (from 0.01% to 99.99%). For example, if the tested network has, at any given time, a 10% chance of moving from good to bad and a 50% chance of moving from bad to good, you specify 10 and 50 in the respective Change State fields. Import network conditions from a recorded file Note: Importing recorded network conditions is supported only for Vuser groups that share the specified bandwidths not by Vuser groups that allocate the specified bandwidth to each Vuser. For details, see "Allocate bandwidth to Vusers" on page 54. You can import recorded parameters from a.ntx file directly into a virtual location, as described in "Import virtual location parameters from an NV profile" on page 51. Import the HPE Network Virtualization for Performance Testing (12.53) Page 34 of 137

35 Chapter 4: Configure NV profiles with NV Network Editor recorded parameters into an NV profile (using NV Network Editor) only if you want to edit the recorded parameters, before importing them into a virtual location. 1. Select Recorded and then click Choose file. 2. Locate and select the required.ntx file, and then click Open. A summary of the network conditions is displayed. 3. Modify the latency and packet loss by selecting a percentage in the Multiply latency values by and Multiply loss value by lists. This factor allows you to predict future load cases or to prepare some margins of service level over the current status. In this field, you can specify a number between 0 and 200 percent to have the latency and packet loss values multiplied by it. For example, the value of 100% means that NV will use the recorded values as they are. The value 200 means that each value will be doubled (multiplied by 200%). 4. Select a Packet Loss value. 5. Select a Playback Mode. Play imported values sequentially Play imported values randomly While emulating the network, HPE NV will cyclically use the latency values one after the other in the order in which they were recorded. While emulating the network, HPE NV will pick one of the recorded values randomly, for each packet. This allows you to experience random values from the real network to predict more network scenarios. Packet effects [WAN Parameters] Emulate IP routing effects such as disruption of packet order (reordering), packet duplication, and fragmentation. Select the effect you want to emulate, and then configure the associated parameters: Out of Order To emulate packet reordering, HPE NV can generate an out-of-order event, based on the probability set in the Chance parameter (from 1% to 50%), by randomly removing a packet from the data stream. HPE Network Virtualization for Performance Testing (12.53) Page 35 of 137

36 Chapter 4: Configure NV profiles with NV Network Editor Then it starts counting the incoming packets. It returns the removed packet after the nth packet has entered. The nth packet is picked randomly from the range that you indicate in the Maximum and Minimum (Packet offset from original location) parameters (from 1 to 64 packets). For example, if you indicate a range of 5 to 10, HPE NV will return the removed packet randomly to the data stream (after the 5th, or 6th or 10th packet that follows the removed one). In order to have HPE NV return all removed packets after a predefined number of incoming packets, the same value can be used for both Minimum and Maximum offset. Duplicate Packets HPE NV emulates duplication by copying a packet that it selects randomly. The number of copies that will be created when the event occurs is specified in the Minimum and Maximum parameters (from 1 to 20 packets). For example, if you specify a range of 2-4 packets, HPE NV duplicates a packet 2, 3, or 4 times (randomly), when it decides to do so (according to the probability specified in Chance from 1% to 99%). To create a pre-defined number of copies, enter the same value in both the Minimum and Maximum parameters. For example, entering 3 in both parameters causes HPE NV to create 3 copies of the packet when the event occurs. Fragmentation To emulate the packet fragmentation effect you need to set a packet size, in bytes, in the Maximum Transmission Unit parameter (from 64 to 1,460 bytes). This would be the maximum size a gateway along the path route would allow (MTU). Bigger packets are likely to be fragmented or discarded when the fragmentation event occurs (according to the probability set in Chance from 1% to 99%). Whether or not the packets will be eventually fragmented depends on the state of the Do Not Fragment (DF) bit in the packet's IP header and on the policy you define via the options included in the Fragmentation group. If the flag is OFF, the packet is fragmented anyway. If the flag is set to ON, the policy you select applies. A description of the DF=ON policies is as follows: Option Ignore Do Not Fragment (DF) bit Generate ICMP Error Messages to Source Do Not Generate ICMP Error Messages Effect Fragment the packet. Discard the packet and inform the packet's source, using an ICMP packet. Discard the packet without informing the packet's source. HPE Network Virtualization for Performance Testing (12.53) Page 36 of 137

37 Chapter 4: Configure NV profiles with NV Network Editor Link faults [WAN Parameters] Link faults consist of effects caused by physical link malfunctions, specifically bit errors and disconnections. HPE NV emulates bit errors by toggling bits at a given frequency. To achieve "disconnection" in the WAN Cloud, you can tell HPE NV to stop responding for a given period. Click the Link Faults tab to set the parameters that help emulate the link faults. You can emulate any of the link faults or all of them. Select the check box of the fault type that you want to emulate, and then enter the parameters described below: Bit Error In the Average Frequency box, type a number of bits (from 10 2 to ). Bit toggling will occur every time that the indicated number of bits (on average) has crossed the WAN Cloud. In the Minimum and Maximum parameters (Number of toggled bits), enter a range of numbers from 1 to 500. HPE NV randomly picks a number in the range and toggles as many bits. Disconnect The parameters in this group allow you to emulate a physical disconnection of the network. In the Average Frequency box type the desired number of seconds (from 3 to 300). HPE NV will emulate a disconnection once every so many seconds on average. In the Minimum and Maximum boxes (Disconnection time span), indicate a time range in milliseconds (from 10 to 30,000 ms). HPE NV will randomly pick a time value from this range and the lines will remain "disconnected" during this time, by dropping all packets that go through the WAN Cloud during this period. Congestion The Congestion feature helps you emulate a periodic and momentary rise in WAN Cloud traffic, which results in increased latency and packet loss. Note: The Latency and Packet Loss values during the congestion event override the original values (either manually defined or imported). HPE Network Virtualization for Performance Testing (12.53) Page 37 of 137

38 Chapter 4: Configure NV profiles with NV Network Editor Congestion Parameter Average congestion frequency Congestion time span Description Type the frequency in seconds (from 1 to 300). For example, entering 20 means that HPE NV will increase latency and packet loss every 20 seconds on average. The duration of the event in milliseconds (from 10 to 65,535). Type a range in the Maximum and Minimum parameters. HPE NV randomly picks a value from this range and sets the latency and packet loss to the values you indicate in the Fixed Latency and Loss fields (below), for the selected duration. After this time, latency and packet loss return to their original settings (as described under the Latency and Packet Loss branches). Congestion event properties Fixed Latency Lose x% of all packets Enter the latency, in milliseconds, that will prevail during the congestion time (from 0 to 8,000 ms). This value replaces the original latency setting. Enter a percentage of packets to be lost during the congestion time (from 0% to 90%). This value replaces the original setting for packet loss. NV profile parameters You can use NV Network Editor to configure the following parameters in a NV profile. WAN Specifications Latency Fixed Latency Uniform Distributed Latency (minimum/maximum): Latency change limit 0-8,000 ms (Granularity = 1 ms) 0-8,000 ms (Granularity = 1 ms) 0-4,000 ms (Granularity = 1 ms) Normal Distributed Latency: Average Standard Deviation Linear Latency (minimum/maximum): Cycle Duration 0-8,000 ms (Granularity = 1 ms) 0-8,000 ms (Granularity = 1 ms) 0-8,000 ms (Granularity = 1 ms) 1-65,535 sec Packet Loss HPE Network Virtualization for Performance Testing (12.53) Page 38 of 137

39 Chapter 4: Configure NV profiles with NV Network Editor Periodic Loss - Lose every n th packet n=2-65,535 Random Loss 0-100% (Granularity = 0.01%) Burst Loss: Probability % (Granularity = 0.01%) Burst Size 1-65,534 packets Gilbert-Elliot Loss Two State Loss Model: Loss Probability % (Granularity = 0.01%) Transition to Other State % (Granularity = 0.01%) Packet Effects Out Of Order: Probability 1-50% Offset 1-64 packets Packet Duplication: Probability 1-99% Number of duplicates 1-20 packets Fragmentation: Probability 1-99% MTU 64-1,460 bytes DF Policy Ignore DF and fragment packet. Link Faults Bit Error: Discard packet and generate ICMP message. Discard packet and do not send a message. Frequency Number of Toggled Bits One every bits bits Disconnection: HPE Network Virtualization for Performance Testing (12.53) Page 39 of 137

40 Chapter 4: Configure NV profiles with NV Network Editor Average Frequency Disconnection Time sec 10-30,000 ms Congestion: Frequency Time span sec 10-65,535 ms Event Properties: Fixed Latency 0-8,000 ms Packet Loss Probability 0%-90% (Granularity = 0.01%) Gateway Specifications Bandwidth Settings Symmetric Bandwidth 2.4 Kbps - 10 Gbps Asymmetric Bandwidth: Uplink/Downlink Packet Overhead Length 2.4 Kbps - 10 Gbps bytes Queue Limitations Queue Size Drop Mode: RED Queue (Minimum/Maximum Threshold) Fill Mode (Queue Size Policy) 2-65,535 KB Drop Tail or RED 2-65,535 KB Byte Mode or Packet Mode Packet Overhead Ethernet (Add 18 bytes to each packet) PPP (Add 9 bytes to each packet) Traffic Type Support IPv4 HPE Network Virtualization for Performance Testing (12.53) Page 40 of 137

41 Chapter 5: Add and configure virtual locations You create virtual locations in the LoadRunner Controller or Performance Center, and then include them in your LoadRunner scenarios or Performance Center tests. You can also use several predefined virtual locations that NV provides. What do you want to do? "Read more about virtual locations" below "Learn about predefined virtual locations" below "Open the Virtual Location Editor in LoadRunner" on the next page "Open the Virtual Location Editor in Performance Center" on page 43 Read more about virtual locations The characteristics of a network are influenced by a variety of factors. One factor that has a significant effect on the character of a network is the geographic locations of the clients and the servers in the network. For example, consider a server based in New York. This server is accessed by clients that are based in London and in Sydney. One can expect the London-New York network to behave at least somewhat differently from the Sydney-New York network. The difference in behavior between the two networks may be significant. When you implement network virtualization for this system, you add London and Sydney as two virtual locations. For each of these virtual locations, you define the characteristics of the network to emulate the network between the client and the server in New York. When you develop a LoadRunner scenario or a Performance Center test, you specify which virtual location is associated with each Vuser group or with each load generator. The virtual location defines the network conditions that are emulated when the Vusers run. See also: "Enable network virtualization" on page 43 "Set the network virtualization mode" on page 44 Learn about predefined virtual locations NV includes a number of predefined virtual locations. Each predefined virtual location includes parameters that define a common network testing environment. You can select one of these predefined locations when you configure scenarios or tests in LoadRunner or Performance Center. Note: You cannot edit, delete, or rename a predefined location. HPE Network Virtualization for Performance Testing (12.53) Page 41 of 137

42 Chapter 5: Add and configure virtual locations Virtual location name Latency [ms] Packet loss [%] Bandwidth-in [kb/s] Bandwidth-out [kb/s] 2.5G Busy G Busy G Typical G Good , G Good ,000 7,500 AUS to US EC ,000 2,000 (Australia to U.S. East Coast) China to US EC ,000 2,000 India to US EC ,000 2,000 London to US EC ,000 4,000 Satellite 0 0 2,000 1,000 Open the Virtual Location Editor in LoadRunner 1. In the LoadRunner Controller, click the Network Virtualization Settings button on the toolbar to open the Network Virtualization Settings dialog box. 2. Make sure that Enable Network Virtualization is selected. 3. Under Virtual Locations, in the Virtual Location column, type the name of a location. You cannot use the following characters: \ / :? < > * % ^,! { } ( ) ; = # 4. Enter a description for the virtual location. You cannot use the following characters: \ / :? < > * % ^,! { } ( ) ; = # 5. Click Configure to open the Virtual Location Editor. For details on how to proceed with the Virtual Location Editor, see "Configuring virtual locations" on page 48. HPE Network Virtualization for Performance Testing (12.53) Page 42 of 137

43 Chapter 5: Add and configure virtual locations Open the Virtual Location Editor in Performance Center 1. In Performance Center > Performance Test Designer, click Groups & Workload. 2. Click Virtual Location Editor. 3. Make sure that Virtual Location Mode is turned on. 4. In the Add new virtual location box, type the name of the new virtual location. You cannot use the following characters: \ / :? < > * % ^,! { } ( ) ; = # 5. Click Add to open the Virtual Location Editor. For details on how to proceed with the Virtual Location Editor, see "Configuring virtual locations" on page 48. For more details on how to work with virtual locations in LoadRunner and Performance Center, see the LoadRunner Help Center or the PC Help section in the ALM Help Center. What's next? "Select virtual location definition method" on page 48 Enable network virtualization You can enable or disable network virtualization for a LoadRunner scenario or a Performance Center test. When NV is enabled, the network conditions that are specified in the virtual locations are applied to the specified running Vusers. When NV is not enabled, network conditions are not emulated. When you create a new LoadRunner scenario or a Performance Center test, NV is enabled by default. After you enable network virtualization, you must specify the mode that will be used to implement the network virtualization. For details, see "Set the network virtualization mode" on the next page. HPE Network Virtualization for Performance Testing (12.53) Page 43 of 137

44 Chapter 5: Add and configure virtual locations To enable network virtualization for a LoadRunner scenario or a Performance Center test 1. LoadRunner: On the LoadRunner Controller toolbar, click the Network Virtualization Settings button open the Network Virtualization Settings dialog box. Performance Center: On the Performance Center > Workload tab toolbar, click the Network Virtualization Settings button to open the Network Virtualization Settings dialog box. 2. Select Enable Network Virtualization. See also: "Set the network virtualization mode" below Set the network virtualization mode When you implement network virtualization for a LoadRunner scenario or a Performance Center test, you select the mode that LoadRunner or Performance Center uses to implement the virtualization: Per Group. This option performs network virtualization per Vuser group. You specify the NV profile (virtual location) that will be applied to all the Vusers in each Vuser group. You can specify a different virtual location for each Vuser group, even if the Vuser groups are running on the same load generator. Per Load Generator. This option performs network virtualization per load generator. You specify the network emulation profile (virtual location) that will be applied to all Vusers running on each load generator. Vuser protocol limitations The following Vuser protocols do not support the Per Group emulation mode: Citrix ICA, Java Record Replay, Java Vuser, COM/DCOM, MAPI. If you select the Per Group emulation mode, and a particular Vuser group runs a Vuser protocol that does not support the Per Group functionality, the Per Load Generator emulation mode is applied to the Vuser group. When the Vuser group runs, the load generator's default virtual location is applied to the Vusers, and not the virtual location that is specified for the Vuser group. See "Allocate bandwidth to Vusers" on page 54 for other limitations of these Vuser protocols. To specify the NV emulation mode for a LoadRunner scenario or a Performance Center test 1. Make sure that network virtualization is enabled for the test or the scenario. For details, see "Enable network virtualization" on the previous page. to HPE Network Virtualization for Performance Testing (12.53) Page 44 of 137

