| 1 |
> emm, from my very personal point of view: no |
| 2 |
> |
| 3 |
> swap is horrible slow. Its use must be avoided at any cost. Swap sucks. |
| 4 |
> Everything is faster than accessing swap. So hitting the disk to read or |
| 5 |
> write some files is IMHO better than hitting the disk to shove X into swap. |
| 6 |
> |
| 7 |
> X in swap is another problem. You can be sure, if X is forced into swap, |
| 8 |
> because gcc uses up all ram for itself, everything sucks. Mouse is jerky, |
| 9 |
> windows need ages to get displayed. |
| 10 |
> |
| 11 |
> Everything that makes swapping more likely is bad advise and shall be |
| 12 |
> avoided. |
| 13 |
|
| 14 |
I'm that opinion too. Particularly when system and swap are on the same disk. |
| 15 |
And even if swap and system are located on different disks, it's one of the |
| 16 |
worst things that can happen to the performance of a system. |
| 17 |
|
| 18 |
Anyway, I've never experienced such behaviour since I've always had enough |
| 19 |
ram. And I never set portage niceness. Even on desktops with single cores, |
| 20 |
100Hz timer frequency and without kernel preemption the system was |
| 21 |
responsive. I think if a system gets "freezes" it must have something to do |
| 22 |
with heavy IO-load and to less main memory. |
| 23 |
|
| 24 |
Rgds |
| 25 |
Bernhard |
| 26 |
-- |
| 27 |
gentoo-amd64@g.o mailing list |
Archives