#Benchmark results created with go test
Test data was generated on a 4-core i7 (16GB ram) Lenovo ThinkPad running Ubuntu 14.04.
Each benchmark was run for 10 seconds (10x the default) to reduce noise seen when the benchmarks ran for 1 second.
go test -run=NONE -bench=Reader -benchtime=10s
I've included the raw test results for each of the master branch, the original pull request, and the pull request modified to avoid heap allocation by inlining the copyWriter struct literal. For the latter two I've included benchcmp output comparing them to the master branch. There is also a comparison between the pull request and it's modified form.
NOTE: The "NoCopy" benchmarks should all be noise because they don't use the
WriteTo method.
NOTE: The test output has been passed through column -t for readability. I'm
not sure if they will properly run through benchcmp in this form.
##Master branch
Raw go test output:
PASS
BenchmarkReaderManpage 1000000 20223 ns/op 126.09 MB/s
BenchmarkReaderManpage_buffered 1000000 20176 ns/op 126.39 MB/s
BenchmarkReaderManpageNoCopy 1000000 20179 ns/op 126.36 MB/s
BenchmarkReaderJSON 50000 320459 ns/op 76.23 MB/s
BenchmarkReaderJSON_buffered 50000 332016 ns/op 68.00 MB/s
BenchmarkReaderJSONNoCopy 50000 334736 ns/op 67.45 MB/s
BenchmarkReaderRandom 10000 1859415 ns/op 5640.66 MB/s
BenchmarkReaderRandom_buffered 10000 1837270 ns/op 5707.95 MB/s
BenchmarkReaderRandomNoCopy 10000 2782352 ns/op 3769.13 MB/s
BenchmarkReaderConstant 1000 16053094 ns/op 30.88 MB/s
BenchmarkReaderConstant_buffered 1000 16138921 ns/op 30.59 MB/s
BenchmarkReaderConstantNoCopy 1000 16921298 ns/op 29.17 MB/s
ok github.com/mreiferson/go-snappystream 246.718s
##Original PR (w/ extra allocation)
Raw go test output:
PASS
BenchmarkReaderManpage 1000000 19525 ns/op 130.60 MB/s
BenchmarkReaderManpage_buffered 1000000 19605 ns/op 130.07 MB/s
BenchmarkReaderManpageNoCopy 1000000 19686 ns/op 129.53 MB/s
BenchmarkReaderJSON 50000 314747 ns/op 77.61 MB/s
BenchmarkReaderJSON_buffered 50000 330732 ns/op 68.27 MB/s
BenchmarkReaderJSONNoCopy 50000 332856 ns/op 67.83 MB/s
BenchmarkReaderRandom 10000 1893232 ns/op 5539.91 MB/s
BenchmarkReaderRandom_buffered 10000 1842477 ns/op 5691.82 MB/s
BenchmarkReaderRandomNoCopy 10000 2765457 ns/op 3792.16 MB/s
BenchmarkReaderConstant 1000 16210969 ns/op 30.58 MB/s
BenchmarkReaderConstant_buffered 1000 16321791 ns/op 30.24 MB/s
BenchmarkReaderConstantNoCopy 1000 16937709 ns/op 29.14 MB/s
ok github.com/mreiferson/go-snappystream 245.194s
Difference from "master" created with benchcmp
benchmark old ns/op new ns/op delta
BenchmarkReaderManpage 20219 19615 -2.99%
BenchmarkReaderManpage_buffered 20150 19615 -2.66%
BenchmarkReaderManpageNoCopy 24238 24093 -0.60%
BenchmarkReaderJSON 318067 314540 -1.11%
BenchmarkReaderJSON_buffered 329900 322778 -2.16%
BenchmarkReaderJSONNoCopy 383377 372792 -2.76%
BenchmarkReaderRandom 1851622 1847109 -0.24%
BenchmarkReaderRandom_buffered 1842622 1831186 -0.62%
BenchmarkReaderRandomNoCopy 2763322 2737107 -0.95%
BenchmarkReaderConstant 16101400 16250396 +0.93%
BenchmarkReaderConstant_buffered 16059995 16193529 +0.83%
BenchmarkReaderConstantNoCopy 16823845 16849812 +0.15%
benchmark old MB/s new MB/s speedup
BenchmarkReaderManpage 126.11 130.00 1.03x
BenchmarkReaderManpage_buffered 126.55 130.00 1.03x
BenchmarkReaderManpageNoCopy 105.21 105.84 1.01x
BenchmarkReaderJSON 76.80 77.66 1.01x
BenchmarkReaderJSON_buffered 68.44 69.95 1.02x
BenchmarkReaderJSONNoCopy 58.89 60.56 1.03x
BenchmarkReaderRandom 5664.40 5678.24 1.00x
BenchmarkReaderRandom_buffered 5691.37 5726.92 1.01x
BenchmarkReaderRandomNoCopy 3795.09 3831.44 1.01x
BenchmarkReaderConstant 30.79 30.50 0.99x
