| 1 |
On Thu, Dec 2, 2010 at 3:53 PM, Harry Putnam <reader@×××××××.com> wrote: |
| 2 |
> Can anyone tell me how determine what these kind of useless names |
| 3 |
> really mean? |
| 4 |
> |
| 5 |
> From df -h |
| 6 |
> Filesystem Size Used Avail Use% Mounted on |
| 7 |
> rootfs 1.9G 283M 1.6G 15% / |
| 8 |
> /dev/root 1.9G 283M 1.6G 15% / |
| 9 |
> |
| 10 |
> How are you supposed to tell what actual device these things are on. |
| 11 |
> |
| 12 |
> I know I can look in fstab... but that is something of a crap shoot |
| 13 |
> since it is user configured. |
| 14 |
|
| 15 |
And I think in this case it is related to the linux boot process, |
| 16 |
where fstab doesn't even come into play yet anyway. In fact that's |
| 17 |
probably the reason for the "makebelieve baloney" names, I would guess |
| 18 |
(more on that later). :) |
| 19 |
|
| 20 |
> So what commands will show real devices not makebelieve baloney, and |
| 21 |
> allow me to see the usage devices are put to? |
| 22 |
|
| 23 |
In addition to what others suggested (ls or readlink), you can use |
| 24 |
stat, or you can look early-on in dmesg, for example mine shows this: |
| 25 |
[ 7.139281] EXT4-fs (sda2): mounted filesystem with ordered data |
| 26 |
mode. Opts: (null) |
| 27 |
[ 7.153529] VFS: Mounted root (ext4 filesystem) readonly on device 8:2. |
| 28 |
|
| 29 |
which seems pretty conclusive. In the case of rootfs and /dev/root you |
| 30 |
can also probably just look at your kernel commandline: |
| 31 |
$ cat /proc/cmdline |
| 32 |
root=/dev/sda2 doscsi raid=noautodetect rootfstype=ext4 |
| 33 |
video=uvesafb:1280x720p-59,mtrr:2,ywrap |
| 34 |
|
| 35 |
Which shows my root device as /dev/sda2 which matches what dmesg and |
| 36 |
ls and readlink and stat tell me. |
| 37 |
|
| 38 |
If your root is a device-mapper array or you're booting on a diskless |
| 39 |
machine over the network or other more "exotic" configurations maybe |
| 40 |
this info might not be so straightforward. In my case I have a normal |
| 41 |
PC with a non-raid boot/root drive so it's easy. |
| 42 |
|
| 43 |
> Why do we use these kind of names anyway? |
| 44 |
|
| 45 |
Kernel documentation sheds some light on it. Check out |
| 46 |
/usr/src/linux/Documentation/filesystems/ramfs-rootfs-initramfs.txt in |
| 47 |
particular. The most relevant snippet from my brief perusal: |
| 48 |
|
| 49 |
[snip] |
| 50 |
What is rootfs? |
| 51 |
--------------- |
| 52 |
|
| 53 |
Rootfs is a special instance of ramfs (or tmpfs, if that's enabled), which is |
| 54 |
always present in 2.6 systems. You can't unmount rootfs for approximately the |
| 55 |
same reason you can't kill the init process; rather than having special code |
| 56 |
to check for and handle an empty list, it's smaller and simpler for the kernel |
| 57 |
to just make sure certain lists can't become empty. |
| 58 |
|
| 59 |
Most systems just mount another filesystem over rootfs and ignore it. The |
| 60 |
amount of space an empty instance of ramfs takes up is tiny. |
| 61 |
[/snip] |
Archives