1 |
On Sat, Sep 17, 2011 at 2:16 AM, Joost Roeleveld <joost@××××××××.org> wrote: |
2 |
|
3 |
> |
4 |
> Except that Redhat and Centos use LVM by default. Which will also mean that |
5 |
> "simple users" also end up using LVM. |
6 |
> Then again, they also end up with an initr* and a generic kernel for |
7 |
> everything under the sun. |
8 |
> I haven't properly looked at the kernel-configs from redhat lately, but I |
9 |
> don't think they include all the possible hardware options be default? |
10 |
> |
11 |
> |
12 |
> |
13 |
The typical mainstream binary distro approach is to: |
14 |
|
15 |
1. Build every module under the sun that won't cause more support headaches |
16 |
than benefits. |
17 |
2. Build a really smart initramfs that can find root whether it is on raid, |
18 |
lvm, luks, or on a SAN behind luks and a VPN (ok, I'm stretching it a |
19 |
little). |
20 |
3. Deploy that on everything. |
21 |
|
22 |
With an initramfs you can essentially build a completely modular kernel, |
23 |
since the kernel can rely on any module it wishes right from the start. |
24 |
However, once the initramfs is done the kernel is still a minimal size |
25 |
since unused modules don't occupy space (the initramfs wipes itself out of |
26 |
ram as its last step, unless in a debug mode). |
27 |
|
28 |
Sure, the fancy initramfs takes work, but since the install image is |
29 |
one-size-fits-all it is less maintenance in the long haul. Plus you can |
30 |
replace your motherboard and still boot the same drive image. The downside |
31 |
is that it might take an extra second or two to boot, but dracut is pretty |
32 |
fast. |
33 |
|
34 |
Honestly, if I were running a server setup I'd probably consider using an |
35 |
intiramfs. They're a lot less susceptible to being messed up if for |
36 |
whatever reason the drives get re-ordered in the BIOS (root=UUID, or with |
37 |
LVM the device names can usually be trusted). I once booted off of a rescue |
38 |
CD that for whatever reason changed the major numbers assigned to my raid |
39 |
devices and for a while I could never predict what /dev/md# my root would |
40 |
end up with. That is what started my quest to get dracut working, which |
41 |
I'll continue whenever git.kernel.org is back up... |
42 |
|
43 |
Rich |