1 |
On Saturday 29 November 2008 11:19:47 Daniel Iliev wrote: |
2 |
> Now let's put the assumptions aside and do a test. |
3 |
> |
4 |
> root@localhost # test $ cat /usr/portage/packages/All/* > test1 |
5 |
> root@localhost # test $ cp test1 test2 |
6 |
> root@localhost # test $ ls -lah |
7 |
> total 2.3G |
8 |
> drwxr-xr-x 2 root users 4.0K 2008-11-29 01:38 . |
9 |
> drwxr-xr-x 44 root users 4.0K 2008-11-29 01:36 .. |
10 |
> -rw-r--r-- 1 root users 1.2G 2008-11-29 01:38 test1 |
11 |
> -rw-r--r-- 1 root users 1.2G 2008-11-29 01:40 test2 |
12 |
> localhost test # filefrag * |
13 |
> test1: 1125 extents found, perfection would be 10 extents |
14 |
> test2: 1923 extents found, perfection would be 10 extents |
15 |
> localhost test # time cat test1 > /dev/null |
16 |
> |
17 |
> real 0m26.747s |
18 |
> user 0m2.110s |
19 |
> sys 0m1.450s |
20 |
> localhost test # time cat test2 > /dev/null |
21 |
> |
22 |
> real 0m29.825s |
23 |
> user 0m1.780s |
24 |
> sys 0m1.690s |
25 |
|
26 |
This is not a test unfortunately. You did one run on one file and one run on |
27 |
another file. We do not know what else the machine was doing at that time, |
28 |
and that unknown is a considerable one. |
29 |
|
30 |
Repeat your test eliminating this factor. Preferably, remount the filesystems |
31 |
after each run and repeat 1000 times. Then analyze the statistical |
32 |
distribution of your results. This should eliminate most random factors and |
33 |
give a more realistic real-world view. |
34 |
|
35 |
-- |
36 |
alan dot mckinnon at gmail dot com |