45 Chapter 5: Add and configure virtual locations 2. Select a method for implementing the network virtualization, Per Group or Per Load generator. See also: "Define virtual location test level settings" below "Add and configure virtual locations" on page 41 Chapter 5: Define virtual location test level settings The NV Virtual Location Test Level Settings apply to all virtual locations and load generators that are included in the test or a scenario. Packet Capture "IP filter" on page 47 After you specify these global settings, you can specify the settings for a particular virtual location or load generator. For details, see "Specify virtual location runtime settings" on page 53. Packet Capture You can configure Network Virtualization for Performance Testing to capture the packets that are transferred across a network while a LoadRunner scenario or Performance Center test runs. Note: Packet capture test level settings apply to all virtual locations and load generators participating in the scenario or test. After you enable packet capture on the global level, you must enable packet capture for particular virtual locations in the scenario or test, as required, and specify the packet capture settings. For details, see "Set packet capture for a virtual location" on page 55. When you configure the Packet Capture common settings, you specify the following: "Packet capture (On/Off)" below "Maximum total packet list size" on the next page "Packet capture method" on the next page Packet capture (On/Off) You can select to capture the packets that are transferred across the network while the scenario or test runs. If you turn on Common/Runtime Settings > Packet Capture, then you can turn on Packet Capture for the virtual locations in the scenario or test. If you turn off Common/Runtime Settings > Packet Capture, then you cannot turn on Packet Capture for the virtual locations in the scenario or test. Note: Make sure to consider the use of disk space on the Load Generator. Depending on your HPE Network Virtualization for Performance Testing (12.53) Page 45 of 137

46 Chapter 5: Add and configure virtual locations packet capture settings, each test can save a packet list of up to 1GB. Consider configuring the NV cleanup threshold to clean the captured packets, or do a manual cleanup as needed. Maximum total packet list size The maximum memory space that is allocated on each load generator in the scenario or test to store captured packets. The default size of the maximum space is 300 MB per load generator. Note: The Total packet list size cannot be more than 25% of the RAM on the load generator. For details on customizing the default maximum size, see "Set the default maximum packet list buffer size for each load generator" on page 62. Packet capture method Defines what happens when packets are captured and the maximum total packet size is reached. Cyclic: When the maximum total packet size is reached, the oldest packets are erased, and the newest packets continue to be recorded. Non-cyclic: When the maximum total packet size is reached, new packets are no longer recorded. To set the common packet capture settings 1. In LoadRunner Controller or Performance Center, display the virtual location test level settings. Show me how for LoadRunner a. In the LoadRunner Controller, click the Show Virtual Location Settings button on the toolbar. b. Make sure that Enable Network Virtualization is selected. c. Click Common Settings. Show me how for Performance Center a. In Performance Center, click the Virtual Location Editor button on the toolbar. b. Make sure that Virtual Location Mode is turned on. c. Click Runtime Settings. 2. Under Packet Capture, click the switch to enable packet capture. 3. In Total packet list size, specify the maximum memory space that is allocated on each load generator in the scenario or test to store captured packets. HPE Network Virtualization for Performance Testing (12.53) Page 46 of 137

47 Chapter 5: Add and configure virtual locations 4. Select the packet capturing method: Cyclic or Non-cyclic. For details, see Packet capture method above. 5. Click OK to save the settings. IP filter While you run a scenario or test that incorporates network virtualization, LoadRunner or Performance Center applies various network conditions to the networks that are used by the Vusers that run during the scenario or test. These network conditions are applied to network traffic to-and-from all load generators that are associated with the virtual locations in the scenario or test. But what if, for some reason, you do not want to apply network virtualization to traffic associated with a specific IP address? You can create IP filters that include lists of IP addresses to which you do not want to apply network virtualization. Network conditions are not applied to network traffic to-or-from a specific IP address if that IP address is included in the relevant IP filter, either as the source or the destination of network traffic. For reasons why you may want to exclude an IP address from network virtualization, see "Exclude machines from network virtualization" on page 57. In LoadRunner and Performance Center, there are two types of IP filters: Global IP filter. A global IP filter is a list of IP addresses to which you do not want to apply network virtualization. The global IP filter applies to all virtual locations in the scenario or test. Network impairments will not be applied to network traffic if an IP address in the global IP filter is either the source or the destination of the network traffic. Local IP filters. You can create a local IP filter for each virtual location in a scenario or test. The local IP filter includes those IP addresses to which you do not want to apply network virtualization. These locally "excluded addresses" are in addition to the IP addresses that are excluded because they appear in the global IP filter. Note: A local IP filter applies only to the virtual location with which the local IP filter is associated. For details on creating a local IP filter, see "Create a local IP filter for a virtual location" on page 56. To maintain a global IP filter: 1. In the LoadRunner Controller or Performance Center, display the test level network virtualization settings, as described in "Define virtual location test level settings" on page Under Exclude these IP's box, type the IP address that you want to exclude, and then click the "+" sign. The IP address is added to the global IP filter. 3. Repeat the previous step for all the IP addresses that you want to exclude. HPE Network Virtualization for Performance Testing (12.53) Page 47 of 137

48 Chapter 5: Add and configure virtual locations Note: IP addresses that are included in the global IP filter are displayed in gray in the local IP filter when you configure the runtime settings for a virtual location. For details, see "Create a local IP filter for a virtual location" on page 56. To remove an IP address from the global IP filter list, select the IP address, and then click "X". 4. Click OK to save the settings. Select virtual location definition method After you create a new virtual location in either the LoadRunner Controller or Performance Center, you use the Virtual Location Editor to configure the virtual location. Virtual location configuration methods The first page in the Virtual Location Editor provides three methods that you can use to configure the NV profile parameters for a virtual location: 1. Custom. Manually specify the network profile parameters. For details, see "Define custom virtual location parameters" on the next page. 2. Global Library. Import the network profile parameters from the NV Global Library. For details, see "Extract virtual location parameters from the NV Global Library" on page Import from a file. Import the network profile parameters from a file. The file may have been produced by HP Network Editor, HPE Network Capture, or HPE Network Capture Express. For details, see "Import virtual location parameters from an NV profile" on page 51. What's Next? After you select the method to use to configure the virtual location, you proceed to the next page of the Virtual Location Editor where you configure the virtual location. For details, see "Configuring virtual locations" below. Configuring virtual locations After you create a new virtual location, you choose the method that you will use to configure the virtual location. The configuration process differs according to the method that you choose. HPE Network Virtualization for Performance Testing (12.53) Page 48 of 137

49 Chapter 5: Add and configure virtual locations Note: To learn more about virtual locations, see "Add and configure virtual locations" on page 41. Define custom virtual location parameters Custom virtual location parameters Specify the following parameters to define a custom virtual location: Latency. The amount of time that it takes for an IP packet to travel across the network, from the client to the server, or from server to client. The valid range is 0 to 8,000 msec. Packet Loss. The percentage of IP packets that are sent through the network, but do not reach their destination. The valid range is 0 to 100%. Download Bandwidth. The maximum throughput through the network [in KBs per second], for data traveling from the server to the client. Upload Bandwidth. The maximum throughput through the network [in KBs per second], for data traveling from the client to the server. HPE Network Virtualization for Performance Testing (12.53) Page 49 of 137

50 Chapter 5: Add and configure virtual locations To manually specify virtual location parameters 1. In the LoadRunner Controller or Performance Center, create a new virtual location, and then open the virtual location in the Virtual Location Editor. For details, see "Add and configure virtual locations" on page Click Custom. 3. Specify all of the NV profile parameters: Latency, Packet loss, Download Bandwidth, and Upload Bandwidth. See above for details about each of the parameters. Note: For Download Bandwidth and Upload Bandwidth, you can select Unlimited to impose no bandwidth limitation on the network. What's Next? After you specify custom parameters for a virtual location, you can configure various runtime settings for the virtual location. For details, see "Specify virtual location runtime settings" on page 53. Extract virtual location parameters from the NV Global Library When you configure a virtual location, you can extract the network parameters from the NV Global Library. NV Global Library The NV Global Library contains extensive real-world network-related data that helps you to define the parameters of a virtual location. You supply a number of input parameters to the library, and the library returns parameters that define the virtual location. Note: To access the full set of geographic locations in the NV Global Library, you must be connected to the internet and be licensed to access the NV Global Library. If either of these two conditions is not met, then you will have access to a limited set of geographic locations. HPE Network Virtualization for Performance Testing (12.53) Page 50 of 137

51 Chapter 5: Add and configure virtual locations After you extract the parameters for a virtual location profile, you can configure various runtime settings for the virtual location. For details, see "Specify virtual location runtime settings" on page 53. To extract virtual location parameters from the NV Global Library 1. In the LoadRunner Controller or Performance Center, create a new virtual location, and then open the virtual location in the Virtual Location Editor. For details, see "Add and configure virtual locations" on page Click Global Library. 3. Specify the Client Location and the Server Location. These locations are typically the cities or states in which the client and server are located. Note: To access the full set of geographic locations in the NV Global Library, you must be connected to the internet and be licensed to access the NV Global Library. If either of these two conditions is not met, then you will have access to a limited set of geographic locations. 4. Click the Next button. 5. Specify the network conditions for the connection between the client and server: Technology. The technology on which the network is based. Carrier. The carrier that operates the network. Time. Specifies if the network communication occurs during business hours or outside of business hours. Communication Quality. The quality of the connection. Note: The editor displays only those options that are available for the network client and server locations that you specified. What's Next After you import parameters for a virtual location, you can configure various runtime settings for the virtual location. For details, see "Specify virtual location runtime settings" on page 53. Import virtual location parameters from an NV profile HPE Network Virtualization for Performance Testing (12.53) Page 51 of 137

52 Chapter 5: Add and configure virtual locations You can import network profile parameters from an NV profile file into LoadRunner or Performance Center, as part of a virtual location definition. There are various HPE software tools that you can use to create an NV profile file that contains the required network profile parameters. Tools to develop network profiles NV Network Editor allows you to define a complex NV profile, and to save the parameters in a NV profile file. You can then import the parameters into LoadRunner or Performance Center, as part of a virtual location definition. NV Network Editor is automatically installed whenever the HPE NV for Load Generator is installed. For details about installing Network Editor and using the Network Editor to create and save NV profiles, see "NV Network Editor overview" on page 27. HP Network Capture and HP Network Capture Express can analyze a specified network, and then determine and record the parameters that define the network. These parameters include, amongst others, the latency, packet loss and bandwidth of the network. The recorded network parameters can then be imported into LoadRunner or Performance Center, as part of a virtual location definition. For details about installing Network Capture and using Network Capture to record and save NV profiles, refer to the Network Capture User Guide. You can download Network Capture Express free from AppStore (ios version) and Google Play (Android version). To import NV profile parameters from a file 1. In the LoadRunner Controller or Performance Center, create a new virtual location, and then open the virtual location in the Virtual Location Editor. For details, see "Add and configure virtual locations" on page Click Import from a file. 3. Under Specify network profile, click Browse. 4. In the Choose File to Upload dialog box, locate and select the file that contains the required NV parameters, and then click Open. 5. If the network profile file contains more than one flow for the network, select the required flow from the list of flows. Note: An NV profile file has either a.ntx or a.ntxx extension depending on the software application that was used to create the NV profile file. When you import an NV profile file that was created using the Network Editor, IP addresses and Excluded IPs are not imported into the virtual location. If the imported file contains recorded bandwidth data, then the resulting virtual location can use Shared Bandwidth bandwidth allocation only, not Individual Bandwidth. For details, see "Allocate bandwidth to Vusers" on page 54. HPE Network Virtualization for Performance Testing (12.53) Page 52 of 137

53 Chapter 5: Add and configure virtual locations What's Next? After you import parameters for a virtual location, you can configure various runtime settings for the virtual location. For details, see "Specify virtual location runtime settings" below. Specify virtual location runtime settings After you specify the parameters for a virtual location, configure the runtime settings for the virtual location. Virtual location runtime settings Bandwidth Allocation. Allows you to specify if the specified network bandwidth is shared by all Vusers on the Load Generator, or is available to each individual Vuser. Packet Capture. Allows you to specify if packets are captured on the Load Generator. You can enable this option only if packet capture has been enabled globally for the scenario or test. For details on enabling packet capture for the scenario or test, see "Packet Capture" on page 45. For details on enabling packet capture for a Load Generator, see "Set packet capture for a virtual location" on page 55. IP Filters. Enables you to specify the list of IP addresses to which network impairments will not be applied when a LoadRunner scenario or Performance Center test runs. For details on specifying the global IP filter list that applies to the scenario or the test, see "IP filter" on page 47. For details on specifying the local IP filter list that applies to an individual load generator in a scenario or test, see "Create a local IP filter for a virtual location" on page 56. HPE Network Virtualization for Performance Testing (12.53) Page 53 of 137

54 Chapter 5: Add and configure virtual locations To specify the runtime settings for a virtual location 1. In LoadRunner or Performance Center, create a new virtual location, and then open the virtual location in the Virtual Location Editor. For details, see "Add and configure virtual locations" on page Specify the NV profile parameters as described in "Add and configure virtual locations" on page Click Next to display the Runtime settings page. 4. Specify the required runtime settings for the virtual location. For details, see: "Allocate bandwidth to Vusers" below "Set packet capture for a virtual location" on the next page "Create a local IP filter for a virtual location" on page 56 What's Next? After you specify the runtime settings for a virtual location, you can incorporate the virtual location in a scenario or test. For details, see the LoadRunner Help Center or the PC Help section in the ALM Help Center. Allocate bandwidth to Vusers When you configure the runtime settings for a virtual location, you specify how the bandwidth of the network is allocated when the Vusers run. Shared bandwidth. The specified bandwidth is shared between all Vusers in the Vuser group. This is the default option. If the Vuser group using the virtual location runs on multiple Load Generators, the specified bandwidth is assigned to each Load Generator. It is not shared between Load Generators. Individual bandwidth. The network's specified bandwidth is available to each Vuser that uses the network. This option is recommended for Vusers that emulate mobile-based users and home-based users because each device has access to the full bandwidth. HPE Network Virtualization for Performance Testing (12.53) Page 54 of 137