BenchmarkReaderConstant_buffered 30.74 30.48 0.99x
BenchmarkReaderConstantNoCopy 29.34 29.29 1.00x
##Modified PR (inlined struct literal)
Raw go test output:
PASS
BenchmarkReaderManpage 1000000 19398 ns/op 131.45 MB/s
BenchmarkReaderManpage_buffered 1000000 19347 ns/op 131.80 MB/s
BenchmarkReaderManpageNoCopy 1000000 19314 ns/op 132.02 MB/s
BenchmarkReaderJSON 50000 319364 ns/op 76.49 MB/s
BenchmarkReaderJSON_buffered 50000 333882 ns/op 67.62 MB/s
BenchmarkReaderJSONNoCopy 50000 336380 ns/op 67.12 MB/s
BenchmarkReaderRandom 10000 1870011 ns/op 5608.70 MB/s
BenchmarkReaderRandom_buffered 10000 1843386 ns/op 5689.01 MB/s
BenchmarkReaderRandomNoCopy 10000 2748211 ns/op 3815.95 MB/s
BenchmarkReaderConstant 1000 16192587 ns/op 30.61 MB/s
BenchmarkReaderConstant_buffered 1000 16193977 ns/op 30.48 MB/s
BenchmarkReaderConstantNoCopy 1000 16944641 ns/op 29.13 MB/s
ok github.com/mreiferson/go-snappystream 244.296s
Difference from "master" created with benchcmp
benchmark old ns/op new ns/op delta
BenchmarkReaderManpage 20219 19324 -4.43%
BenchmarkReaderManpage_buffered 20150 19341 -4.01%
BenchmarkReaderManpageNoCopy 24238 24155 -0.34%
BenchmarkReaderJSON 318067 318259 +0.06%
BenchmarkReaderJSON_buffered 329900 328553 -0.41%
BenchmarkReaderJSONNoCopy 383377 375546 -2.04%
BenchmarkReaderRandom 1851622 1858935 +0.39%
BenchmarkReaderRandom_buffered 1842622 1839555 -0.17%
BenchmarkReaderRandomNoCopy 2763322 2745518 -0.64%
BenchmarkReaderConstant 16101400 16116817 +0.10%
BenchmarkReaderConstant_buffered 16059995 16095118 +0.22%
BenchmarkReaderConstantNoCopy 16823845 16858126 +0.20%
benchmark old MB/s new MB/s speedup
BenchmarkReaderManpage 126.11 131.96 1.05x
BenchmarkReaderManpage_buffered 126.55 131.84 1.04x
BenchmarkReaderManpageNoCopy 105.21 105.57 1.00x
BenchmarkReaderJSON 76.80 76.75 1.00x
BenchmarkReaderJSON_buffered 68.44 68.72 1.00x
BenchmarkReaderJSONNoCopy 58.89 60.12 1.02x
BenchmarkReaderRandom 5664.40 5642.12 1.00x
BenchmarkReaderRandom_buffered 5691.37 5700.86 1.00x
BenchmarkReaderRandomNoCopy 3795.09 3819.70 1.01x
BenchmarkReaderConstant 30.79 30.76 1.00x
BenchmarkReaderConstant_buffered 30.74 30.67 1.00x
BenchmarkReaderConstantNoCopy 29.34 29.28 1.00x
Finally, the difference between the original PR ("old") and the modified PR ("new") from
benchcmp though I think this is more of academic interest. The real
important numbers are the difference from master and these numbers can
obfuscate that.
benchmark old ns/op new ns/op delta
BenchmarkReaderManpage 19615 19324 -1.48%
BenchmarkReaderManpage_buffered 19615 19341 -1.40%
BenchmarkReaderManpageNoCopy 24093 24155 +0.26%
BenchmarkReaderJSON 314540 318259 +1.18%
BenchmarkReaderJSON_buffered 322778 328553 +1.79%
BenchmarkReaderJSONNoCopy 372792 375546 +0.74%
BenchmarkReaderRandom 1847109 1858935 +0.64%
BenchmarkReaderRandom_buffered 1831186 1839555 +0.46%
BenchmarkReaderRandomNoCopy 2737107 2745518 +0.31%
BenchmarkReaderConstant 16250396 16116817 -0.82%
BenchmarkReaderConstant_buffered 16193529 16095118 -0.61%
BenchmarkReaderConstantNoCopy 16849812 16858126 +0.05%
benchmark old MB/s new MB/s speedup
BenchmarkReaderManpage 130.00 131.96 1.02x
BenchmarkReaderManpage_buffered 130.00 131.84 1.01x
BenchmarkReaderManpageNoCopy 105.84 105.57 1.00x
BenchmarkReaderJSON 77.66 76.75 0.99x
BenchmarkReaderJSON_buffered 69.95 68.72 0.98x
BenchmarkReaderJSONNoCopy 60.56 60.12 0.99x
BenchmarkReaderRandom 5678.24 5642.12 0.99x
BenchmarkReaderRandom_buffered 5726.92 5700.86 1.00x
BenchmarkReaderRandomNoCopy 3831.44 3819.70 1.00x
BenchmarkReaderConstant 30.50 30.76 1.01x
BenchmarkReaderConstant_buffered 30.48 30.67 1.01x
BenchmarkReaderConstantNoCopy 29.29 29.28 1.00x