NumaTOP

Binary and source code for a runtime memory locality analysis tool.

Overview

Latest Download

Use this observation tool for runtime memory locality characterization and analysis of processes and threads running on a non-uniform memory access (NUMA) system. NumaTOP helps you characterize the NUMA behavior of processes and threads and identify where the NUMA-related performance bottlenecks reside. The tool uses performance counter sampling technologies from Intel and associates the performance data with Linux* system runtime information to provide real-time analysis in production systems.

Step-by-Step NUMA Analysis

Latest Download

Added support for products formerly code named Broadwell.
December 15, 2015

Download NumaTOP 1.0.4

Release Notes

User Guide

Features

Characterize the locality of all running processes and threads to identify those with the poorest locality in the system.

Identify the hot memory areas, report average memory access latency, and provide the location where accessed memory is allocated.

Provide the call chains in the process or thread code that accesses a given hot memory area.

Provide the call chains when the process or thread generates certain counter events. The call chains help to locate the source code that generates the events.

Provide per-node statistics for memory and CPU use.

Show a list of processes and threads sorted by metrics (default sort is by CPU use), with the top process having the highest CPU use in the system and the bottom process having the lowest CPU use.

Use hot keys to resort the output by these metrics: remote memory accesses (RMA), local memory accesses (LMA), RMA and LMA ratio, cycles per instruction (CPI), and CPU use.

Who This Is For

This project is for data center IT and facilities managers and developers, testers, and debuggers.

Project Specifics

This open source project is distributed under the BSD license. With an established code base containing over 12,000 lines of code, NumaTOP is written in the C programming language.

Intel's Involvement

Intel is the leading contributor to Node Manager. This project is optimized through open source for Intel® Xeon® processors.

Figure 1. Image of the home screen

Most modern systems use a Non-Uniform Memory Access (NUMA) design for multiprocessing. In NUMA systems, memory and processors are organized in such a way that some parts of memory are closer to a given processor, while other parts are farther from it. A processor can access memory that is closer to it much faster than the memory that is farther from it. Hence, the latency between the processors and different portions of the memory in a NUMA machine may be significantly different.

NumaTOP is a tool that periodically tracks and analyzes the NUMA activity of processes and threads and displays useful metrics. Users can scroll up and down by using the up or down key to navigate in the current window and can use several hot keys, shown at the bottom of the window, to switch between windows or to change the running state of the tool. For example, hot key R refreshes the data in the current window.