55 Chapter 5: Add and configure virtual locations Bandwidth allocation limitations Some Vuser protocols support only the Shared bandwidth mode of bandwidth allocation: Citrix ICA, Java Record Replay, Java Vuser, COM/DCOM, MAPI. If you specify Individual bandwidth mode for a protocol that does not support this type of bandwidth allocation, the Shared bandwidth mode will be applied when the Vusers run. When you import NV profile parameters from a file, if the imported file contains recorded bandwidth data, then the resulting virtual location can use Shared Bandwidth bandwidth allocation only, not Individual Bandwidth. See "Set the network virtualization mode" on page 44 for other limitations of these Vuser protocols. Specify the bandwidth allocation for a virtual location: 1. In the LoadRunner Controller or Performance Center, open a virtual location in the Virtual Location Editor. For details, see "Add and configure virtual locations" on page Display the Runtime settings page. 3. Under Bandwidth Allocation, click one of the available options, Share bandwidth or Individual bandwidth. See above for definitions of each of these options. 4. Click Next to display a summary of the virtual location parameters and runtime settings. See also: "Specify virtual location runtime settings" on page 53 Set packet capture for a virtual location You can configure LoadRunner or Performance Center to capture the packets that are transferred across networks while a scenario or test runs. You can then use NV Analytics to analyze the captured packets. After enabling and configuring packet capture globally for a scenario or test, you can enable packet capture for the individual virtual locations, as required. Note: HPE Network Virtualization for Performance Testing (12.53) Page 55 of 137

56 Chapter 5: Add and configure virtual locations Capturing packets is relevant only when a Vuser group (that is associated with a specific virtual location) contains only a single Vuser. You can enable the Capture Packets option for a virtual location only if packet capture has been enabled globally for the scenario or test, using the NV virtual location test level settings. For details, see "Packet Capture" on page 45. To enable or disable packet capture for a specific virtual location 1. In the LoadRunner Controller or Performance Center, open a virtual location in the Virtual Location Editor. For details, see "Add and configure virtual locations" on page Display the Runtime settings page. 3. Under Capture Packets, click On to enable packet capture, or Off to disable packet capture for the virtual location. 4. Click Next to display a summary of the virtual location parameters and runtime settings. See also: "Specify virtual location runtime settings" on page 53 Create a local IP filter for a virtual location While you run a scenario or test that incorporates network virtualization, LoadRunner or Performance Center applies various network conditions to the networks that are used by the Vusers that run during the scenario or test. These network conditions are applied to network traffic to-and-from all load generators that are associated with the virtual locations in the scenario or test. But what if, for some reason, you do not want to apply network virtualization to traffic associated with a specific IP address? You can create IP filters that include lists of IP addresses to which you do not want to apply network virtualization. Network conditions are not applied to network traffic to-or-from a specific IP address if that IP address is included in the relevant IP filter, either as the source or the destination of network traffic. HPE Network Virtualization for Performance Testing (12.53) Page 56 of 137

57 Chapter 5: Add and configure virtual locations For reasons why you may want to exclude an IP address from network virtualization, see "Exclude machines from network virtualization" below. In LoadRunner and Performance Center, there are two types of IP filters: Global IP filter. A global IP filter is a list of IP addresses to which you do not want to apply network virtualization. The global IP filter applies to all virtual locations in the scenario or test. Network impairments will not be applied to network traffic if an IP address in the global IP filter is either the source or the destination of the network traffic. Local IP filters. You can create a local IP filter for each virtual location in a scenario or test. The local IP filter includes those IP addresses to which you do not want to apply network virtualization. These locally "excluded addresses" are in addition to the IP addresses that are excluded because they appear in the global IP filter. Note: A local IP filter applies only to the virtual location with which the local IP filter is associated. For details on how to create a local IP filter, see below. For details on maintaining a global IP filter, see "IP filter" on page 47. To create a local IP filter to exclude IP addresses from a virtual location 1. In the LoadRunner Controller or Performance Center, open a virtual location in the Virtual Location Editor. For details, see "Add and configure virtual locations" on page Display the Runtime settings page. 3. In the box under Exclude these IP's, type the IP address that you want to exclude, and then click the "+" sign. 4. Repeat the previous step for all the IP addresses you want to exclude. Note: IP addresses that are included in the global IP filter are displayed in gray. To remove an IP address from the local IP filter, select the address and click "X". 5. Click Next to display a summary of the virtual location parameters and runtime settings. See also: "Specify virtual location runtime settings" on page 53 Exclude machines from network virtualization In some situations, you may need to exclude certain machines from network virtualization. For example, consider the following real-life situation. Users operate an application that accesses a number of servers. All the servers are located far away from the users, except for a single server that is located close to the users. When this scenario is emulated in a testing environment, network virtualization should be applied HPE Network Virtualization for Performance Testing (12.53) Page 57 of 137

58 Chapter 5: Add and configure virtual locations to all the emulated distant servers, whereas the near-by server should be excluded from network virtualization. To exclude a machine from network virtualization, you add the IP address of the machine to an IP filter. For details, see "IP filter" on page 47 and "Create a local IP filter for a virtual location" on page 56. The following are situations to consider excluding a machine from network virtualization: In a multi-protocol scenario or test that includes a Web server and a database server. If information from the database server is not required as a part of the load test, then you should exclude the database server. Deployment and software upgrade servers. Servers that run and store scripts on a shared network drive. HPE Network Virtualization for Performance Testing (12.53) Page 58 of 137

59 Chapter 6: Customize NV integration settings You can customize various NV settings to meet your needs. What do you want to do? Set the active adapter on NV for Load Generator 60 Change the NV for Controller port 60 Change the NV for Load Generator and NV Network Editor port 61 Set the default maximum packet list buffer size for each load generator 62 Enable the cleanup threshold on NV for Load Generator 63 HPE Network Virtualization for Performance Testing (12.53) Page 59 of 137

60 Chapter 6: Customize NV integration settings Set the active adapter on NV for Load Generator LoadRunner and Performance Center use the active adapter on each Load Generator to impose the required network conditions for network virtualization. For each Load Generator machine that has more than one NIC [network interface card], perform the following task on the NV for Load Generator machine to specify the NIC to be used. 1. On the NV for Load Generator computer, select Start > All Programs > HP Software > HP Network Virtualization for Performance Testing > HP NV for Load Generator > NV Agent. The NV Agent icon appears in the system tray of the Load Generator machine. 2. Right-click the NV Agent icon and select Active Adapter Settings. 3. In the Active Adapter dialog box, click Set Active Adapter, and select the required NIC. Change the NV for Controller port By default, port 8182 is the port number that is used to access NV for Controller, for both LoadRunner and Performance Center. Although there is typically no need to change the port number, if required, you can change the port number as described below. Note: Perform the procedure below only if NV for Controller is installed, and NV for Load Generator is not installed on the same computer. If both NV for Controller and NV for Load Generator are installed on the same computer, see "Change the NV for Load Generator and NV Network Editor port" on the next page. On the NV for Controller computer: 1. Make sure that no tests are running. 2. Make sure that the new port has been allowed in the firewall. 3. Navigate to the following folder: <installation_folder>\conf. By default, <installation_folder> is C:\Program Files (x86)\hp\nv\ or C:\Program Files\HP\NV\. 4. Open the config.properties file in a text editor. 5. Change: to com.shunra.bootstrapper.port=<current port number> com.shunra.bootstrapper.port=<new port number> 6. Save the file. 7. Restart the HP Network Virtualization for Performance Testing service (NVWatchDogService). 8. Navigate to the following folder: <installation_folder>\conf. HPE Network Virtualization for Performance Testing (12.53) Page 60 of 137

61 Chapter 6: Customize NV integration settings By default, <installation_folder> is C:\Program Files (x86)\hp\nv\ or C:\Program Files\HP\NV\. 9. Right-click the file License.url and select Properties. In the URL, replace the old port number with the new port number. 10. In all browser shortcuts to the License Manager, replace the old port number with the new port number. Change the NV for Load Generator and NV Network Editor port By default, port 8182 is the port number that is used by a browser to access the NV Network Editor, and the port number that is used to access the NV for Load Generator component. Although there is typically no need to change the port number, if required, you can change the port number as described below. For details on how to change the NV for Controller port number, see "Change the NV for Controller port" on the previous page. On each computer on which the NV for Load Generator or NV Network Editor are installed: 1. Make sure that no tests are running. 2. Make sure that the new port has been allowed in the firewall. 3. Navigate to the following folder: <installation_folder>\conf. By default, <installation_folder> is C:\Program Files (x86)\hpe\nv\ or C:\Program Files\HPE\NV\. 4. Open the vcat.properties file in a text editor. 5. Change: to com.shunra.vcat.port=<current port number> com.shunra.vcat.port=<new port number> 6. Save the file. 7. Open the config.properties file in a text editor. 8. Change: to com.shunra.bootstrapper.port=<current port number> com.shunra.bootstrapper.port=<new port number> HPE Network Virtualization for Performance Testing (12.53) Page 61 of 137

62 Chapter 6: Customize NV integration settings 9. Navigate to the following folder: <installation_folder>\lib\shunra\vcat\. By default, <installation_folder> is C:\Program Files (x86)\hpe\nv\ or C:\Program Files\HPE\NV\. 10. Open the NVAgent.exe.config file in a text editor. 11. Search for and replace "8182" (or the current port) with the new port number. 12. Restart the HPE Network Virtualization for Performance Testing service (NVWatchDogService). 13. Navigate to the following folder: <installation_folder>\conf. By default, <installation_folder> is C:\Program Files (x86)\hpe\nv\ or C:\Program Files\HPE\NV\. 14. Right-click the file Network Editor.url and select Properties. In the URL, replace the old port number with the new port number. 15. Right-click the file License.url and select Properties. In the URL, replace the old port number with the new port number. 16. In all browser shortcuts to the NV Network Editor, replace the old port number with the new port number. Set the default maximum packet list buffer size for each load generator You can specify that when a LoadRunner or Performance Center test runs, that the network packets that are transmitted across the network are captured, and stored in a buffer for future analysis. The NV Common Settings let you specify the maximum size of the packet list buffer that is available for each load generator in a scenario or a test. For details, see "Define virtual location test level settings" on page 45. You can specify the default maximum packet list buffer size that is applied to each new scenario or test. Packet capture is configured within a virtual location. One of the packet capture settings you configure in a virtual location is the maximum size of the packet list. For details, see "Define virtual location test level settings" on page 45. When using the option to generate NV Analytics reports in VuGen, the maximum packet list size is defined by the PacketListMaxSizeMB property, in the NV configuration file. You can modify the maximum buffer size as described below. To modify the default maximum packet list buffer size: 1. Stop all network virtualization tests that are currently running. 2. Navigate to the following folder: <installation_folder>\conf. By default, <installation_folder> is C:\Program Files (x86)\hp\nv\ or C:\Program Files\HP\NV\. 3. Open the file userconfiguration.xml file in a text editor. 4. Locate the PacketListMaxSizeMB property, and set it to the required value. The permitted range is 10-1,000 MB. HPE Network Virtualization for Performance Testing (12.53) Page 62 of 137

63 Chapter 6: Customize NV integration settings 5. Save the file. 6. Restart the HP Network Virtualization for Performance Testing service (NVWatchDogService). Enable the cleanup threshold on NV for Load Generator The NV cleanup threshold is a mechanism for deleting NV packet lists saved on the NV for Load Generator to free up disk space. Test results are saved in the NV for Load Generator's Data folder, as defined during setup. The Data folder is located by default in \%programdata%\hp\nv\emulation\runs. When starting a test, if the defined free disk threshold on the NV for Load Generator isn't met, the test results folders will be deleted one by one from the oldest until the threshold is restored, or until only two folders are left. If only two folders are left and the threshold wasn't restored and the test does not start on the HPE Load Generator, consider deleting unnecessary data or reducing the threshold. The cleanup threshold is 3 times that of the packet list size that is defined on the NV for Load Generator. For details on customizing the packet list size, see "Set the default maximum packet list buffer size for each load generator" on the previous page. By default the cleanup threshold mechanism is disabled. You can enable it on each machine where NV for Load Generator is installed. 1. Stop all network virtualization tests that are currently running. 2. Navigate to the following folder: <installation_folder>\conf. By default, <installation_folder> is C:\Program Files (x86)\hp\nv\ or C:\Program Files\HP\NV\. 3. Open the file userconfiguration.xml file in a text editor. 4. Locate the IsCleanupEnabled property, and set it to true. 5. Save the file. 6. Restart the HPE Network Virtualization for Performance Testing service (NVWatchDogService). HPE Network Virtualization for Performance Testing (12.53) Page 63 of 137

64 Chapter 7: Analytics This section describes the various NV reports. NV Analytics / NV Insights This comprehensive network analysis report provides information about how your application performs over various types of networks (virtual locations). This standalone HTML report covers all of the workload groups in a single report. For details, see "NV Analytics / NV Insights Report" on the next page. NV Analytics for Vugen The NV Analytics report consists of several sub-reports, each displaying different aspects of network data captured during the replay of a Vuser script. The report includes performance optimization recommendations as well as HTTP analysis and resources breakdown, as well as load times, component download analysis, response time breakdown, and details of errors received. For details, see "NV Analytics Report from VuGen" on page 82. Standalone NV Analytics Provides in-depth analysis of your application s performance during a test, and helps to identify factors that negatively impact your application's performance. Note: You can now generate NV Insights and NV Analytics reports directly during load tests, and from VuGen. If you choose to continue using the standalone NV Analytics application, you can install the updated version, included with the NV installation files. Make sure to install it on the machine where the NV Analytics for VuGen license is installed. The standalone NV Analytics application uses the same license. For details, see "NV Analytics" on page 90. NV Predictor NV Predictor analyzes test results from LoadRunner, and generates reports for SLO (service level objective) compliance. NV Predictor evaluates results of one or more key metrics for SLO compliance, based on static values and/or baseline performance. The customizable reports display a comprehensive analysis of application performance based on virtual location. For details, see "NV Predictor overview" on page 129. HPE Network Virtualization for Performance Testing (12.53) Page 64 of 137

65 Chapter 7: Analytics NV Analytics / NV Insights Report NV Analytics report (also known as NV Insights) is a comprehensive network analysis report that provides information about how your application performs over various types of networks (virtual locations), via network emulation. You have full control over what NV analyzes. You configure the groups to include in the report by selecting the NV Insights check box in Performance Center's Groups & Workloads view > Groups area. After running a test, you can open the report from the results. The standalone HTML report is packaged in a.zip file, making it easy to share with others. Save the file on your local machine and extract the files to view the pages. The files must be saved locally to display correctly. This report is suitable for various audiences and is divided into multiple sections, enabling you or others to drill down and focus on specific interests. Looking for the big picture? The "Dashboard" on the next page provides test- and script-related summary information, such as analysis coverage, network impact statistics, optimization grade breakdowns, and transaction distribution by script. Want to drill down? From the script list on the dashboard, you can drill down to the network analysis for specific transactions. You can view and filter graphs and tables that illustrate how networks impact the transactions in your applications, including transaction duration over different virtual locations and total throughput. The interactive sidebar highlights graph statistics, including network impact range and optimization grades indicating application efficiency and performance. See these sections for more details: "Group by Transaction - Graph View" on page 70 "Group by Virtual Location - Graph View" on page 75 "Group by Virtual Location - Graph View" on page 75 "Group by Virtual Location - List View" on page 79 Looking for more details? From a transaction, you can drill down to the resources and view analyses and optimization recommendations to improve your application's performance over slow networks. See these sections for more details: HPE Network Virtualization for Performance Testing (12.53) Page 65 of 137

