1 |
On Sat, 2007-09-22 at 17:22 +0200, Volker Armin Hemmann wrote: |
2 |
> On Samstag, 22. September 2007, Richard Freeman wrote: |
3 |
> > P.V.Anthony wrote: |
4 |
> > >> If following the old rule, 8G should be allocated for swap. I feel |
5 |
> > >> that is too much. Does 2.6 kernel really need so much of swap with 4G |
6 |
> > >> of ram? |
7 |
> > >> |
8 |
> > >> Was thinking of just using a 1G swap file for safety. Please share |
9 |
> > >> some thoughts on the this swap size issue. |
10 |
> > > |
11 |
> > > Please ignore this email. It looks like I have asked something similar |
12 |
> > > to this before. I will read the old thread. |
13 |
> > |
14 |
> > That's Ok, I got a chuckle out of it. You have Duncan who doesn't use |
15 |
> > swap at all (I think), and you have me: |
16 |
> > |
17 |
> > total used free shared buffers cached |
18 |
> > Mem: 2058448 2041388 17060 0 82084 420860 |
19 |
> > -/+ buffers/cache: 1538444 520004 |
20 |
> > Swap: 17514480 1152972 16361508 |
21 |
> > |
22 |
> > I guess there is a happy medium. But what else am I going to do with |
23 |
> > that odd space that doesn't fit easily into a RAID-5? I figure that if |
24 |
> > the kernel can find a use for it I might as well let it... :) I |
25 |
> > probably have 50G more of unpartitioned space lying around since I've |
26 |
> > installed my 2 RAID-5s on non-identical drives. I guess I'll just have |
27 |
> > to wait until ZFS takes off on linux... :) |
28 |
> |
29 |
> why? zfs is slow and is mixing things that should be in different layers. One |
30 |
> argument against reiser4 always was 'it violates the layering' - well this is |
31 |
> even more true for zfs. |
32 |
> |
33 |
> And from this numbers: |
34 |
> http://tastic.brillig.org/~jwb/zfs-xfs-ext4.html |
35 |
> |
36 |
> it doesn't look so great. |
37 |
Maybe it's just me, but I tend to take comparisons of apples and oranges |
38 |
with a salt block. To make that a useful comparison, the kernel |
39 |
parameters would have to be shown as well as the level of support for |
40 |
the various hardware. The second is his write-up does not match his |
41 |
numbers. He writes ZFS "...has very bad sequential transfer with |
42 |
hardware RAID and appalling sequential transfer with software RAID" but |
43 |
fails to mention how equally bad EXT4 was at the same task. XFS does |
44 |
appear to blow them both out of the water in this area. The rest of the |
45 |
tests have ZFS being roughly equal to (or in some cases better then) |
46 |
then the other two options. |
47 |
|
48 |
Maybe the next time, the test will be run using the userland ZFS on the |
49 |
same system as the EXT4 and XFS system. At least that way you can |
50 |
adjust the numbers for the overhead of userland talking to the kernel. |
51 |
Apples to Apples with a handicap is far more accurate then apples to |
52 |
oranges. |
53 |
|
54 |
I don't know enough about the layers you mention so I can't say whether |
55 |
that's a real problem or not. |
56 |
|
57 |
|
58 |
|
59 |
-- |
60 |
gentoo-amd64@g.o mailing list |