Here is some cool distcc benchmark result.
The results are given in seconds reported by distcc-2.18.3/bench/Build.py, and averaged over 5 run(s). Smaller is better.
| bench | local run1 | ccache run1 | ccache run2 | ccache+distcc run1 | ccache+distcc run2 | distcc-j4 run1 | distcc-j8 run1 |
| hello-2.1.1 | 0.3 | 0.5 | 0.2 | 0.3 | 0.3 | 0.5 | 0.5 |
| make-3.80 | 5.7 | 5.8 | 1.8 | 1.4 | 1.6 | 2.3 | 2.0 |
| glib-2.0.7 | 31.6 | 30.2 | 14.7 | 13.4 | 10.7 | 13.9 | 12.1 |
| gdb-5.3 | 54.2 | 101.4 | 37.6 | 39.2 | 26.3 | 40.7 | 32.0 |
| httpd-2.0.43 | 77.5 | 85.5 | 43.0 | 61.7 | 43.6 | 54.9 | 54.5 |
| samba-2.2.7 | 90.1 | 103.1 | 34.7 | 56.1 | 25.8 | 71.0 | 45.2 |
| linux-2.5.51 | 260.9 | 259.0 | 85.5 | 77.1 | 50.5 | 105.5 | 66.3 |
| samba-3.0.7 | 291.7 | 318.3 | 123.6 | 150.1 | 72.0 | 235.9 | 149.8 |
| wine-0.9.3 | 1320.0 | 1096.6 | 470.4 | 491.1 | 348.6 | 679.0 | 481.8 |
| wine-0.9.4 | 1335.0 | 1446.0 | 534.7 | 582.9 | 393.8 | 687.1 | 486.5 |
Normalized results against local build (each column is divided by local run1). Smaller is better.
| bench | local run1 | ccache run1 | ccache run2 | ccache+distcc run1 | ccache+distcc run2 | distcc-j4 run1 | distcc-j8 run1 |
| hello-2.1.1 | 1.00 | 1.67 | 0.67 | 1.00 | 1.00 | 1.67 | 1.67 |
| make-3.80 | 1.00 | 1.02 | 0.32 | 0.25 | 0.28 | 0.40 | 0.35 |
| glib-2.0.7 | 1.00 | 0.96 | 0.47 | 0.42 | 0.34 | 0.44 | 0.38 |
| gdb-5.3 | 1.00 | 1.87 | 0.69 | 0.72 | 0.49 | 0.75 | 0.59 |
| httpd-2.0.43 | 1.00 | 1.10 | 0.55 | 0.80 | 0.56 | 0.71 | 0.70 |
| samba-2.2.7 | 1.00 | 1.14 | 0.39 | 0.62 | 0.29 | 0.79 | 0.50 |
| linux-2.5.51 | 1.00 | 0.99 | 0.33 | 0.30 | 0.19 | 0.40 | 0.25 |
| samba-3.0.7 | 1.00 | 1.09 | 0.42 | 0.51 | 0.25 | 0.81 | 0.51 |
| wine-0.9.3 | 1.00 | 0.83 | 0.36 | 0.37 | 0.26 | 0.51 | 0.36 |
| wine-0.9.4 | 1.00 | 1.08 | 0.40 | 0.44 | 0.29 | 0.51 | 0.36 |
Here is our experiment setup
Here is the script patch we created to run the experiment distcc_script.html