66 Chapter 7: Analytics "Summaries" on page 83 "HTTP Analysis" on page 84 "Optimization" on page 86 "Resources" on page 88 "Endpoint Latencies" on page 88 Dashboard The main page of the NV Analytics / NV Insights report provides a dashboard showing summary information for the workload and each of its group's scripts. The impact of different networks on transaction duration is clearly illustrated, as are optimization grades that include links to optimization recommendations for your applications. Summary data The start date and time of the run. Analysis coverage Selected scripts. The total number of unique scripts included in the report. Virtual locations. The total number of unique virtual locations included in the report. A virtual location can represent a geographical location, such as London, or a mobile network type, such as 4G. Selected groups. The total number of groups that were selected for NV Insights analysis. (Network Virtualization for Performance Testing analyzes every group for which the NV Insights check box was selected when defining the groups to include in the run. The NV Insights check box is located in: HPE Network Virtualization for Performance Testing (12.53) Page 66 of 137

67 Chapter 7: Analytics LoadRunner: Scenario Scripts pane Performance Center: Groups area of the Groups & Workloads view Network impact Network impact illustrates the effect of networks on transaction duration. Click anywhere in this box to show/hide a bar displaying the number of transactions in each range in the test, as well as how much longer it took to run the transactions with virtual locations than it took to run them without. The time is displayed as a percentage range. For example: Example: If a transaction ran: 10 seconds without a virtual location 12 seconds with a virtual location The network impact percentage would be "Up to 25% longer than baseline" because it took 20% longer to run the transaction with a virtual location. If a transaction ran: 10 seconds without a virtual location 25 seconds with a virtual location The network impact percentage would be "Over 200% longer than baseline" because it took more than twice as long to run the transaction with a virtual location. Determining network impact During a LoadRunner or Performance Center run session, Network Virtualization for Performance Testing runs two additional iterations with one Vuser on every group for which the NV Insights check box is selected. HPE Network Virtualization for Performance Testing (12.53) Page 67 of 137

68 Chapter 7: Analytics The first iteration, known as the base iteration, on each group runs without virtual locations at the beginning of the test. This creates a baseline indicating how long transactions in a script take without network emulation. If the same script was included in more than one group, the baseline represents the average time it took to run the script's transactions on all groups running that script. Later in the test, Network Virtualization for Performance Testing runs an additional iteration on every group using the specified virtual locations. This is know as the analysis run. Analysis runs emulate realworld networks and can differ greatly from baseline runs. The resulting response time gap between the baseline and analysis runs is the network impact. This indicates how network properties impact your application. The smaller the time gap, the better your application is optimized. Transactions with virtual locations The highlighted numbers represent the number of transactions that ran with virtual locations (emulating networks) out of (/) the total number of transactions. For example, 10/22 indicates that out of a total of 22 transactions, 10 ran with emulated networks. Donut Chart Summary of all transactions selected for NV Analytics analysis. Each colored piece represents the total number of transactions that ran within a specific network impact range. Ranges for successful transactions that ran with network emulation are shown as percentages (%) above or below the baseline transaction duration. Hover over the pieces in the chart to view the highlighted network impact ranges. Optimization grades Optimization grades indicate how well your application's transactions perform. Grades range from A-F. The higher the grade, the better your transaction is optimized. HPE Network Virtualization for Performance Testing (12.53) Page 68 of 137

69 Chapter 7: Analytics Click anywhere in the box to see how many transactions there are for each grade. Importance of optimization Optimized applications perform better and run faster and more efficiently. For example, two web pages that look exactly the same can load at very different speeds if one of the pages is optimized and the other is not. As part of the analysis process, NV Analytics provides recommendations for optimizing your application so that network emulation is as efficient as possible. Each action is measured against accepted industry standards as well as HPE's enhanced best practices to detect performance issues. Grading transactions Each transaction is given a score comparing the transaction to programming best practices, as well as priority, based on the potential performance improvement to be obtained if the recommendation is implemented. The scores are totaled and converted to grades, based on the level of compliance, where 'A' is the most compliant. Bar chart Each bar is a different color, representing a grade. Hover over a bar to see how many transactions received a particular grade. Transaction distribution by script This area lists all of the scripts that were included in the run. Click a script to drill down and view transaction details. For each script, the following is displayed: Number of transactions for each virtual location Network impact. The number of transactions that ran with virtual locations (emulating networks) out of (/) the total number of transactions. The pieces in the donut chart indicate the total number of transactions that ran within a specific network impact range. HPE Network Virtualization for Performance Testing (12.53) Page 69 of 137

70 Chapter 7: Analytics Click anywhere in this box to show/hide a bar displaying the number of transactions in each range in the script, as well as how much longer it took to run the transactions with virtual locations than it took to run them without. Optimization grades. Notice how the grades change when you click the mobile/desktop toggle. Next steps: "NV Analytics / NV Insights Report" on page 65 "Group by Transaction - Graph View" below "Group by Virtual Location - Graph View" on page 75 "Summaries" on page 83 "HTTP Analysis" on page 84 "Optimization" on page 86 "Resources" on page 88 "Endpoint Latencies" on page 88 Group by Transaction - Graph View This page, displayed as graphs, illustrates how your application's transactions perform when run on various virtual locations. HPE Network Virtualization for Performance Testing (12.53) Page 70 of 137

71 Chapter 7: Analytics The page includes: Item Graphs Table Filter Graph View / List View toggle Description Visual display of emulated network transaction time. Details for each transaction. Filters the report by virtual locations, transaction names, optimization grades, and/or percentage of network impact. The filter button changes color when a filter is applied. Click to display either the graph view or an expandable table sorted by transaction. HPE Network Virtualization for Performance Testing (12.53) Page 71 of 137

72 Chapter 7: Analytics Item Sidebar Description Hover over a piece of the donut or an optimization grade bar to view more details. In graph view, the relevant areas of the graph are highlighted. Learn more button Click to learn more about the terms used in this report, and to access the Network Virtualization for Performance Testing Help Center. See how network impacts your transactions Graph name Description What does it tell me? Transaction Duration Total Throughput per Transaction The top graph provides a graphical representation of your application's transactions. Hover over a transaction to view more details. The bottom graph indicates the total throughput of each transaction for each virtual location. Hover over a transaction in the Transaction Duration graph to view the throughput upload and download details. Highlights the time in seconds it took to run each transaction over a virtual network. The line that crosses each group represents the baseline 1 for that transaction. The total amount of data transferred during the transaction (in KB). 1 The average length of time it takes to run transactions in a script without network emulation. HPE Network Virtualization for Performance Testing (12.53) Page 72 of 137

73 Chapter 7: Analytics Graph name Description What does it tell me? Transaction details This popup provides details for a specific transaction. Hover over a transaction in the Transaction Duration graph to view its details. Details for the transaction including: name and virtual location whether it was run from a mobile or desktop application optimization grade network impact range indicator transaction duration in seconds average baseline duration in seconds difference between the baseline and analysis durations total throughput of uploaded and downloaded data (in KB) Upload Bandwidth. Maximum throughput for data traveling from the client to the server. Download Bandwidth. Maximum throughput for data traveling from the server to the client. Group by Transaction - List View This page, displayed as a table, illustrates how your application's transactions perform when run on various virtual locations. The page includes: HPE Network Virtualization for Performance Testing (12.53) Page 73 of 137

74 Chapter 7: Analytics Item Graphs Table Filter Graph View / List View toggle Sidebar Description Visual display of emulated network transaction time. Details for each transaction. Filters the report by virtual locations, transaction names, optimization grades, and/or percentage of network impact. The filter button changes color when a filter is applied. Click to display either the graph view or an expandable table sorted by transaction. Hover over a piece of the donut or an optimization grade bar to view more details. In graph view, the relevant areas of the graph are highlighted. Learn more button Click to learn more about the terms used in this report, and to access the Network Virtualization for Performance Testing Help Center. See how network impacts your transactions Column Location Duration (sec) Description The virtual location on which the transaction ran. The total time in seconds between the transaction start and end. HPE Network Virtualization for Performance Testing (12.53) Page 74 of 137

75 Chapter 7: Analytics Column Network Impact Description The percentage showing how much longer or shorter it took to run a transaction with a virtual location (analysis iteration) than it took to run without ( base iteration 1 ). Hover over a Network Impact value to view more details. Total Throughput The total amount of data transferred during the transaction (in KB). Upload Bandwidth. Maximum throughput for data traveling from the client to the server. Download Bandwidth. Maximum throughput for data traveling from the server to the client. Optimization Grades A letter grade for the category indicating level of compliance with the best practices, with A representing the highest level of compliance. A score comparing the transaction to the best practice, based on the level of compliance to website programming rules. Group by Virtual Location - Graph View This page, displayed as graphs, illustrates how your application's transactions perform when run on various virtual locations. 1 The first iteration run on a script at the beginning of a scenario runs without a virtual location. HPE Network Virtualization for Performance Testing (12.53) Page 75 of 137

76 Chapter 7: Analytics The page includes: Item Graphs Table Filter Graph View / List View toggle Description Visual display of transactions grouped by virtual locations. Transaction details for each virtual location. Filters the report by virtual locations, transaction names, optimization grades, and/or percentage of network impact. The filter button changes color when a filter is applied. Click to display either the graph view or an expandable table sorted by virtual location. HPE Network Virtualization for Performance Testing (12.53) Page 76 of 137

77 Chapter 7: Analytics Item Sidebar Description Hover over a piece of the donut or an optimization grade bar to view more details. In graph view, the relevant areas of the graph are highlighted. Learn more button Click to learn more about the terms used in this report, and to access the Network Virtualization for Performance Testing Help Center. See how network impacts your transactions Column Transaction Duration (sec) Description The transaction that ran on this virtual location. The total time in seconds between the transaction start and end. HPE Network Virtualization for Performance Testing (12.53) Page 77 of 137

78 Chapter 7: Analytics Column Network Impact Description The percentage showing how much longer or shorter it took to run a transaction with a virtual location (analysis iteration) than it took to run without ( base iteration 1 ). Hover over a Network Impact value to view more details. Total Throughput The total amount of data transferred during the transaction (in KB). Upload Bandwidth. Maximum throughput for data traveling from the client to the server. Download Bandwidth. Maximum throughput for data traveling from the server to the client. Optimization Grades A letter grade for the category indicating level of compliance with the best practices, with A representing the highest level of compliance. A score comparing the transaction to the best practice, based on the level of compliance to website programming rules. 1 The first iteration run on a script at the beginning of a scenario runs without a virtual location. HPE Network Virtualization for Performance Testing (12.53) Page 78 of 137

79 Chapter 7: Analytics Group by Virtual Location - List View This page, displayed as graphs, illustrates how your application's transactions perform when run on various virtual locations. The page includes: Item Graphs Table Filter Graph View / List View toggle Description Visual display of transactions grouped by virtual locations. Transaction details for each virtual location. Filters the report by virtual locations, transaction names, optimization grades, and/or percentage of network impact. The filter button changes color when a filter is applied. Click to display either the graph view or an expandable table sorted by virtual location. HPE Network Virtualization for Performance Testing (12.53) Page 79 of 137

80 Chapter 7: Analytics Item Sidebar Description Hover over a piece of the donut or an optimization grade bar to view more details. In graph view, the relevant areas of the graph are highlighted. Learn more button Click to learn more about the terms used in this report, and to access the Network Virtualization for Performance Testing Help Center. See how network impacts your transactions Column Transaction Duration (sec) Description The transaction that ran on this virtual location. The total time in seconds between the transaction start and end. HPE Network Virtualization for Performance Testing (12.53) Page 80 of 137

81 Chapter 7: Analytics Column Network Impact Description The percentage showing how much longer or shorter it took to run a transaction with a virtual location (analysis iteration) than it took to run without ( base iteration 1 ). Hover over a Network Impact value to view more details. Total Throughput The total amount of data transferred during the transaction (in KB). Upload Bandwidth. Maximum throughput for data traveling from the client to the server. Download Bandwidth. Maximum throughput for data traveling from the server to the client. Optimization Grades A letter grade for the category indicating level of compliance with the best practices, with A representing the highest level of compliance. A score comparing the transaction to the best practice, based on the level of compliance to website programming rules. 1 The first iteration run on a script at the beginning of a scenario runs without a virtual location. HPE Network Virtualization for Performance Testing (12.53) Page 81 of 137

82 Chapter 7: Analytics NV Analytics Report from VuGen The report opens to the Overview page. The Overview page displays all actions contained in your script. Each action is assigned a performance score with a letter (from A to F) and a percentage, as well as details of the throughput and time taken to complete the action. Action scores and the performance optimization recommendations can be different for Mobile and Desktop clients. By default, the Overview page opens with details for desktop performance. Click Mobile to see the results for mobile. Click any action to view the detailed report for that action or click the dropdown link at the top to see the list of actions. This report includes: "Summaries" on the next page "Endpoint Latencies" on page 88 "HTTP Analysis" on page 84 "Optimization" on page 86 "Resources" on page 88 Shared Analytics Report Views These report views are available from both the NV Analytics / NV Insights Report and the NV Analytics Report from VuGen. "Summaries" on the next page "Endpoint Latencies" on page 88 "HTTP Analysis" on page 84 "Optimization" on page 86 "Resources" on page 88 HPE Network Virtualization for Performance Testing (12.53) Page 82 of 137

83 Chapter 7: Analytics Summaries Displays the Client Network Server Breakdown of the action and different metrics for the following protocols: HTTP SSL (secure communication) TCP UDP Duration Breakdown The values for the fields shown in each pie are: HPE Network Virtualization for Performance Testing (12.53) Page 83 of 137

