Changes between Version 22 and Version 23 of performance_tuning

12/19/16 10:47:48 (6 months ago)

Add measuring performance / lmbench sections


  • performance_tuning

    v22 v23  
    411411[ Modulation Table Online ] 
     413== Measuring Performance == 
     414In most cases, measuring the performance of a particular system function can be done with specialized tests like {{{iperf}}}. When a more thorough measurement is desired of the overall system, a small and portable benchmarking suite like {{{lmbench}}} is more appropriate. 
     416=== lmbench === 
     417{{{lmbench}}} is a micro-benchmark suite designed to focus attention on the basic building blocks of many common system applications, such as databases, simulations, software development, and networking. It provides a suite of benchmarks that attempt to measure the most commonly found performance bottlenecks in a wide range of system applications. {{{lmbench}}} is designed to identify, isolate, and reproduce these performance bottlenecks using a set of small microbenchmarks which measure system latency and bandwidth of data movement among the processor and memory, network, file system, and disk. 
     419Installation can be done on Ubuntu systems via {{{sudo apt-get install lmbench}}}. Alternatively, you can compile the source yourself which can be found at the [ lmbench website]. 
     421Run the following commands to compile the source:  
     423tar xvf lmbench3.tar.gz 
     424cd lmbench3 
     425mkdir SCCS; touch SCCS/s.ChangeSet 
     426make -C src/ 
     429Running the benchmark suite is done via {{{make results}}}. The binaries for the individual benchmarks can be found in the {{{bin/}}} directory. 
     431The {{{src/Makefile}}} has additional make targets for your convenience: 
     433# lmbench       [default] builds the benchmark suite for the current os/arch 
     434# results       builds, configures run parameters, and runs the benchmark 
     435# rerun         reruns the benchmark using the same parameters as last time 
     436# scaling       reruns the benchmark using same parameters as last time, 
     437#               except it asks what scaling value to use 
     438# hardware      reruns the hardware benchmarks using the same parameters 
     439# os            reruns the OS benchmarks using the same parameters 
     440# clean         cleans out sources and run configuration 
     441# clobber       clean and removes the bin directories 
     442# shar          obsolete, use cd .. && make shar 
     443# depend        builds make dependencies (needs gcc) 
     444# debug         builds all the benchmarks with '-g' debugging flag 
     445# assembler     builds the .s files for each benchmark3 
     448After the test has been completed, you can compare results from multiple runs with {{{make -C results/ LIST=*}}} 
     450Read the man pages for the individual benchmarks in the {{{doc/}}} directory, or the [ lmbench introduction] to learn more about the test suite.