William Lee Irwin III <wli@holomorphy.com> wrote:
>> 32x/48GB NUMA-Q
>> Throughput 257.986 MB/sec 128 procs
>> dbench 128 95.36s user 4833.06s system 2832% cpu 2:53.97 total
>> vma samples %-age symbol name
>> c01dc9ac 4532033 21.4566 .text.lock.dec_and_lock
>> c0169c0b 3835802 18.1603 .text.lock.dcache
>> c0106ff4 1741849 8.24666 default_idle
On Sat, Mar 15, 2003 at 02:37:18PM -0800, Andrew Morton wrote:
> Looks like it's gone nuts when 128 processes all try to close lots of
> files at the same time.
> One possible reason for this leaping out is that all the instances are now
> achieving more uniform runtimes. You can tell that by comparing the dbench
> dots.
For some reason this version of dbench doesn't produce dots. I logged
what it did produce, though. It looks something like this:
8 103.63 MB/sec^M 128 4008 105.52 MB/sec^M 128 4397 108.04 MB/sec^M 128
4811 109.90 MB/sec^M 128 5243 111.89 MB/sec^M 128 5637 114.19 MB/sec
128 6039 117.42 MB/sec^M 128 6421 120.99 MB/sec^M 128 6779 124.12 M
B/sec^M 128 7120 127.06 MB/sec^M 128 7467 128.75 MB/sec^M 128 7799 1
30.19 MB/sec^M 128 8146 131.55 MB/sec^M 128 8551 132.97 MB/sec^M 128
8975 134.09 MB/sec^M 128 9374 135.67 MB/sec^M 128 9737 137.73 MB/sec^M 12
8 10123 140.34 MB/sec^M 128 10503 142.81 MB/sec^M 128 10847 145.13 MB/s
ec^M 128 11161 146.17 MB/sec^M 128 11511 147.09 MB/sec^M 128 11857 147.
92 MB/sec^M 128 12293 149.22 MB/sec^M 128 12711 149.91 MB/sec^M 128 1309
6 151.01 MB/sec^M 128 13470 152.52 MB/sec^M 128 13808 154.25 MB/sec^M 128
14176 156.10 MB/sec^M 128 14517 157.65 MB/sec^M 128 14842 158.75 MB/sec
128 15200 159.51 MB/sec^M 128 15558 159.99 MB/sec^M 128 15947 160.84 M
B/sec^M 128 16372 161.64 MB/sec^M 128 16805 162.56 MB/sec^M 128 17175 1
63.49 MB/sec^M 128 17523 164.99 MB/sec^M 128 17884 166.28 MB/sec^M 128 1
8237 167.82 MB/sec^M 128 18575 168.78 MB/sec^M 128 18919 169.10 MB/sec^M 12
8 19246 169.26 MB/sec^M 128 19600 169.73 MB/sec^M 128 19983 170.34 MB/s
ec^M 128 20398 170.91 MB/sec^M 128 20782 171.59 MB/sec^M 128 21126 172.
44 MB/sec^M 128 21456 173.34 MB/sec^M 128 21792 174.53 MB/sec^M 128 2213
8 175.44 MB/sec^M 128 22499 176.01 MB/sec^M 128 22821 176.11 MB/sec^M 128
... and dos2unix just annihilated the log from the last run ...
-- wli
-
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 : Sat Mar 15 2003 - 22:00:44 EST