1 |
On Tuesday 31 October 2006 11:04, Uwe Thiem wrote: |
2 |
> On 31 October 2006 09:17, Alan McKinnon wrote: |
3 |
> > I find it useful to keep in mind that XFS is a file-system (i.e. a |
4 |
> > system for files), and not necessarily a severly disk-bound |
5 |
> > filesystem |
6 |
> |
7 |
> Would you mind to elaborate on this? I simply do not get your point. |
8 |
|
9 |
Historically SGI was very strong in graphics, and those applicatiosn |
10 |
tended to generate massive amounts of temporary files that had a short |
11 |
life and only the final version needs to be written to persistent |
12 |
storage, very well suited to aggressive caching and other similar |
13 |
speedups. |
14 |
|
15 |
SGI's engineers could get away with this because they could guarantee |
16 |
that power loss to the machine wouldn't happen, so the potential data |
17 |
loss on a power outage didn't happen either. This sounds a bit weird to |
18 |
those of us raised on Intel where we pay close attention to getting |
19 |
everything on disk ASAP with as little performance loss as possible, |
20 |
but it's a perfectly reasonable system for an engineer to implement on |
21 |
the kind of hardware SGI were building. |
22 |
|
23 |
That's why I say XFS is designed to not be tightly bound to the physical |
24 |
disk if the admin chooses to set it up that way, and the file system |
25 |
becomes more of a collection of directories and files that might never |
26 |
even be stored on a disk at all |
27 |
|
28 |
alan |
29 |
-- |
30 |
gentoo-user@g.o mailing list |