84 Chapter 7: Analytics Client: Portion of time that the client process ran (does not include time waiting for a server response) Connect: Portion of time in which the client connects to the server, such as the request in TCP or SSL handshake (the establishment of a secure channel) Transmission: Portion of time data was downloaded or uploaded Response wait: Portion of time spent waiting for the server s response Metrics Total bytes: The total amount of transmitted data (KB) Packet overhead: The percentage of the total bytes per protocol used by non-data elements Total packets: Total number of packets associated with the protocol Application turns: The number of times a communication flow change occurs from the request to the response, per protocol KB per turn: The average of the throughput per application turn, per protocol HTTP Analysis The HTTP Analysis page includes both as a table and a graph displaying throughput values. The report displays the list of requests/resources sent over the network. You can locate bottlenecks in your application and correct them accordingly. You can: Click a resource to see more details. Click any of the table headers to sort the table according to the column. Highlight a section of the throughput graph (displayed below) to view that section's details only in the table. HPE Network Virtualization for Performance Testing (12.53) Page 84 of 137

85 Chapter 7: Analytics Item Graph Description Visual display of emulated network transaction time. Drag the markers in the timeline. to focus on a specific time frame Filter Filters the report by resource name, hosts, IP addresses, byte size, and/or time frame (in milliseconds). Columns Highlight Status code Click to highlight other resources in the report that were downloaded: From the same host During the same session The HTTP status code, such as 404 (page not found) or 200 (OK). HPE Network Virtualization for Performance Testing (12.53) Page 85 of 137

86 Chapter 7: Analytics Item Optimization Description Indicates if there are steps you can perform to optimize the transaction. Switch between desktop and mobile views to see mode-specific optimization suggestions. Resource Host IP Size (bytes) Time (ms) Timeline The URL. Hover over the resource to see the full path. The host name. The IP address. The response body size. Response time in milliseconds The position of the resource in the timeline. HTTP Metrics. Click a row to view the following for the selected resource. The graph highlights the time frame. Request Headers Response Headers Timeline Breakdown Resource Optimizations The request header details. The response header details. Breakdown of the time taken for the resource to be processed. The resource's full URL. Suggestions for optimizing the selected resource. Optimization There are many factors beyond bandwidth and latency that affect the speed at which network requests are fulfilled. For example, two web pages that look exactly the same can load at very different speeds if one of the pages is optimized and the other is not. As part of the analysis process, NV provides recommendations for optimizing network traffic in your script so that it is as efficient as possible. To detect performance issues, each transaction is measured against accepted industry standards as well as HPE's enhanced best practices. HPE Network Virtualization for Performance Testing (12.53) Page 86 of 137

87 Chapter 7: Analytics The Optimization page shows: A specific category for each best-practice Individual best-practice violations in each category Optimization recommendations for each category A grade for each category A total optimization grade for the entire transaction.the higher your grade, the better your transaction is optimized. The optimization rules differ slightly between desktop and mobile clients. Click Desktop or Mobile to switch between the clients. Note: Not every recommendation will be suitable to your environment. The recommendations in the report are not a replacement for a full investigation of each case. HPE Network Virtualization for Performance Testing (12.53) Page 87 of 137

88 Chapter 7: Analytics You can: Click the optimization rule to see which resource is not compliant with the rule. Click a resource to open the resource in the HTTP Analysis. Resources The Resources page displays the Instances and Size for each type of resource that exists in the transaction. Hover over each section of the pie chart to see the resource it represents. Endpoint Latencies The Endpoint Latencies page displays details of the latency observed at the client and server endpoints. HPE Network Virtualization for Performance Testing (12.53) Page 88 of 137

89 Chapter 7: Analytics Source: The source IP address Destination: The destination IP address URL/IP: The URL or IP address of the server Average (ms): The average value observed in the packet capture Min (ms): The minimum value observed in the packet capture 5th percentile (ms): The minimum value observed in the packet capture, not including outermost conditions 95th percentile (ms): The maximum value observed in the packet capture, not including outermost conditions Max (ms): The maximum value observed in the packet capture Sample Count: The number of packets used in the calculation of the latency HPE Network Virtualization for Performance Testing (12.53) Page 89 of 137

90 Chapter 8: NV Analytics - Standalone Provides in-depth analysis of your application s performance during a test, and helps to identify factors that negatively impact your application's performance. Note: You can now generate NV Insights and NV Analytics reports directly during load tests, and from VuGen. If you choose to continue using the standalone NV Analytics application, you can install the updated version, included with the NV installation files. Make sure to install it on the machine where the NV Analytics for VuGen license is installed. The standalone NV Analytics application uses the same license. To make the test results available to NV Analytics, make sure that packets are captured while the test runs. For details, see "Packet Capture" on page 45 and "Set packet capture for a virtual location" on page 55. After the test run, the LoadRunner Controller and Performance Center save the captured packets on the load generator computers, in the following default location: C:\ProgramData\Shunra\Emulation\Runs\ NV Analytics NV Analytics assists in pinpointing factors that negatively impact an application's performance across a network. NV Analytics performs an analysis based on packet list data, then displays the resulting data in informative reports that provide insight into an application's operation. Analysis of HTTP, HTTPS and other protocols in waterfall diagrams provide a visual look into individual resource sizes and load times, enabling rapid analysis of transaction response times and the ability to quickly identify areas for optimization. This section describes: System Requirements: Reviews the minimum host requirements for NV Analytics. See "System Requirements" below for more information. Installing Software: Provides step-by-step instructions on how to install NV Analytics. See "Install NV Analytics" on the next page for more information. Log Files: Gives the location of the NV Analytics log files. See "Log Files" on page 94 for more information. System Requirements The minimum requirements for NV Analytics are as follows: HPE Network Virtualization for Performance Testing (12.53) Page 90 of 137

91 Chapter 8: NV Analytics - Standalone Processor Memory Hard Disk Quad core 2.5 GHz or stronger 4 GB RAM 1 GB of free disk space Desktop Operating System Windows Server 2008 R2 SP1 (64 bit) Windows 7 SP1 (32/64 bit) Windows Server 2012 R2 Windows 8.0 (32/64 bit) Windows 8.1 (64 bit) Windows 2012 Browsers Internet Explorer 10 and higher Firefox Chrome Microsoft Office (for exporting reports) Office 2007 Office 2010 Office 2013 Install NV Analytics Note: You can now generate NV Insights and NV Analytics reports directly during load tests, and from VuGen. If you choose to continue using the standalone NV Analytics application, you can install the updated version, included with the NV installation files. Make sure to install it on the machine where the NV Analytics for VuGen license is installed. The standalone NV Analytics application uses the same license. Software Prerequisites Wireshark version or later must be installed prior to installing NV Analytics. Note: While NV Analytics supports WireShark and higher, we recommend using the most recent stable version of Wireshark to ensure that the latest security updates are applied. For release information and downloads, refer to the Wireshark website. The following products will be installed during the installation of NV Analytics if they are not already installed: HPE Network Virtualization for Performance Testing (12.53) Page 91 of 137

92 Chapter 8: NV Analytics - Standalone Java Runtime Environment [JRE] 8 (32-bit) Update 45 will be installed if JRE 6 (32-bit) Update 24 or higher is not present. Microsoft.NET Framework Full - if.net 4 or higher is not installed. Microsoft Silverlight Installing Software To install NV Analytics, run the NV Analytics setup file Analytics_setup.exe (as an administrator), click Install, and follow the on-line directions. Upgrading NV Analytics: When you run the installation, the installer checks if there is a previous version of NV Analytics installed. Depending on the version you have installed, the installer does one of the following: Uninstalls the previous version and then continues the installation. Prompts you to first uninstall the previous version and then run the installation again. After installing NV Analytics, you must reboot the computer. After installation, access NV Analytics from the Windows Start menu: All Programs > HPE NV for Performance Testing > HPE NV Analytics > NV Analytics. In Windows 8.x or higher, you can access NV Analytics directly from the Start or Apps screen. Silent Installation Note: If Wireshark is not installed prior to installation of NV Analytics, installation will abort. To silently install NV Analytics: 1. Copy the NV Analytics setup file from the installation package to a convenient location. 2. From the Windows Start menu, click All Programs > Accessories, right-click Command Prompt, and then click Run as administrator. 3. In the Command window, navigate to the location of the file copied in step 1, and enter the following command together with the required command line options: Analytics_setup.exe /s /v"/qn command_line_options" Command line options [* indicates a mandatory command line option]: PORT=<port number> The port used to connect to NV Analytics. If another Network Virtualization component is already installed, NV Analytics will use the same port. HPE Network Virtualization for Performance Testing (12.53) Page 92 of 137

93 Chapter 8: NV Analytics - Standalone ENABLE_REMOTE=TRUE FALSE Opens the port in the firewall. Default is TRUE. DATA_FOLDER=\"<path to data folder>\" The location where internal application data and user data is saved. Default is <Common App Data folder>\hp\nv (C:\ProgramData\HP\NV in Win 7). INSTALLDIR=\"<path to installation folder>\" The location where the application files will be installed. Default is C:\Program Files (x86)\hp\nv\ REBOOT_IF_NEED=TRUE FALSE If a reboot is needed, automatically reboots the computer after installation completes. Default is TRUE. Silent Uninstallation To silently uninstall NV Analytics: 1. Copy the NV Analytics setup file from the installation package to a convenient location. 2. From the Windows Start menu, click All Programs > Accessories, right-click Command Prompt, and then click Run as administrator. 3. In the Command Prompt window, navigate to the location of the file copied in step 1, and enter the following command together with the required command line options: Analytics_setup.exe /s /removeonly /v"/qn command_line_options" Command line options [All the command line options are optional]: REBOOT_IF_NEED=TRUE FALSE If a reboot is needed, automatically reboots the computer after uninstall completes. Default is TRUE. FORCE_REBOOT=TRUE FALSE Automatically reboots the computer after uninstall completes, whether or not a reboot is needed. Default is FALSE. DELETE_DATA=TRUE FALSE Deletes all stored NV Analytics data. Default is FALSE. HPE Network Virtualization for Performance Testing (12.53) Page 93 of 137

94 Chapter 8: NV Analytics - Standalone Installation log files Installation logs are located under C:\HP Log. Log files are named as follows: <Product name>_<date>_<time>.log For example: NV Analytics_ _ log HP NV for Load Generator_ _ log HP NV for Controller_ _ log Log Files HPE Network Virtualization for Performance Testing products log files are located in the <installation directory>\logs, by default in \Program Files\HP\NV\logs or \Program Files (x86)\hp\nv\logs. Analyzing Results This section describes how to view and analyze the results generated by NV Analytics. NV Analytics reports consist of data captured using a network virtualization packet list which is subsequently processed and displayed in intuitive reports. These reports are then examined to decipher problems that may exist. The NV Analytics reports provide detailed data about the breakdown of each transaction. Statistics of each resource being uploaded or downloaded in a transaction are displayed in both tabular and graphic format. Precise performance data includes load times, component download analysis, response time breakdown and details of errors received. Performance optimization recommendations how to improve and optimize mobile and non-mobile transaction performance are provided. This section includes: Settings To configure NV Analytics settings, click the Settings icon (wrench) on the Welcome page. The settings are also available from the main page. HTTP Port Settings To modify the default ports, select the HTTP ports tab and add one or more ports or range, each separated by a comma. Add the ports that your application under test is using. HTTP/S analysis will run on the specified ports. HPE Network Virtualization for Performance Testing (12.53) Page 94 of 137

95 Chapter 8: NV Analytics - Standalone HTTPS Keys To enable analysis of secure data, enter the HTTPS key. To add an HTTPS Key: 1. Click the "+" sign, and in the New HTTPS Key window, enter the required information. 2. To edit the information, click the pencil icon; to delete the Key click the "x". The IP, port and private key of the application server, or of the debugging proxy if installed, should HPE Network Virtualization for Performance Testing (12.53) Page 95 of 137

96 Chapter 8: NV Analytics - Standalone be provided. General Settings Displays the NV Analytics server port. Analyzing a Test Result File 1. Start NV Analytics, and click Open File (to specify the port and define other settings, see "Settings" on page 94). The following file types are supported: *.shunra, *.ved, *.cap, *.pcap and *.enc. 2. When the test result file contains more than one packet list, select some or all of the packet lists to be included in the analysis. HPE Network Virtualization for Performance Testing (12.53) Page 96 of 137

97 Chapter 8: NV Analytics - Standalone 3. After the NV Analytics window opens, in the toolbar click All Transactions to select a display of all the transactions, or select a specific transaction. The main page displays all the transactions, and each square shows one transaction. Exporting NV Analytics Results Data of the selected view and recommendations can be exported to a.csv file and in MS Word format. Exporting in.csv format The export to.csv includes the throughput, the network conditions displayed in the waterfall graph, and recommendations such as the rules, grade achieved and the violations for each rule. To export results in.csv format: 1. Click the icon in the toolbar. Note: When using a zoom view, the export will use the selected portion of the transaction, and not export the entire transaction. 2. In the Export settings dialog, select all or some of the reports, and any or all of the options, then click Export. Exporting in MS Word format When exporting in MS Word format, the HTTP Report and Recommendations can be exported. Note: When using a zoom view, the export will use the selected portion of the transaction, and not export the entire transaction. To export results in MS Word format: HPE Network Virtualization for Performance Testing (12.53) Page 97 of 137

98 Chapter 8: NV Analytics - Standalone Click the icon in the toolbar. When using a zoom view, the export will use the selected portion of the transaction, and not export the entire transaction. Results of searches that are highlighted are also highlighted in the exported reports. Viewing Reports The Overview page shows you transactions contained in the test result file you analyzed. Click on any transaction to view details of that transaction. Each Transaction section provides details of the throughput and errors, and shows a performance score with a letter and a percentage. The display is interactive, so that clicking any metric, such as "Total" displays the detailed report for that metric. The NV Analytics Overview page displays detailed breakdown of each transaction in the following categories: "Response Time" on the next page "Summaries" on the next page "General Analysis" on page 101 "Endpoint Latencies" on page 104 "TCP/UDP Errors & Sessions" on page 105 "HTTP Analysis" on page 106 HPE Network Virtualization for Performance Testing (12.53) Page 98 of 137

99 Chapter 8: NV Analytics - Standalone "HTTP Optimization" on page 110 "HTTP Resources and Responses" on page 112 Below the list of reports (on the left side), the total Duration and Network Time for the transaction is listed. If a description was added when creating the transaction, it is also displayed. In each of the report views, to return to the main page click the Home icon, or use the Back icon to return to the previous view, or the Forward icon to advance to the next view. Response Time In the All Transactions tab, click the of all transactions. icon at the top left to display the Transaction Response Times Click any bar to display the HTTP Analysis of that transaction. Summaries Displays the Client Network Server Breakdown of the transaction according to protocol, showing results in a pie chart and also additional details for the following protocols: TCP UCP HPE Network Virtualization for Performance Testing (12.53) Page 99 of 137

