1 |
On Mon, Nov 15, 2010 at 07:09, Iain Buchanan <iaindb@××××××××××××.au> wrote: |
2 |
> On Tue, 2010-11-09 at 23:24 +0200, Fatih Tümen wrote: |
3 |
>> On Fri, Nov 5, 2010 at 08:45, Iain Buchanan <iaindb@××××××××××××.au> wrote: |
4 |
>> > OK so vm.swappiness seemed to help a bit but today I notice that swap |
5 |
>> > usage is up again. It's firefox: |
6 |
>> > |
7 |
>> > PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND |
8 |
>> > 14072 iain 20 0 1369m 897m 15m S 3 29.5 113:14.91 firefox |
9 |
>> > |
10 |
>> > I think that's 1.3Gb + 900Mb... sounds like a memory leak to me. |
11 |
>> > |
12 |
>> > Anyone else run firefox for 113+ hours? I'm using 3.6.9-r1. |
13 |
>> > |
14 |
>> |
15 |
>> 1.3G is the grant total of Res and Swap. You need to read man top |
16 |
>> before judging not-entirely-accurate values reported by top. |
17 |
> |
18 |
> judging? I only said "I think"! |
19 |
> |
20 |
> sure, top has it's quirks, but it's ok for comparing against itself. |
21 |
> |
22 |
>> 900M is resident on your main memory. '113+ hours' is not a decent |
23 |
>> information to draw conclusion from. Running firefox for 113+ hours |
24 |
>> with a single tab on a text-only website is not same as running dozens |
25 |
>> of tabs with dozens of multimedia/embedded objects. |
26 |
> |
27 |
> sure, but running it for 10 or 100 or 1000 hours should produce roughly |
28 |
> the same characteristics for the same browsing behaviour if all other |
29 |
> things are equal. A few months ago this didn't cause any issues at all, |
30 |
> now I'm seeing high swap usage. I usually never use my 3G of physical |
31 |
> RAM. |
32 |
> |
33 |
|
34 |
Can you recall what significant change have you made to the system? |
35 |
For emerged packages you can try smth like genlop --list --date 1 |
36 |
month ago and then check against the versions upgraded from. |
37 |
|
38 |
> Again today I see it is using about 900Mb in total, which seems quite |
39 |
> large. vm.swappiness is set to 0. I've upgraded firefox to 3.6.12. |
40 |
> |
41 |
> I had to reboot, but I'll check the usual statistics next time I see it. |
42 |
> |
43 |
|
44 |
You say swappiness is set to 0 but dont give any swap usage info. If |
45 |
there is any swap usage while swapiness is 0 then it would be weird |
46 |
and we could blame it on the kernel. |
47 |
|
48 |
I just googled mem usage firefox as I am running out of ideas. It seem |
49 |
like you are not the only one complaining about this. Take a look at |
50 |
these top results. There are some tweaking advice, see it they work |
51 |
for you. |
52 |
http://kb.mozillazine.org/Memory_Leak |
53 |
http://support.mozilla.com/en-US/kb/high+memory+usage |
54 |
|
55 |
Chromium, which I have been happily using for almost a year now, has a |
56 |
task manager which shows mem usage of every extension and tab. If |
57 |
firefox has switched to multiprocessing, which was a feature plan some |
58 |
time ago, similar tool should likely be available for firefox as well. |
59 |
|
60 |
-- |
61 |
Fatih |