1 |
Hi, |
2 |
|
3 |
I have a fast desktop computer and a slow laptop. Both are ~amd64 |
4 |
Gentoo. After some of the recent discussions about Gentoo on slow |
5 |
devices, I thought I'd dust off the laptop and try to bring it up to |
6 |
date. |
7 |
|
8 |
I'd like to use distcc to make the desktop do all the compiling during |
9 |
emerges. I've never been able to get distcc working properly, or, at |
10 |
least, I've never been able to get it working to the point where using |
11 |
distcc is any faster than not using it at all. |
12 |
|
13 |
Desktop: 192.168.0.100 |
14 |
Laptop: 192.168.0.107 |
15 |
|
16 |
My goal is to just don't use the laptop at all. Do all compiling on |
17 |
the desktop. So far, while it appears to be working (some things are |
18 |
being compiled on the desktop, a very insignificant amount -- CPU |
19 |
usage on the desktop is basically 0% during emerges on the laptop), |
20 |
compile times are no better than just compiling everything on the |
21 |
laptop. The laptop still has 100% CPU activity during emerges. I |
22 |
realize the laptop is still going to do work during emerges, but |
23 |
thought at least distcc might be able to help offload most of the |
24 |
heavy part (the compiling). |
25 |
|
26 |
Both machines contain "distcc" in FEATURES. It's not using |
27 |
-march=native. I've tried various -jN values with no real difference |
28 |
in performance. |
29 |
|
30 |
On the desktop, /etc/conf.d/distcc contains (among other things): |
31 |
DISTCCD_OPTS="${DISTCCD_OPTS} --allow 192.168.0.0/24" |
32 |
DISTCCD_OPTS="${DISTCCD_OPTS} --listen 192.168.0.100" |
33 |
|
34 |
And /etc/distccd/hosts contains: |
35 |
192.168.0.0/24 |
36 |
|
37 |
On the laptop, /etc/conf.d/distcc contains (among other things): |
38 |
DISTCCD_OPTS="${DISTCCD_OPTS} --allow 192.168.0.0/24" |
39 |
DISTCCD_OPTS="${DISTCCD_OPTS} --listen 192.168.0.107" |
40 |
|
41 |
And /etc/distccd/hosts contains: |
42 |
192.168.0.100 |
43 |
|
44 |
Does anyone know any tricks or can tell me if I'm doing something |
45 |
wrong? How much of a speed-up should I realistically expect? Should |
46 |
distcc be able to exploit the fast machine to the fullest of its |
47 |
abilities? |
48 |
|
49 |
Thanks |