100 Chapter 8: NV Analytics - Standalone HTTP HTTPS (secure communication) HLS (HTTP Live Streaming) Client Network Server Breakdown The legend for the chart is below the table. The values for the fields shown in each pie are: Connect duration: portion of time in which the client connected to the server, such as the request in TCP or the triple handshake in SSL (the establishment of a secure channel) Client duration: portion of time that the client processes (does not include time waiting for a server response) Response wait duration: portion of time spent waiting for the server s response Transmission duration: portion of time data was downloaded or uploaded Parameters Total packets: total number of packets associated with the protocol Total KB: total throughput associated with the protocol; includes headers, such as TCP, IP, etc. HPE Network Virtualization for Performance Testing (12.53) Page 100 of 137

101 Chapter 8: NV Analytics - Standalone Application turns: the number of times a communication flow change occurs from the request to the response, per protocol KB per application turn: the average of the throughput per Application Turn, per protocol Max concurrent connections: the highest number of concurrent connections, per protocol Protocol overhead %: the percentage of the total throughput per protocol used by non-data elements such as headers Errors: the number of errors that occurred in the transaction General Analysis The General Analysis displays details of all subtransactions for all protocols. Display Options The following options are available from the toolbar of the General Analysis and HTTP Analysis reports. The first two options are also available in the right-click menu in the table: Highlight options: click to select from the various options that highlight the resources according to the same Source IP, etc. Display in graph options: click to select to display a session, service, etc. in the graph HPE Network Virtualization for Performance Testing (12.53) Page 101 of 137

102 Chapter 8: NV Analytics - Standalone Filter options: click to select parameters that limit the display of the subtransactions to the selected criteria Subtransaction Parameters To adjust the display of the subtranscations: Sort the rows according to ascending or descending order in each column (available in all tables) Expand or decrease the area by dragging the borders of the area (available in the General and HTTP Analysis). The following data is provided in table format. Each row represents a subtransaction: Low-level Protocol: includes TCP and UDP High-level Protocol: includes HTTP and HTTPS, DNS, etc. Source IP Source Port Destination IP Destination Port Session: number of the TCP or UDP session in which the resource was uploaded or downloaded; color coding scheme indicates the tasks occurring in the download, such as DNS resolution, TCP setup, etc. Time: response time of the request in milliseconds Size: the number of bytes used in the request/response Information: protocol specific information (when available) Timeline: the position of the Resource within the Transaction sequence Request/Response Details To display the Request/Response details, expand or collapse a row with the icon in the left column. The Client and Server portion of each Request/Response are displayed in bytes. Breakdown The breakdown of the share of the traffic for each protocol is displayed in the graph at the lower left of the window, including TCP, UDP, DNS Resolution and HTTP. Throughput Graph Note: The breakdown and throughput area can be collapsed and expanded by clicking the double arrow below the main table. Throughput The Throughput edit options are used to modify the display in the graph. HPE Network Virtualization for Performance Testing (12.53) Page 102 of 137

103 Chapter 8: NV Analytics - Standalone Use the zoom bar to focus on a specific period within the transaction. Top Sessions: displays up to five Sessions with the most traffic Top Services: displays up to five Services with the most traffic In each tab of the Throughput area, click a bar to include that item, such as a specific session in the graph. The Raw data displays the throughput in KB/sec. To add/delete components in the Throughput graph display: 1. Click one of the tabs, for example "TCP Session 18"; the graph displays the throughput for TCP Session 18 in maroon. 2. To select other sessions, services, etc., click the pencil icon for each session or hosts that is to be displayed. 3. The following tabs are available and each displays the traffic (in KB/Sec) for each of the following values: HPE Network Virtualization for Performance Testing (12.53) Page 103 of 137

104 Chapter 8: NV Analytics - Standalone TCP Sessions UDP Session Source IP Destination IP Services (IP Address and Port) 4. Select the required item, for example "TCP session 16" or a Service Name. The metric is displayed in the graph with the selected color. Endpoint Latencies The Endpoint Latencies report displays details of the latency observed at the client and server endpoints. Source IP Destination IP Names: The name of the server Best estimate (ms): Best estimate of the latency between the client and server, as deduced from the TCP connections between the two, taking into consideration any additional latency found in the packet capture due to bandwidth constraints (thus, it's possible for this value to be lower than the Min value) Min (ms): the minimum value observed in the packet capture 95th percentile (ms): the maximum value observed in the packet capture, excluding outermost conditions Max (ms): the maximum value observed in the packet capture Samples: the number of packets used in the calculation of the latency HPE Network Virtualization for Performance Testing (12.53) Page 104 of 137

105 Chapter 8: NV Analytics - Standalone TCP/UDP Errors & Sessions Select TCP errors & sessions to display the details of transaction that used the TCP protocol. Select UDP errors & sessions to display the details of transaction that used the UDP protocol. HPE Network Virtualization for Performance Testing (12.53) Page 105 of 137

106 Chapter 8: NV Analytics - Standalone Session details for TCP and UDP are displayed separately, and each includes: Session Source Address Source Port Destination Address Destination Port Total Responses Average Response Time Total Errors Error Types HTTP Analysis The details of the transaction are displayed in both table and graph format. HPE Network Virtualization for Performance Testing (12.53) Page 106 of 137

107 Chapter 8: NV Analytics - Standalone The following options are available in the toolbar of the General Analysis and HTTP Analysis reports. The first two options are also available in the right-click menu in the table: Highlight options: click to select from the various options that highlight the resources according to the same Source IP, etc. This example shows all the Session s participants. HPE Network Virtualization for Performance Testing (12.53) Page 107 of 137

108 Chapter 8: NV Analytics - Standalone Display in graph options: click to select to display a session, service, etc. in the graph To search for similar resources according to URL: 1. In the toolbar, enter the required string in the Search area, in this example "in". All matching resources are highlighted, in this case strings that contain "in". Use the arrows below the Search area to navigate between the results. 2. Clear the search results by closing the Search area (clicking the "X"). Subtransaction Details To adjust the display of the request/responses: Sort the rows according to ascending or descending order in each column. Expand or decrease the area by dragging the borders of the area. HPE Network Virtualization for Performance Testing (12.53) Page 108 of 137

109 Chapter 8: NV Analytics - Standalone The following columns are displayed in the report: Expand/Collapse icon : Show or hide details of each Request. Up/down arrows: The up arrow indicates a POST or PUT; down arrow indicates a GET; a head indicates an icon header; N/A indicates HTTPS; for all other types a star is shown. Status: The HTTP status, such as 404 (page not found) or 200 (OK). Type: The icon indicates the type of file requested, for example a graphic file. Resource: Displays the path that was accessed by the subtransaction. Host: The host (domain, server, etc.) from which the resource is uploaded or downloaded. Session: Number of the TCP session in which the resource was uploaded or downloaded. Instances: Number of times a resource with the same name appears in the transaction. Time: Response time of the request in milliseconds. Size: The number of bytes used in the request/response. Timeline: The position of the Resource within the Transaction sequence. The following data is displayed per Transaction (at the left): Network Time: The time it takes the transaction to complete between the first packet of the request and the last packet of the transaction. Duration: The total time between the transaction start and end. Description: Displays the transaction's description. To customize the display of the graph: The graph at the bottom-right displays the throughput. The handles can be used to focus on a specific period within the Request/Response. Select HTTP Throughput to display the HTTP data only, or Total Throughput to display all the data in the Request/Response. When the mouse moves over this area, a line moves with the mouse to connect the Request/Response area and the graph view, and indicates the time (in seconds) when this occurred during the capture. Use the zoom icons to zoom in, zoom out and to zoom out or to select no zoom HTTP Parameters. When a resource is highlighted, the area in the below the table displays the following breakdown: DNS Resolution: Includes the wait time from the resolve time to the start of the connection (the time between the DNS query and the first SYN to the server whose name was queried). Connect Time: The TCP Setup time and any wait time between the initialization of the connection and sending of the first request data packet. TLS Time: SSL/TLS secure channel establishment. Request: Time required for the client to send the request to the server. Wait: The time (in ms) between the last packet of the request and the first packet of the response. HPE Network Virtualization for Performance Testing (12.53) Page 109 of 137

110 Chapter 8: NV Analytics - Standalone Response: The time (in ms) between the first packet and the last packet of the response. Encrypted Data Transmission: The duration of an encrypted HTTPS session. To display the details of a Request/Response: Double-click the Request/Response name, or click the Expand/Collapse icon; the following columns are displayed: Request Headers: The syntax of the request header. Response Headers: The syntax of the response header. Request Content/Response Content: Displays the image, HTML data, etc. (non-printable characters may be replaced by a ". " Content in certain formats such as PDF, Audio, Video, Flash, and Fonts are not displayed. Details: Throughput of the request/response; for a HTTP POST the full URL is also displayed. Note: When the Resource is selected (highlighted), details of the Resource are displayed below. HTTP Optimization NV Analytics provides a number of Best Practice recommendations, based on data obtained from external sources, in addition to the knowledge obtained within HPE from current application testing methodology. Each transaction is given a score comparing the transaction to the best practice. The Total Score (out of 100) is the summary of the individual scores, based on the level of compliance to website programming rules. Individual numeric scores: Prioritization emphasizes which transactions most affect the results. Each recommendation is weighted, based on the potential performance improvement to be obtained if the recommendation is performed, and is displayed in a negative numeric point value. For example, an individual score of -8 points indicates that non-compliance with the specific recommendation more significantly affects transaction response time than non-compliance with another recommendation that received a score of -2 points. HPE Network Virtualization for Performance Testing (12.53) Page 110 of 137

111 Chapter 8: NV Analytics - Standalone Letter grades: In addition, a value from A - F is provided, indicating level of compliance with the best practice, with F representing the lowest level of compliance. Example: In the example below, there are multiple best practices that received grades of 'F', while the individual numeric scores vary. This indicates that while the transaction shows low compliance with all of these recommendations, some are more problematic, and will more significantly affect transaction results. For the best practice recommendation to "Reduce the size of your images" - perhaps the images are quite large, but there are not many of them. Therefore, a greater benefit would be obtained by implementing the recommendation with a grade of 'F' and a score of -8, than by implementing a recommendation with a grade of 'F' and a score of -4. To display the list of transactions with the relevant recommendations: 1. Select Optimization and select the rules to be included in the report; by default all are selected. The report can be displayed more than once with different rules. 2. Select Desktop or Mobile; you can also sort the results according to the Priority (according to the number of points given to each result) or by Name (alphabetical order). 3. To display additional details, click a recommendation. For example, for the recommendation "Don t download the same data twice", the details show the number of times the file appeared. Each can be clicked and viewed in the HTTP Analysis. Note: The score and recommendations may be different for desktop and mobile results, due to platform specific optimizations. To filter the list of rules for both mobile and desktop: 1. Click the Edit (pencil) icon at the top right. 2. Select or deselect a rule, then click Save. When only some of the rules are selected, an "information icon" appears beside the Edit icon. HPE Network Virtualization for Performance Testing (12.53) Page 111 of 137

112 Chapter 8: NV Analytics - Standalone HTTP Resources and Responses The Resources Breakdown displays the Instances and Total Throughput for each type of resource that was present in the transaction. Resources Breakdown The Instances pie chart shows the number of times that each type of resource appears in the transaction. The chart is divided according to type of resource, so that.jpeg images are one category, and css files are another category. The Total Throughput chart shows the breakdown according to the same categories as the Instances chart, but the size is calculated according to the Total Throughput in KB. HTTP Errors The table displays each error according to the subtransaction for which it occurred. The Errors by Class graph displays the totals according to the class, such as 4xx for HTTP client errors, and 5xx for server errors. HPE Network Virtualization for Performance Testing (12.53) Page 112 of 137

113 Chapter 8: NV Analytics - Standalone HLS Errors The table displays a summary of errors that occurred during video streaming. Response Summary Displays the number of occurrences of each HTTP response code. Secure Communication Secure HTTPS communication can be viewed in the NV Analytics report, including the Host Name, TCP Session Establishment and SSL Session Establishment data. The private key must be a text file in the Privacy Enhanced Mail (PEM) format, as in this example: -----BEGIN RSA PRIVATE KEY----- MIICXAIBAAKBgQC20yQBpiE1atCfIuiOqa9fnPyHrCsYCoCJ/hKlS8z6IOcmBsNq kzckzlmqcqp7i9o3sgvbxv1rasgn3snkqnud4pfukuhqrjppc0kpx9kddcclfu5f Bxq+7/7hDzQA+rKWqixO3WmBVcKQm+3WvpF5M+jJuIsRY8O6rOxb+FrpvwIDAQAB AoGBAKWpszHPUL4vmPTqU+KZ/bDc9rMVnkL9mTXxRQIFKqroT6vUaxTQ8i1GzHNj zyelu+nmnwjd6cwixjj/fap3hjnwmf0byzewpyc5ykkezqdkt3n549ntpxm wdo9gehjqwjbanzu4lxs/+4wkqwn5yzh8vckmmuc0pqvxw+nixrjsucc5k5vdhw4 qoobxalp1pyrafhj6yv8pfx9xhpprazs/j0cqbq8ijqhttbpzc+obucthtpr7xht 5BTuuG3rPFoW+R/D8K2apQSoj2uEgxSFLcvpcaninPHEo0bO8SfqLqCmZxkCQADb dka13lwq7wktdq2k4biwu8gd+d/gcjtbajvjj1uzmnqbspoglnaxivc6ezxpayvs CdT0yDKhjqsWggkjMWkCQHZvP1E28M51k1pLsQx43nq7zbueKZWkDg/biA3yOaLb FJ9TSJeufAXAmG/US+zCfgLuzrSuJwHiCMnhRrBOm+Y= -----END RSA PRIVATE KEY----- HPE Network Virtualization for Performance Testing (12.53) Page 113 of 137

114 Chapter 8: NV Analytics - Standalone If it is not in this format, use Open SSL to change the format of the private key. Determine the operating system of the Server. On Microsoft Windows, keys are often stored in PKCS7/DER format (locally) or in.net format (from any directory server). To convert use these commands: # for PKCS7/DER keys (as held on disk) openssl pkcs8 -nocrypt -in derfile.key -inform DER -out key.pem -outform PEM # for NET keys (from the directory server) openssl pkcs8 -nocrypt -in file.ick -inform NET -out key.pem -outform PEM On Mac OSX, Solaris, and other systems the file format used is often PKCS#12. To convert use this command: openssl pkcs12 -nodes -in file.p12 -out key.pem -nocerts -nodes On Linux use these commands: openssl x509 -nocrypt -in foo.der -informat DER -out key.pem -outformat PEM openssl x509 -nocrypt -in foo.net -informat NET -out key.pem -outformat PEM Note: To analyze secure communication, refer to the SSL options in "HTTPS Keys" on page 95. NV Analytics API NV Analytics API uses Representation State Transfer (REST) web services architecture. The analysis API requests all have a same URL structure, the prefix is: [base address]/shunra/api/analysis Note: A code sample "analyzer.py" in Python is available in the installation folder. Updates can be found in It can be used to access the API. Segments of the code are also provided in this document. The following methods are exposed: AnalysisEngines Provides a JSON contents list of the installed analysis engines. GET [base address]/shunra/api/analysis/engines Response Example: HPE Network Virtualization for Performance Testing (12.53) Page 114 of 137

