1 |
On Wednesday, 15 February 2023 13:18:24 GMT Rich Freeman wrote: |
2 |
> On Wed, Feb 15, 2023 at 4:56 AM Peter Humphrey <peter@××××××××××××.uk> |
3 |
wrote: |
4 |
> > Not long ago I read that we should allow 2GB RAM for every emerge job - |
5 |
> > that is, we should divide our RAM size by 2 to get the maximum number of |
6 |
> > simultaneous jobs. I'm trying to get that right, but I'm not there yet. |
7 |
> > |
8 |
> > I have these entries in make.conf: |
9 |
> > EMERGE_DEFAULT_OPTS="--jobs=16 --load-average=32 --autounmask=n --quiet- |
10 |
> > unmerge-warn --ke> |
11 |
> > MAKEOPTS="-j16" |
12 |
> > |
13 |
> > Today, though, I saw load averages going up to 72. Can anyone suggest |
14 |
> > better values to suit my 24 threads and 64GB RAM? |
15 |
> |
16 |
> First, keep in mind that --jobs=16 + -j16 can result in up to 256 |
17 |
> (16*16) tasks running at once. Of course, that is worst case and most |
18 |
> of the time you'll have way less than that. |
19 |
> |
20 |
> Keep in mind that you need to consider available RAM and not just |
21 |
> total RAM. Run free under the conditions where you typically run |
22 |
> emerge and see how much available memory it displays. Depending on |
23 |
> what you have running it could be much lower than 64GB. |
24 |
> |
25 |
> Beyond that, unfortunately this is hard to deal with beyond just |
26 |
> figuring out what needs more RAM and making exceptions in package.env. |
27 |
> |
28 |
> Also, RAM pressure could also come from the build directory if it is |
29 |
> on tmpfs, which of course many of us use. |
30 |
> |
31 |
> Some packages that I build with either a greatly reduced -j setting or |
32 |
> a non-tmpfs build directory are: |
33 |
> sys-cluster/ceph |
34 |
> dev-python/scipy |
35 |
> dev-python/pandas |
36 |
> app-office/calligra |
37 |
> net-libs/nodejs |
38 |
> dev-qt/qtwebengine |
39 |
> dev-qt/qtwebkit |
40 |
> dev-lang/spidermonkey |
41 |
> www-client/chromium |
42 |
> app-office/libreoffice |
43 |
> sys-devel/llvm |
44 |
> dev-lang/rust (I use the rust binary these days as this has gotten |
45 |
> really out of hand) |
46 |
> x11-libs/gtk+ |
47 |
> |
48 |
> These are just packages I've had issues with at some point, and it is |
49 |
> possible that some of these packages no longer use as much memory |
50 |
> today. |
51 |
|
52 |
Thank you all. I can see what I'm doing better now. (Politicians aren't the |
53 |
only ones who can be ambiguous!) |
54 |
|
55 |
I'll start by picking up the point I'd missed - putting MAKEOPTS in |
56 |
package.env. |
57 |
|
58 |
-- |
59 |
Regards, |
60 |
Peter. |