User Guide

Contents

Web Server Interface

Use Intel® VTune™ Profiler in a web server mode (preview feature) to get an easy on-boarding experience, benefit from a collaborative multi-user environment, and access a common repository of collected performance results.
This is a
PREVIEW FEATURE
. A preview feature may or may not appear in a future production release. It is available for your use in the hopes that you will provide feedback on its usefulness and help determine its future. Data collected with a preview feature is not guaranteed to be backward compatible with future releases. Please send your feedback to parallel.studio.support@intel.com or to intelsystemstudio@intel.com.
The web server interface helps you quickly get started with the tool since you do not need to install the VTune Profiler as a desktop application on every client system. You can use the VTune Profiler Server to configure and control analysis on arbitrary target systems and view collected results.
To run an analysis via a web interface:
  1. (Personal mode) Run the VTune Profiler Server to get a URL to access the web interface.
    (Reverse proxy/SAML SSO modes) Get the server URL from your admin.
  2. Access the server via the URL.
  3. Select your target system:
To control
VTune
Profiler
agents, use the Administrator Dashboard.

Run VTune Profiler Server

Prerequisite:
VTune Profiler Server is installed with VTune Profiler GUI.
In the personal/evaluation usage mode, run the VTune Profiler Server as follows:
  1. Start the VTune Profiler Server:
    <vtune-install-dir>/bin64/vtune-backend
    If you want the VTune Profiler Server to access a specific TCP port, specify it with the
    --web-port
    option. For example:
    vtune-backend --web-port=8080
    VTune Profiler Server outputs a URL to access the GUI. For the first run, the URL includes a one-time token. For example:
    Serving GUI at https://127.0.0.1:64880?one-time-token=0160852eeff593e0ab0a0f90991b4efa
    By default, access to the VTune Profiler Server is limited to the local host only. To enable access from remote client and target systems, restart the server with the
    --allow-remote-ui
    option.
  2. Open the URL with the provided one-time token.
    If you start the VTune Profiler Server in the personal/evaluation mode with no signed TLS certificate provided, your web browser warns you that the default self-signed server certificate is not trusted and asks for your confirmation to proceed.
  3. Set a passphrase in the
    Set Passphrase
    dialog box.
In the reverse proxy or SAML SSO usage modes, use the URL provided by your admin to access the VTune Profiler Server instance installed in a lab.

Deploy the
VTune
Profiler
Agent

You can choose between automated and manual deployment of the
VTune
Profiler
Agent.
Deploy the Agent automatically
VTune Profiler Server uses SSH for automated agent deployment. Running an SSH server on the target machine is required for automated deployment.
To deploy the Agent automatically:
  1. Enter the target machine username.
  2. Enter the credentials for target machine:
    • For
      Public key authentication
      , add your public SSH key to the
      authorized_keys
      file on the target system for the user account that you specify in the
      Username
      field. Then, select the
      Private key
      file on your client machine. If your private key is encrypted, specify the
      Private key passphrase
      .
    • Alternatively, switch to
      Password authentication
      and provide the username and password.
  3. Optionally, specify the deployment directory.
  4. Click the
    Deploy Agent
    button.
Deploy VTune Profiler Agent
Deploy the Agent manually
To deploy the Agent manually:
  1. Click the
    Download Agent Manually
    button In the
    WHERE
    pane of the
    Configure Analysis
    window or access the
    http://
    <VTune Profiler Server URL>
    /api/collection-agent/download
    URL to download the Agent.
    You can use tools such as
    wget
    to download the Agent directly to the target system.
  2. Extract the Agent archive with your tool of choice and copy its contents to the target system.
  3. Run the
    vtune-agent
    executable on the target system and specify the agent owner using the
    -owner
    <vtune-user-id>
    option.
    You can find your
    VTune
    Profiler
    user ID in the
    About
    dialog.
  4. Compare the
    Agent key fingerprint
    in the
    WHERE
    pane of the
    Configure Analysis
    window with the fingerprint printed out by the agent upon startup. If they match, click the
    Admit Agent
    button.

Shared Agents

You can run a shared
VTune
Profiler
Agent. In this case, the Agent will be available to all users of an instance of
VTune
Profiler
Server. This means that any user of this
VTune
Profiler
Server instance will be able to run data collection using this agent. It is recommended to only run shared agents using dedicated faceless accounts.
To deploy a shared agent, check the
Share the agent with all VTune Profiler users
checkbox in the
WHERE
pane of the
Configure Analysis
dialog, or use the
--shared
command line option when deploying an agent manually.

Select a Client System

To profile a
client system
, which is the same machine that you use to access the VTune Profiler Server via a web browser, do the following:
  1. Click
    New Project
    and specify a name for the new project.
    VTune Profiler opens the project configuration with your
    localhost
    pre-selected as a target system.
  2. Configure your analysis target and analysis type.

Select a Remote System

To profile a remote target system, do the following:
  1. In the
    WHERE
    pane of the
    Configure Analysis
    window, click the down arrow to see available target systems.
  2. Select
    Add new remote target...
    .
    VTune
    Profiler
    maintains a list of used remote systems, if any, and displays it under
    Remote Targets
    .
  3. Enter the hostname or IP address.

Run the Analysis

Once the Agent is running, the
Configure Analysis
pane displays information that
VTune
Profiler
is detecting the device configuration.
The Agent downloads the collectors and the target package, which is approximately 100MB in size. Once the target package is downloaded, the Agent analyzes the target system configuration and displays the applicable analysis types.
To run an analysis:
  1. Install the Intel sampling drivers manually by running these commands:
    On Windows* OS:
    <vtune-agent-dir>
    \bin64\amplxe-sepreg.exe
    On Linux* OS:
    <vtune-agent-dir>
    /sepdk/src/build-driver
    <vtune-agent-dir>
    /sepdk/src/insmod-sep
    The
    <vtune-agent-dir>
    is the
    <vtune_profiler_<version>>
    installation folder created on the client system by
    VTune
    Profiler
    .
  2. Configure your analysis target and analysis type.
  3. Click the
    Start
    button to run the analysis.

Control
VTune
Profiler
Agents

The
Administrator Dashboard
feature of
VTune
Profiler
Server enables you to monitor and manage one or multiple agents from a single point.
To open the Administrator Dashboard:
  1. Open the
    VTune
    Profiler
    Server interface in your browser.
  2. In the main toolbar, open the drop-down menu and select
    Administrator Dashboard
    .
    The dashboard opens in a new tab and shows all agents that are related to this instance of
    VTune
    Profiler
    Server. This includes both connected and disconnected agents.
The dashboard enables you to:
  • View information related to this agent:
    • Target system IP address and hostname
    • The username of the agent's user.
    • Current connection status.
  • Admit or stop one or multiple agents. To admit or stop multiple agents, select the agents by ticking the checkboxes and click
    Admit selected
    or
    Stop selected
    .

Product and Performance Information

1

Intel's compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice.

Notice revision #20110804