115 Chapter 8: NV Analytics - Standalone The response includes the id and names of all the analysis engines. { "supportedanalysisengines":[{"name":"harexport","id":"harexport"}, {"name":"networkmeasurements","id":"networkmeasurements"}, {"name":"generalwaterfall","id":"generalwaterfall"},{"name":"http","id":"http"}, {"name":"iostats","id":"iostats"},{"name":"metrics","id":"metrics"}, {"name":"best practices","id":"best practices"}]} } Returns 200 OK 404 Not Found 500 Internal Server Error Code sample def get_engine_id(engine_name): """ Returns the analysis engine id, given its name. This can also be used as a sort of a sanity test for the analysis api. >>> get_engine_id('best practices') u'best practices' """ resp = get('/shunra/api/analysis/engines') engines = dict([(entry['name'], entry['id']) for entry in resp ['supportedanalysisengines']]) return engines[engine_name] Extract Packet Lists Provides a JSON contents of packet list names, IDs, endpoints and.pcap and.ved file unique IDs. PUT [base address]/shunra/api/analysis/packetlistmetadata Body Example: The JSON defines the analyzed emulation result (.ved or.pcap file) ID. It is a file system path for NV Analytics: HPE Network Virtualization for Performance Testing (12.53) Page 115 of 137

116 Chapter 8: NV Analytics - Standalone { "id":"c:\\tmp\\sample.ved" } The response includes the ID of the analyzed run result and the packet list metadata (names, IDs, endpoints): { } Returns 200 OK "packetlists": [{ "endpoints": [{ "name":"tokyo Office", "id":"6d0652db88c349de9382a54dc350349f" }], "name":"packet List 3", "id":"c6064d9bf25d405382e374795fef35fe" }, { "endpoints": [{ "name":"london Office", "id":"de c4eea8caeef62bfbbb493" }], "name":"packet List 2", "id":"59220e1cb4d248eba3b89a695918be91" }, { "endpoints": [{ "name":"ny Office", "id":"8c95498f7bb04c7598dde1d5e609082a" }], "name":"packet List 1", "id":"620984c9a31b4ef694a1ac47d61b6a7e" }], "runresultid":"b80de7f5ffa97428b2324c8b3a9d469b" 404 Not Found 500 Internal Server Error Code sample def get_packetlists(inputfilepath): """ Returns a dictionary of the available packet lists in the given file. The dictionary keys are the packet lists names, and the dictionary values are the packet lists ids. >>> packetlists = get_packetlists(os.path.join(sample_folder, 'Sample.ved')) HPE Network Virtualization for Performance Testing (12.53) Page 116 of 137

117 Chapter 8: NV Analytics - Standalone >>> len(packetlists) 3 >>> 'Packet List 1' in packetlists True """ resp = put('/shunra/api/analysis/packetlistmetadata', {'id':inputfilepath}) return dict([(entry['name'], entry['id']) for entry in resp['packetlists']]) def get_run_result_id(inputfilepath): resp = put('/shunra/api/analysis/packetlistmetadata', {'id':inputfilepath}) return resp['runresultid'] AnalysisRequest Represented by JSON; provides contents of current status of analysis process per packet list, per transaction and per analysis engine. The response is a dictionary with the following entries: transactionanalysisstatus a list of transactions as described below reportid an identifier for the analysis process name name of the analyzed packet list id id of the analyzed packet list The list of transactions contains an entry for each transaction assoicated with the packet list. Each entry is a dictionary, containing the transaction id (id), name (name) and analysis status (analysisstatusperengine). Analysis status is a dictionary whose keys are the analysis engine, and the values are their status as specified in the API documentation. PUT [base address]/shunra/api/analysis/request/{plid} Example: Where "plid" is a packet list unique ID that has been returned by the Extract Packet List request. Body Contains the analysis parameters such as ports, SSL Encryption Key, and the analyzed emulation result (.ved or.pcap file) ID. and the file system path, because the file is not persisted (retained?) by the system. The body is in JSON format: { "ports":"80, 8080", HPE Network Virtualization for Performance Testing (12.53) Page 117 of 137

118 Chapter 8: NV Analytics - Standalone "sslencryptionkey": " ,443,http,C:\\keys\\secret.key", "runresulthandle":"c:\\tmp\\sample.ved" } The response includes a current analysis status per transaction, per installed analysis engines and the generated analysis report id identifying the analysis parameters: { } "transactionanalysisstatus": [{ "analysisstatusperengine": { "networkmeasurements":"started", "harexport":"started", "generalwaterfall":"started", "http":"started", "iostats":"started", "metrics":"started", "best practices":"started" }, "name":"undefined", "id":"ccb8713e522241c9a691c4ed1ce72d27" }], "reportid":" ", "name":"packet List 1", "id":"620984c9a31b4ef694a1ac47d61b6a7e" Where possible analysis statuses are: public enum WorkStatus { // a job still has not been started, not proceeded, analyzed, etc Idle(0), // a job (for example emulation or analysis) started Started(1), // a job (for example emulation or analysis) finished Finished(2), // a job (for example analysis) failed Failed(3); } Note: The heuristic for analysis process completeness is that all the items have either a Finished or Failed status; otherwise some items in the analysis jobs pool have not completed yet. The client side should continue to process analysis requests until the analysis process has completed. HPE Network Virtualization for Performance Testing (12.53) Page 118 of 137

119 Chapter 8: NV Analytics - Standalone Returns 200 OK 404 Not Found 500 Internal Server Error Code Sample def analyze(inputfilepath, packetlist_id, settings={}): """ calls analysis on a given file (use settings to pass special analysis parameters such as port numbers and ssl keys) packetlist_id should be the id return by get_packetlists for a specific packet list. The response is a dictionary with the following entries: * transactionanalysisstatus - a list of transactions as described below * reportid - an identifier for the analysis process * name - name of the analyzed packet list * id - id of the analyzed packet list The list of transactions contains an entry for each transaction assoicated with the packet list. Each entry is a dictionary, containing the transaction id (id), name (name) and analysis status (analysisstatusperengine). Analysis status is a dictionary whose keys are the analysis engine, and the values are their status as specified in the API documentation. >>> inputfilepath = os.path.join(sample_folder, 'Sample.ved') >>> packetlists = get_packetlists(inputfilepath) >>> packetlist_id = packetlists['packet List 1'] >>> resp = start_analysis(inputfilepath, packetlist_id)['transactionanalysisstatus'] >>> len(resp) # only one tranasction is associated with this packet list 1 >> resp[0]['name'] u'undefined' >>> resp[0]['analysisstatusperengine']['http'] in ['Idle', 'Started', 'Finished', 'Failed'] True """ params = dict(settings) params['runresulthandle'] = inputfilepath resp = put('/shunra/api/analysis/request/'+packetlist_id, params) return resp def get_report_id(inputfilepath, packetlist_id, settings={}): return analyze(inputfilepath, packetlist_id, settings)['reportid'] """ def get_transactions(inputfilepath, packetlist_id, settings={}): Gets all the transactions associated with a given packetlist. HPE Network Virtualization for Performance Testing (12.53) Page 119 of 137

120 Chapter 8: NV Analytics - Standalone The result is a list of pairs, the first element of each pair is the transaction id, and the second is the transaction's name >>> inputfilepath = os.path.join(sample_folder, 'Sample.ved') >>> packetlists = get_packetlists(inputfilepath) >>> packetlist_id = packetlists['packet List 1'] >>> result = get_transactions(inputfilepath, packetlist_id) >>> len(result) # only one transaction is associated with this packet list 1 >>> result[0][1] u'undefined' """ return [(transaction['id'], transaction['name']) for transaction in analyze (inputfilepath, packetlist_id, settings)['transactionanalysisstatus']] def start_analysis(inputfilepath, packetlist_id, settings={}): """ Starts analysis on a given file. The response is a list, with an entry for each transaction assoicated with the packet list. Each entry is a dictionary, containing the transaction id (id), name (name) and analysis status (analysisstatusperengine). Analysis status is a dictionary whose keys are the analysis engine, and the values are their status as specified in the API documentation. """ return analyze(inputfilepath, packetlist_id, settings) def is_analysis_done(inputfilepath, packetlist_id, settings={}): """ Returns True if all the transactions associate with the given packet list were analyzed and their reports are ready to be fetched. """ resp = analyze(inputfilepath, packetlist_id, settings)['transactionanalysisstatus'] for transaction in resp: for engine_status in transaction['analysisstatusperengine'].values(): if engine_status in ['Idle','Started']: return False return True AnalysisSummary Represented by JSON; provides the contents of analysis summary per packet list, per transaction, and per analysis engine. GET HPE Network Virtualization for Performance Testing (12.53) Page 120 of 137

121 Chapter 8: NV Analytics - Standalone [base address]/shunra/api/analysis/summary/{runresulthandle}/{plid}/{reportid}/ {engineid} GET Example: /620984c9a31b4ef694a1ac47d61b6a7e/ /best%20practice [base address]/shunra/api/analysis/summary/{runresulthandle}/{plid}/{reportid}/{trid}/ {engineid} Example: /620984c9a31b4ef694a1ac47d61b6a7e/ /best%20practices The first call returns the requested analysis report for all transactions in packet list. The second returns it for the specified transaction only. The following report types are currently supported: http: HTTP Analysis best practices: Optimization report iostats: Throughput report general/waterfall: General Analysis metrics: the protocols summary and metrics report networkmeasurements: Endpoints Latencies report harexport: a report containing the HTTP subtransaction in HAR format (experimental) Response GET /shunra/api/analysis/summary/%s/%s/%s/%s/%s'%(run_result_handle, packetlist_id, report_ id, transaction_id, engine_id) Returns resp['successfulltransactionanalysis'][0]['result']??? 200 OK 404 Not Found 500 Internal Server Error Code Sample def get_analysis_report(run_result_handle, packetlist_id, report_id, transaction_id, engine_id): HPE Network Virtualization for Performance Testing (12.53) Page 121 of 137

122 Chapter 8: NV Analytics - Standalone """ Get the result of running one of the analysis engines on a given packet list """ resp = get('/shunra/api/analysis/summary/%s/%s/%s/%s/%s'%(run_result_handle, packetlist_id, report_id, transaction_id, engine_id)) return resp['successfulltransactionanalysis'][0]['result'] AnalysisArtifact A file found within a transaction, such as a picture, movie, doc, text, etc. This section contains: "Structure of an Analysis Report" below "Structure of the HTTP Waterfall Analysis Report" on the next page "Structure of the Best Practices Analysis Report" on page 126 GET [base address]/shunra/api/analysis/artifact/{filehandle} Where the artifact handle is taken from the analysis report (See "Structure of an Analysis Report" below). Returns Example: e%2f %2Fccb8713e522241c9a691c4ed1ce72d27%2F94660f9c01724f63bedfefb370dc4575%2F abf8bde dbe29cab1cecae OK 404 Not Found 500 Internal Server Error Structure of an Analysis Report As a result of executing step (4) Get Analysis Result for a specific transaction, the API returns a JSON document of the following format: { "name": "Packet List 1", "successfulltransactionanalysis": [{ "status": "Finished", "result": { HPE Network Virtualization for Performance Testing (12.53) Page 122 of 137

123 Chapter 8: NV Analytics - Standalone } "type": "Best Practices Report", "subtype": "Web Applications Best Practices Report", "version": "0.5", -- Other, analysis engine dependent fields -- }, "name": "Undefined", "id": "fe15bdf3eafe4ec8bb1b055c49ca622b" }], "reporttype": "best practices", "reportid": " ", "id": "1ae2d2ee02144e69801e0f3d1cb39d89", "failedtransactionanalysis": [] Note: The text in blue will list the actual report. Notice that since the request is for a specific engine for a specific transaction, the results for "successfulltransactionanalysis" contains only one entry, which lists the report for that transaction, wrapped with the transaction s name and ID. The reports all share a common structure, with "type", "subtype" and "version" fields. Structure of the HTTP Waterfall Analysis Report A typical HTTP Waterfall report contains many entries in the "subtransactions" list, each is one of two possible types: HTTP request/response: contains a singles HTTP request coming either from a HTTP session or a decrypted HTTPS session HTTPS session: contains details about an non-decrypted HTTPS session (highlighted in red) Not all fields are mandatory. Below, only the fields marked in red are guaranteed to be available in each entry. For example, if the response to a given request was not captured in the packet list, all the fields associated with a response do not appear in the entry. Thus, only the component that details the request s timestamps is guaranteed to be available. (Currently, the report contains only HTTP request/response pairs where the request was captured). Timestamps are marked in blue and represent the number of seconds since January 1st, The handle to the response data is marked in orange. The response data itself can be retrieved by using the Get Analysis Artifact (5) API call. { "type": "Waterfall report", "subtype": "Http Waterfall report", "version": "0.80", "subtransactions": [{ "type": "HTTP request/response", "start": , "end": , "recomendations": "", HPE Network Virtualization for Performance Testing (12.53) Page 123 of 137

124 Chapter 8: NV Analytics - Standalone "attributes": { "RequestContentSize": 0, "ResponseContentType": "application/json; charset\u003dutf-8", "StatusCode": 401, "TcpReset": false, "Method": "POST", "Scheme": "https", "ResponseContentSize": 104, "RequestHeaders": "POST /setup/ws/1/validate HTTP/1.1\r\nHost: setup.example.com naccept-encoding: gzip, deflate\r\nconnection: keep-alive\r\nproxy-connection: keepalive\r\n", "TcpSession": 4, "RequestData": "", "RequestContentType": "text/plain", "URI": "/setup/ws/1/validate", "ResponseData": "\\74b85bbff75340a9b744bf8b4d1f5f6b\\ \\5d35c39d1db84f3fa16786dc78eff622\\ ccbda491ba6d59944c1ef1114/78820e83d f134389", "ResponseHeaders": "HTTP/ Unauthorized\r\nDate: Thu, 29 Mar :01:04 \r\nconnection: Keep-alive\r\n", "host": "setup.example.com", "Referer": " }, "components": [{ "type": "DNSResolution", "start": , "end": }, { "type": "TCPSetup", "start": , "end": }, { "type": "ClientWaitAfterTCPSetup", "start": , "end": }, { "type": "TLSHandshake", "start": , "end": }, { "type": "request", "start": , "end": }, { HPE Network Virtualization for Performance Testing (12.53) Page 124 of 137

