1 |
Something _really_ weird happened to your quoting; you quoted my |
2 |
email, but your email client said you wrote it. |
3 |
|
4 |
On Mon, Dec 17, 2012 at 2:00 AM, J. Roeleveld <joost@××××××××.org> wrote: |
5 |
>> On Sun, Dec 16, 2012 at 10:39 AM, J. Roeleveld <joost@××××××××.org> wrote: |
6 |
|
7 |
^-- weird --^ |
8 |
|
9 |
>>> On Sunday, December 16, 2012 01:52:46 PM Volker Armin Hemmann wrote: |
10 |
>>>> Am Samstag, 15. Dezember 2012, 20:57:24 schrieb J. Roeleveld: |
11 |
>>>> > Even on a system with only 2 sockets, it can be useful to have NUMA |
12 |
>>>> > available. |
13 |
>>>> |
14 |
>>>> or not, because it costs you performance. |
15 |
>>> |
16 |
>>> When does it cost performance? |
17 |
>>> In all situations? |
18 |
>> |
19 |
>> It adds some additional logic to memory allocation (put an allocation |
20 |
>> near the process that uses it) and to process scheduling (keep the |
21 |
>> process near its memory, but bump it to a more distant idle core if |
22 |
>> necessary). |
23 |
> |
24 |
> That's the way it's supposed to work, yes :) |
25 |
> |
26 |
>> In all honestly, it's not a performance loss you're likely to notice, |
27 |
>> unless you're so in need of squeezing out every spare cycle that you |
28 |
>> most definitely _have_ hardware where there are disconnected memory |
29 |
>> banks. I'm not convinced it's even measurable for us mundanes and our |
30 |
>> hardware. |
31 |
> |
32 |
> I don't think I would notice it either, but as the system I have supports |
33 |
> it, I want to use it. |
34 |
> And then I want to be certain it actually supports it correctly. |
35 |
> |
36 |
> The system I'm talking about is used for testing purposes. Running |
37 |
> multiple VMs. As far as I know, Xen has support for it, just need to |
38 |
> configure it properly. |
39 |
> And for this usecase, I think NUMA with only 2 physical CPUs should make a |
40 |
> positive difference. |
41 |
|
42 |
Don't get me wrong; I was arguing that it shouldn't hurt to have it enabled. :) |
43 |
|
44 |
-- |
45 |
:wq |