Intel® VTune™ Amplifier Platform Profiler provides a holistic view into system behavior, including CPU, memory, network, and disk usage.

Note

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.

Use the Platform Profiler to gain insights into overall system configuration, performance, and behavior. The tool provides special focus on identifying platform-level memory, storage, I/O, and interconnect bottlenecks and imbalances and can collect data on a deployed system running a full load over an extended period of time. To accomplish this, the Platform Profiler reads performance data collected using a data collector component on a target system, analyzes the data, and makes it available via web APIs and an easy-to-use HTML user interface.

Platform Profiler is available pre-installed as part of Intel® VTune™ Amplifier, which can be acquired from https://software.intel.com/en-us/vtune.

Prerequisites

Before you begin collecting data, set up the Platform Profiler.

Initial Configuration

  1. Navigate to /opt/intel/<studio>/vtune_amplifier_<version>/vpp/server.

  2. Run the following commands and follow the prompts using an account with root or sudo privileges.

    ./vpp-server config
    ./vpp-server start

Starting and Stopping Platform Profiler

To start Platform Profiler manually after initial installation or a system reboot, run:

<vtune_install_dir>/vpp/server/vpp-server start

To stop Platform Profiler, run:

<vtune_install_dir>/vpp/server/vpp-server stop

Step 1: Collect System Data

  1. Identify a typical workload for the system under analysis (system on which the collectors are installed). Systems in a data center or those running a continuous workload could use a sample time period.

  2. Set up the environment.

    <install_dir>/vpp/collector/vpp-collect config
  3. Run the following command to start data collection:

    <install_dir>/vpp/collector/vpp-collect start

    Tip

    Use the -c option when starting collection to add a comment to your result file. The comment will be visible after uploading the result for viewing and can help differentiate between results collected on the same system. For instance, you might add a description of your workload.

  4. [Optional] While data collection is running, use the following command to add timeline markers to your result at intervals of your choosing:

    <install_dir>/vpp/collector/vpp-collect mark "comment"

    Tip

    Use the comment to add an optional label to your timeline marker.

  5. Run the following command to stop data collection:

    <install_dir>/vpp/collector/vpp-collect stop

    Collection results are available in a compressed file in the current directory (<machine-name>_<datetime>.tgz/zip).

Step 2: View Results

  1. From the system on which Platform Profiler is installed, navigate to the home page in your browser (Google Chrome* recommended): http://localhost:6543

  2. Click View Results.

  3. If your result does not already appear in the list, click the Upload button and select the result file to upload.

  4. Select the result from the list to open the viewer.

  5. Navigate through the result to identify areas for optimization.

    • The timeline allows you to filter on a smaller range of collection time.

    • Hover over the topology diagram to get quick metrics for each subsystem.

    • Click an area of interest in the topology diagram or select from the Select View drop-down list to focus on that subsystem (socket, core, memory, disk, etc.).

Next Steps

  • Consider whether an upgrade to hardware components (CPU, memory, storage, network) would improve performance. After new hardware is installed, re-run Platform Profiler and compare performance between the older and newer components.

  • Analyze the collected data to determine where the performance bottleneck is most prevalent and which components are impacted. If a specific portion of the workload is causing performance issues, consider running the following analysis types using Intel® VTune™ Amplifier using a targeted collection interval (seconds instead of hours):

    • Microarchitecture Exploration: Identify issues with CPU utilization, cache, or memory

    • Memory Access: Identify memory issues

    • Input and Output: Identify storage usage issues

Find More

Document

Description

Intel VTune Amplifier User Guide

The online VTune Amplifier user guide has additional documentation for the Platform Profiler and other performance analysis opportunities.

Usage Example

This article introduces the VTune Amplifier Platform Profiler and provides an example interpretation of results collected for a movie recommendation system application.

Legal Information

Intel, the Intel logo, Intel Atom, Intel Core, Intel Xeon Phi, VTune and Xeon are trademarks of Intel Corporation in the U.S. and/or other countries.

*Other names and brands may be claimed as the property of others.

Copyright 2016-2019 Intel Corporation

This software and the related documents are Intel copyrighted materials, and your use of them is governed by the express license under which they were provided to you (License). Unless the License provides otherwise, you may not use, modify, copy, publish, distribute, disclose or transmit this software or the related documents without Intel's prior written permission.

This software and the related documents are provided as is, with no express or implied warranties, other than those that are expressly stated in the License.

Para obtener información más completa sobre las optimizaciones del compilador, consulte nuestro Aviso de optimización.