125 Chapter 8: NV Analytics - Standalone ]}, }, { } "type": "wait", "start": , "end": "type": "response", "start": , "end": { "type": "HTTPS session", "start": , "end": , "recomendations": "", "attributes": { "SentBytes": 384, "ReceivedBytes": 5792, "host": " "TcpReset": false, "TcpSession": 0 }, "components": [{ "type": "TCPSetup", "start": , "end": }, { "type": "ClientWaitAfterTCPSetup", "start": , "end": }, { "type": "TLSHandshake", "start": , "end": }, { "type": "EncryptedDataTransmission", "start": , "end": } ]} --- OTHER HTTP and HTTPS ENTRIES -- ]} HPE Network Virtualization for Performance Testing (12.53) Page 125 of 137

126 Chapter 8: NV Analytics - Standalone Structure of the Best Practices Analysis Report The best practices report is quite simple. report is a list of entries, each representing a best practice; In the example below two best practices are highlighted in blue. Each best practice contains the following fields: Name Description List of applicable scenarios (currently DesktopWeb or MobileSafari) Score: which measures the how much the transaction follows the given best practice (a number in [0,1]), Weight: measures the impact of the best practice on the transaction (a number in [0,1]). A dictionary of violations: each entry in this dictionary is a specific type of violation on the best practice and a list of resources (or TCP sessions) that are committing that violation. Notice that a transaction may not have any violation for a given best practice, as is the case with "Compress Components" below. { "type": "Best Practices Report", "subtype": "Web Applications Best Practices Report", "version": "0.5", "report": [{ "violations": {}, "name": "Compress Components", "scenarios": [ "DesktopWeb", "MobileSafari" ], "description": "Checks that textual elements are transferred in a compressed format. Compression usually reduces the response size by about 70%. Approximately 90% of current Internet traffic travels through browsers that claim to support gzip. ", "score": 100.0, "weight": 1.0 }, { "violations": { "An expiration header was not found": [ " ], "Expiration date is within the next two days": [ " " " ]}, "name": "Add long term headers expiration dates", "scenarios": [ "DesktopWeb", HPE Network Virtualization for Performance Testing (12.53) Page 126 of 137

127 Chapter 8: NV Analytics - Standalone "MobileSafari" ], "description": "Near future headers expiration dates prevent effective caching. This results in a repeat visit to your site to be slower than necessary.", "score": 65.0, "weight": 0.8 }] } NV Analytics Protocols This section provides details regarding how NV Analytics identifies and works with various protocols in the legacy reports, including: Supported Protocols The following protocols are supported and analyzed by NV Analytics. Layer 2-3 IP TCP Web HTTP HTTPS UDP Conversation Definition The definition and identification of a conversation depends on the type of analysis being performed. The definitions (identifications) are based on: IP - IP address pair (e.g ) UDP - IP address & port number pair (e.g : :3456) TCP - IP address & port number pair (e.g : :3456) HTTP - URL (e.g. Collecting Conversation Statistics NV Analytics collects statistics per conversation instance (e.g. a single Get of URL). Metrics shown will be for the following groupings: All applications Per application HPE Network Virtualization for Performance Testing (12.53) Page 127 of 137

128 Chapter 8: NV Analytics - Standalone Per application conversation (including sub-conversation) Classification of TCP, UDP, IP A conversation is classified as TCP only if no higher level protocol is present. In this case it is identified as TCP Other. A conversation is classified as UDP only if no higher level protocol is present. In this case it is identified as UDP Other. A conversation is classified as IP only if there is no higher level protocol e.g. not TCP or UDP). In this case it is identified as IP Other. Sub-Transaction Grouping The NV Analytics groups sub-conversations into a single flow so that you can get data of the whole conversation or on each of the sub-conversations that it contains. How this grouping takes places is determined by how you configure NV Analytics for grouping. Understanding Protocol Association Conversations are associated based on the relevant application protocol. This means that if we're looking at HTTP then the underlying TCP / IP communication and communication metrics will be associated with the HTTP conversation they are part of. If we found TCP Retransmissions during a HTTP Get Request-Response conversation, these TCP Retransmissions will be associated with the HTTP conversation. When a sequence number is received that is lower than expected (i.e. either a retransmission, a fast retransmission, or an out of order segment), NV Analytics assumes that it is a fast retransmission if: It has seen >= 2 duplicate ACKs for this segment (i.e. >=3 ACKs). If this segment is the next un-acked segment. If this segment came within 20ms of the last duplicate ACK (20ms is arbitrary; it should be small enough to not be confused with a retransmission timeout). HPE Network Virtualization for Performance Testing (12.53) Page 128 of 137

129 Chapter 9: NV Predictor NV Predictor analyzes test results from LoadRunner, and generates reports for SLO (service level objective) compliance. NV Predictor evaluates results of one or more key metrics for SLO compliance, based on static values and/or baseline performance. The customizable reports display a comprehensive analysis of application performance - based on virtual location. To learn more, see "NV Predictor overview" below. NV Predictor overview HP NV Predictor is a flexible analytical tool that predicts the performance degradation effects when an application under test is operated under various network conditions. NV Predictor analyzes test results from LoadRunner and Performance Center tests that include network virtualization, and generates reports for SLO compliance and baseline measurements. These customizable reports display a comprehensive analysis of application performance, by virtual location. How it works: 1. Use LoadRunner or Performance Center to conduct a performance test that includes one nonvirtual location and one or more virtual locations. 2. Run the test. Result data is generated automatically during the test run, and saved as a standard results file or a LoadRunner Analysis file. 3. Import the results file into NV Predictor and generate the reports. See also: "Install NV Predictor" below Install NV Predictor NV Predictor can be installed on the LoadRunner Controller computer, Performance Center computer, or on any other computer on which LoadRunner Analysis is installed. Note: NV Predictor does not support LoadRunner Controller or LoadRunner Analysis installed with a language pack other than English. System Requirements Before installing NV Predictor, make sure that the target computer meets the following minimum requirements: Memory: 1 GB RAM (recommended) HPE Network Virtualization for Performance Testing (12.53) Page 129 of 137

130 Chapter 9: NV Predictor Operating System: Virtual Machine: Supported operating system for the LoadRunner or Performance Center machine. VMware Workstation 6.0 or later (recommended) VMware ESX 3.0 or later (supported) Prerequisites On the LoadRunner Controller: LoadRunner Controller NV for Controller Microsoft Office 2007, 2010, or 2013 (Word and Excel) On the Performance Center Host: Performance Center NV for Controller Microsoft Office 2007, 2010, or 2013 (Word and Excel) On a LoadRunner Analysis host: LoadRunner Analysis Microsoft Office 2007, 2010, or 2013 (Word and Excel) Installing NV Predictor Run the NVPredictorSetup.exe file as an administrator, and follow the instructions in the online setup wizard. Uninstalling NV Predictor In the Microsoft Windows Control Panel, click Uninstall a program, select NV Predictor, and then click Uninstall. Create reports After running a test in LoadRunner or Performance Center, you use NV Predictor to process the test results and generate a report. Note: The input files for NV Predictor can be either the raw results files [with an.lrr extension], or results files that have been processed by LoadRunner Analysis [with an.lra extension]. To generate a report with NV Predictor 1. Create a LoadRunner or Performance Center test that includes NV virtual locations. It is recommended to include at least two virtual locations, including one "local location" with no HPE Network Virtualization for Performance Testing (12.53) Page 130 of 137

131 Chapter 9: NV Predictor network virtualization (to differentiate between networking and other performance issues). 2. Run the LoadRunner or Performance Center test and wait until all the results are collated. 3. For Performance Center users only: a. Download and unzip the RawResults.zip file for the test. 4. In NV Predictor, click File > Load LoadRunner Analysis or click the Load LoadRunner Analysis button on the NV Predictor toolbar. 5. Locate and select the results file for your test (e.g. Results_178.lra or Test_178.lrr) Launching NV Predictor automatically If you are running a test on the LoadRunner Controller, NV Predictor can be configured to launch automatically after the results are collated. See "Configure NV Predictor to run automatically after analysis" on page 135. NV Predictor will then open the test results and automatically create an NV Predictor Analysis Report after each test run. To modify the location of the test results, see "Define global settings" on page 134. See also: "Analyze results" below Analyze results After generating a NV Predictor report, you can modify various input parameters and then view how your SLOs are affected by the changing input parameters. To modify report input parameters: 1. When NV Predictor is finished processing (a progress bar indicates the advancement), open the generated report in MS Word by double-clicking the top message in the Event Log (or selecting File > Open Last Report). Note: If NV Predictor does not open automatically, it can be opened from the Start menu: HP Software > HP Network Virtualization > HP NV Predictor > NV Predictor. The Event Log refreshes with each action, including errors. To view the details of an event, doubleclick the row in the event log. 2. Once your first analysis is completed, modify classifications and settings as required to assess various scenarios. The classifications are always displayed as percentage increase from the performance of a local user (a user without network virtualization) or from Service Level Objective (SLO). a. To modify the classifications select Options > Classifications, see "Update results with classifications" on the next page. HPE Network Virtualization for Performance Testing (12.53) Page 131 of 137

132 Chapter 9: NV Predictor b. To modify the metrics used in the SLO calculation (Transaction Response Time, Bandwidth Utilization, etc.) select Options > Key Performance Indicators, see "Update results with KPIs" on the next page. 3. Once your settings are customized, you can generate a new report from the same test results without reloading LoadRunner or Performance Center results: a. To manually override any outlier transaction response times, select File > Modify Location Data. For details, see "Modify locations" below. b. Depending on the settings, NV Predictor may save the generated report in a designated location along with the MS Excel file that was used to generate the report. In addition, NV Predictor exports the report data in CVS format to allow for integration with third party applications. For further details, see "Define global settings" on page 134. Note: NV Predictor uses Microsoft Word templates to generate the reports. You can create new templates and modify the templates, but make sure to keep the required merge fields in the templates. Note that all charts in the NV Predictor report are embedded Microsoft Excel objects; you can change their layout, colors and charts type in the report itself without regenerating a new report. See also: "Modify locations" below Modify locations After running a test, you can choose to specify custom values to override the Transaction Response Time (TRT) for some transactions for specific locations. To specify an override value for a transaction: 1. Select File > Modify Location Data. 2. In the Modify Locations dialog box, select Override Location Data, and then select a transaction. 3. In the Override column, type the required value. See also: "Update results with classifications" below Update results with classifications SLO classifications match service levels experienced by users to measurable categories. This categorization assists in rating various aspects of application performance and user experience. All the fields in this dialog box are editable, enabling you to modify the name of each category and the associated values. HPE Network Virtualization for Performance Testing (12.53) Page 132 of 137

133 Chapter 9: NV Predictor The upper table compares each transaction to the defined SLO for that transaction, either the default SLO of 3 seconds, or a custom value or an override as defined in Key Performance Indicators dialog box. The lower table compares each remote Location to the non-virtual location baseline. To modify the classifications: 1. Click Options > Classifications to open the Classifications dialog box. 2. Select the classification or the percentage value and replace the label. The value in each row must be lower than the next row, for example, the 'Excellent' value cannot have a higher Max Increase % than the Good value. 3. Regenerate the reports to view the effect of the updated values. See also: "Update results with KPIs" below Update results with KPIs Some or all of the Key Performance Indicators can be enabled to modify the SLO in the "Increase from Service Level Objective Classifications" in the Classification dialog box: Include Throughput Performance Include Hits per Second Performance Include Errors per Second Performance Include Bandwidth Utilization Performance (includes the Bandwidth Utilization options available in the network virtualization settings. To learn more about bandwidth options, see the Network Virtualization for LoadRunner and Performance Center User Guide.) To enable and modify the KPI values: 1. Click Options > Key Performance Indicators to open the Key Performance Indicators dialog box. 2. Under Key Performance Indicators Service Level Objectives, select the required option, such as "Include Hits per Second Performance". 3. Enter the required values, and if required, modify any fields in the tables, then click OK. 4. Regenerate the reports to view the effect of the updated values. See also: "Select transactions" below Select transactions You can specify that the NV Predictor report include only some of the transactions that are included in the imported results file. You can search for specific instances of the transactions in the Filter area, HPE Network Virtualization for Performance Testing (12.53) Page 133 of 137

134 Chapter 9: NV Predictor scroll through the list of transactions and include the required transactions, or choose automatic transaction selection. To choose specific transactions to include in a report: 1. Click Options > Select Transactions. 2. Select Automatic selection or Manual selection. For manual selection, select the transactions to include in the reports. See also: "Define global settings" below Define global settings You can specify the following settings for NV Predictor: Include Think Time: Includes Vuser think-time in the NV Predictor reports. See the LoadRunner User Guide to learn more about Vuser think time. Export results as.csv files: Creates.csv files that contain the information that is included in the NV Predictor reports. These files can be used for further analysis using third party applications Overwrite existing reports: Overwrites existing reports to reduce disk space usage. Automatically add "Local" Location: If a test was run without a "local" location, i.e. a location in which the test is conducted without any network virtualization, a local user can be added during analysis to provide a "baseline" location in each graph and table. Results and template directories: Lets you specify where the NV Predictor templates are located, where the results files are saved, and the location that can be used by NV Predictor for temporary file storage. To change the reports directory: a. Select Options > Global Settings. b. In the Global Settings dialog box, in the Reports Directory field, type the path of the location where NV Predictor will save the reports. To change the templates directory: a. Select Options > Global Settings. b. In the Global Settings dialog box, in the Templates Directory field, type the path where the NV Predictor templates are located. To change the temporary analysis directory: a. Select Options > Global Settings. b. In the Global Settings dialog box, in the Temporary Analysis Directory field, type the path of the location for the NV Predictor temporary analysis files. HPE Network Virtualization for Performance Testing (12.53) Page 134 of 137

135 Chapter 9: NV Predictor See also: "Configure NV Predictor to run automatically after analysis" below Configure NV Predictor to run automatically after analysis You can configure NV Predictor to run automatically after analyzing the test results. Note: This functionality is available only for LoadRunner installations, not for Performance Center installations. 1. In the LoadRunner Controller, click Tools > Options. 2. Click the Execution tab. 3. In the Post Collate Command box, add the following: PredictorLauncher.exe "%ResultDir%" HPE Network Virtualization for Performance Testing (12.53) Page 135 of 137

136 Send Us Feedback Let us know how we can improve your experience with the User Guide. Send your to: HPE Network Virtualization for Performance Testing (12.53) Page 136 of 137