Re: [PATCH 2.5.53] NUMA scheduler (1/3)

From: Michael Hohnbaum (hohnbaum@us.ibm.com)
Date: Fri Jan 03 2003 - 20:58:31 EST

On Tue, 2002-12-31 at 05:29, Erich Focht wrote:
> Here comes the minimal NUMA scheduler built on top of the O(1) load
> balancer rediffed for 2.5.53 with some changes in the core part. As
> suggested by Michael, I added the cputimes_stat patch, as it is
> absolutely needed for understanding the scheduler behavior.

Thanks for this latest patch. I've managed to cobble together
a 4 node NUMAQ system (16 700 MHZ PIII procs, 16GB memory) and
ran kernbench and schedbench on this, along with the 2.5.50 and
2.5.52 versions. Results remain fairly consistent with what
we have been obtaining on earlier versions.

Kernbench:
                        Elapsed User System CPU
             sched50 29.96s 288.308s 83.606s 1240.8%
             sched52 29.836s 285.832s 84.464s 1240.4%
             sched53 29.364s 284.808s 83.174s 1252.6%
             stock50 31.074s 303.664s 89.194s 1264.2%
             stock53 31.204s 306.224s 87.776s 1263.2%

Schedbench 4:
                        AvgUser Elapsed TotalUser TotalSys
             sched50 22.00 35.50 88.04 0.86
             sched52 22.04 34.77 88.18 0.75
             sched53 22.16 34.39 88.66 0.74
             stock50 27.18 45.99 108.75 0.87
             stock53 0.00 41.96 133.85 1.07

Schedbench 8:
                        AvgUser Elapsed TotalUser TotalSys
             sched50 32.05 46.81 256.45 1.81
             sched52 32.75 50.33 262.07 1.63
             sched53 30.56 46.58 244.59 1.67
             stock50 44.75 63.68 358.09 2.55
             stock53 0.00 59.64 318.63 2.40

Schedbench 16:
                        AvgUser Elapsed TotalUser TotalSys
             sched50 55.34 70.74 885.61 4.09
             sched52 50.64 62.30 810.50 4.31
             sched53 54.04 70.01 864.84 3.68
             stock50 65.36 80.72 1045.94 5.92
             stock53 0.00 78.01 947.06 6.70

Schedbench 32:
                        AvgUser Elapsed TotalUser TotalSys
             sched50 59.47 139.77 1903.37 10.24
             sched52 58.37 132.90 1868.30 7.45
             sched53 57.35 132.30 1835.49 9.29
             stock50 84.51 194.89 2704.83 12.44
             stock53 0.00 182.95 2612.83 12.46

Schedbench 64:
                        AvgUser Elapsed TotalUser TotalSys
             sched50 63.55 276.81 4067.65 21.58
             sched52 66.75 293.31 4272.72 21.06
             sched53 62.38 276.99 3992.97 19.90
             stock50 99.68 422.06 6380.04 25.92
             stock53 0.00 441.42 6723.01 26.83

Note that the 0.00 in the AvgUser column for stock53 was due to
me not applying the cputime patch (03--cputimes_stat-2.5.53.patch).
Not sure how I managed to forget that one. I'll rerun this on
a kernel with that patch on Monday. 

-- 
Michael Hohnbaum            503-578-5486
hohnbaum@us.ibm.com         T/L 775-5486

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

This archive was generated by hypermail 2b29 : Tue Jan 07 2003 - 22:00:24 EST