1 |
Am 10.03.2013 13:48, schrieb Florian Philipp: |
2 |
> Am 10.03.2013 00:53, schrieb cosmoslx lin:> |
3 |
>> 2013/3/10 Volker Armin Hemmann <volkerarmin@××××××××××.com |
4 |
>> <mailto:volkerarmin@××××××××××.com>> |
5 |
>> |
6 |
>> Am 09.03.2013 19:15, schrieb Florian Philipp: |
7 |
>> > Hi list! |
8 |
>> > |
9 |
>> > Whenever I do sequential IO for a long stretch of time (e.g. |
10 |
>> md5summing |
11 |
>> > 40 GB), I'm experiencing high load (ca. 6 on a 4 CPU system) and |
12 |
>> > temporary freezes of most applications. For example, switching between |
13 |
>> > tabs in konsole sometimes takes more than 2 seconds. |
14 |
>> > |
15 |
> [...] |
16 |
>> > |
17 |
>> congratulation. You hit 'the bug'. Been around for ages, but for magical |
18 |
>> reasons kernel dev are unable to see or unable to do something about it. |
19 |
>> If you are using a vanilla kernel, posting on lkml might be the right |
20 |
>> thing to do. |
21 |
>> |
22 |
>> I have try the BFQ patch outside of the kernel mainline, it works well. |
23 |
>> Maybe you would like to see: |
24 |
>> http://algo.ing.unimo.it/people/paolo/disk_sched/ |
25 |
>> |
26 |
> |
27 |
> Thanks. I'll try pf-sources which should to include that patch set. |
28 |
> |
29 |
> Regards, |
30 |
> Florian Philipp |
31 |
> |
32 |
|
33 |
pf-sources (only BFQ enabled, not BFS) helped somewhat. kswapd still |
34 |
likes to keep one CPU busy when IO is happening but I suspect cleancache |
35 |
to be the culprit for that. I'll test without it later. |
36 |
|
37 |
Regards, |
38 |
Florian Philipp |