1 |
Volker Armin Hemmann <volker.armin.hemmann@××××××××××××.de> posted |
2 |
200802131717.57766.volker.armin.hemmann@××××××××××××.de, excerpted below, |
3 |
on Wed, 13 Feb 2008 17:17:57 +0100: |
4 |
|
5 |
> Not real benchmarks. But kdepim with enablefinal, 1gb of ram and -j2 |
6 |
> several hours. With j1 2h. |
7 |
> . |
8 |
> kdepim with 2gb of ram and j2 30m |
9 |
> Just because first case swap storn, last case no swap at all. |
10 |
|
11 |
But that's measuring the effect of the additional memory as much as it is |
12 |
the effect of swap. What you need to do is kdepim with enablefinal, a |
13 |
gig of RAM and -j1 with PORTAGE_TMPDIR set to someplace on disk, against |
14 |
the same thing but with PORTAGE_TMPDIR set to a tmpfs of sufficient |
15 |
size. Then and only then are you measuring the effect of tmpfs vs no |
16 |
tmpfs. He (and I) are arguing that the tmpfs case will take less time |
17 |
because some files will get deleted before they end up being written to |
18 |
disk at all, as opposed to /all/ of them being written to disk (well, if |
19 |
they remain more than a few seconds anyway). |
20 |
|
21 |
If you like you can of course do the test with -j2, but then you must use |
22 |
-j2 for both cases, so again PORTAGE_TMPDIR on tmpfs vs on disk is the |
23 |
only variable. |
24 |
|
25 |
This of course assumes similar system load otherwise, presumably either |
26 |
an idle system (in X/GNOME/KDE/whatever both times or at the CLI both |
27 |
times) or only something steady, say streaming Internet radio, the same |
28 |
station at the same bits per second without visualizations, going on, |
29 |
with the same apps loaded in the background. No answering mails or |
30 |
browsing or anything else interactive or variable since it's possible you |
31 |
could then be doing something different at a critical time in one case |
32 |
vs. the other. |
33 |
|
34 |
-- |
35 |
Duncan - List replies preferred. No HTML msgs. |
36 |
"Every nonfree program has a lord, a master -- |
37 |
and if you use the program, he is your master." Richard Stallman |
38 |
|
39 |
-- |
40 |
gentoo-amd64@l.g.o mailing list |