1 |
Hi there! |
2 |
|
3 |
I still wonder why my KDE4 system starts swapping so early. Until a week |
4 |
ago, I had 6G of RAM, but after a day of being logged in, I usually had some |
5 |
swap usage. Sometimes this goes up to 1.5G, this is when the system becomes |
6 |
way too slow and I log out. |
7 |
|
8 |
Normally I don't mind having swap, I always had. But to me it looks like |
9 |
this is worse than it should be. From the point when swapping starts, it's |
10 |
nearly permanent, like, when switching applications and desktops. As if |
11 |
important stuff were swapped out that will be needed again soon. While |
12 |
unimportant stuff stays in memory. |
13 |
|
14 |
This was even worse when using the ati-drivers/fglrx, but at the moment |
15 |
I'm using xf86-video-ati with xorg-server-1.10.1.901 and KMS. |
16 |
echo 1 > /proc/sys/vm/drop_caches helps a little, but not much. And not |
17 |
for long. |
18 |
|
19 |
Sometimes I quit memory-hungry applications like Kontact, Amarok and TV |
20 |
Browser and restart them immediately afterwards. Memory usage has |
21 |
dropped, and desktop switching is fast again, once I went to every desktop |
22 |
so that stuff will be swapped in. BTW, vm.swappiness is set to 20. |
23 |
|
24 |
Or I do a 'swapoff -a && swapon -a', this empties the swap and also |
25 |
frees memory. But this takes _quite_ _a_ _while_. Once I measured 37 minutes |
26 |
to clear 1G of swap. |
27 |
But I do not remember how much memory was being used then, so I tried |
28 |
again, after closing memory-intensive applications: |
29 |
|
30 |
weird ~ # free -m |
31 |
total used free shared buffers cached |
32 |
Mem: 5721 4184 1536 0 34 111 |
33 |
-/+ buffers/cache: 4039 1682 |
34 |
Swap: 4093 885 3208 |
35 |
|
36 |
weird ~ # time swapoff -a |
37 |
|
38 |
real 27m8.757s |
39 |
user 2m12.284s |
40 |
sys 21m37.089s |
41 |
|
42 |
weird ~ # free -m |
43 |
total used free shared buffers cached |
44 |
Mem: 5721 4785 936 0 53 409 |
45 |
-/+ buffers/cache: 4322 1398 |
46 |
Swap: 0 0 0 |
47 |
|
48 |
So, 27 minutes to put 885MB of swap back into RAM, with the double amount of |
49 |
that being free RAM. I monitored with iotop, and the transfer rate started |
50 |
around 60-100 K/s, later it went higher. But the average transfer rate is |
51 |
550K/s. Shouldn't swap be, like, a little faster? |
52 |
|
53 |
My swap is on an LUKS-encrypted LVM volume, but I tried with a fresh new |
54 |
LVM volume, and it was a little better only. The SATA drive gives 80-100 |
55 |
MB/s according to hdparm -t. dd'ing /dev/zero directly to the swap volume |
56 |
gives around 100 MB/s. |
57 |
|
58 |
I'm running ck-sources 2.6.38, but I experience this for a long time |
59 |
now, and changed from tuxonice-sources (another thing that just doesn't |
60 |
work well for me) to ck-sources, to try if this would improve things. I |
61 |
even cloned a .config from some live cd, in case I had some stupid option |
62 |
activated that slowed things down. |
63 |
|
64 |
Lowering swappiness helped, as did more memory. With 3.7G of RAM, KDE4 |
65 |
was barely usable. But two years ago, when this PC was new, I ran KDE 4.2 on |
66 |
x86, compiled all except OpenOffice in tmpfs, often had a virtual |
67 |
machine running in VMplayer, and it was fine. |
68 |
|
69 |
I sure run lots of applications (Kontact, Amarok, TV Browser which uses an |
70 |
incredible amount of RAM), some Dolphins, some Konsoles, a few Konquerors, |
71 |
about 15-26 Chromium tabs, KMyMoney, some editors, sometimes Qt Creator, and |
72 |
some Okulars. When I file some photos from my camera with Digikam, I |
73 |
normally do not close it afterwards, hoping that it gets swapped out if |
74 |
needed, and unless I start using it again this should not cost too much of |
75 |
my RAM. This has always been my style of working, I don't like to close an |
76 |
applications just after working with it. And it was fine in the past, with |
77 |
much less memory, and with a virtual machine running all of the time. |
78 |
|
79 |
Now I got another 2G of RAM, and things are better. But still, I have 800M |
80 |
of swap right now, without using special applications or VMs. At this |
81 |
moment, it's not problem, there is no paging going on. A little while ago, |
82 |
it was quite noticeable. I was running rdiff-backup stuff at that moment, |
83 |
but so I am doing now. Weird. |
84 |
Oh, even weirder: The phone just rang, and five minutes later, swap has gone |
85 |
to 860M. I was running rdiff-backup --list-increment-sizes, maybe this uses |
86 |
much memory, and caches the stuff. Now the command has finished, and paging |
87 |
has stopped. The rdiff-backup process itself does not use much memory. |
88 |
|
89 |
BTW, does anyone else's kwin use 750M? That's pretty high, I think it used |
90 |
to be more like 300M. |
91 |
|
92 |
Wonko |