| 1 |
On Thursday, September 15, 2011 02:29:20 PM Rich Freeman wrote: |
| 2 |
> On Thu, Sep 15, 2011 at 11:07 AM, Zac Medico <zmedico@g.o> wrote: |
| 3 |
> > On 09/15/2011 07:33 AM, Joost Roeleveld wrote: |
| 4 |
> > > The use for an initrd/initramfs/... will create an additional layer |
| 5 |
> > > of |
| 6 |
> > > complexity a lot of us users are not really waiting for, especially |
| 7 |
> > > as we> |
| 8 |
> > are |
| 9 |
> > |
| 10 |
> > > not seeing any issues with our current systems. |
| 11 |
> > |
| 12 |
> > Like it or not, it's the simplest possible solution if you want separate |
| 13 |
> > /usr. The plan is to provide a minimal initramfs that won't contain any |
| 14 |
> > modules, so it won't have to be rebuilt for each kernel. See the "/usr |
| 15 |
> |
| 16 |
> > vs. initramfs redux" thread: |
| 17 |
> It should be noted that the alternative is to use a more full-featured |
| 18 |
> initramfs like dracut, which will also be updated to support mounting /usr |
| 19 |
> (it already parses /etc/fstab to remount root anyway). |
| 20 |
> |
| 21 |
> The minimal initramfs will not contain mdadm or lvm tools, so it is |
| 22 |
> basically suitable for booting systems that don't currently require an |
| 23 |
> initramfs. Of course, something like dracut is more likely to require you |
| 24 |
> to rebuild the initramfs every time you update your kernel, and won't simply |
| 25 |
> be a static image like the simplified one. |
| 26 |
> |
| 27 |
> The simplified initramfs could be compiled into the kernel as Zac suggests |
| 28 |
> (this is probably the most foolproof method), or it could be loaded from |
| 29 |
> /boot using the appropriate grub settings. |
| 30 |
|
| 31 |
Is there an option in Grub to add a "default" initramfs that is used for all |
| 32 |
boot-options that can be overriden per boot-set? |
| 33 |
|
| 34 |
In other words, if I don't specify an initramfs for a kernel, that this |
| 35 |
default is then automatically applied? |
| 36 |
And will this then also work when using Xen where the kernel is already passed |
| 37 |
as a module? |
| 38 |
|
| 39 |
> Note that dracut does drop you to a shell if it fails (this is |
| 40 |
> configurable), but by default this shell is dash, not bash. No doubt it |
| 41 |
> would work fine either way, but bash is likely to be a little slower. I |
| 42 |
> don't think RAM use is likely to be a problem - it should be completely |
| 43 |
> de-allocated before init runs. |
| 44 |
|
| 45 |
It is my understanding all the options need to be specified every time dracut |
| 46 |
is run to create an initramfs. If this becomes mandatory, will this be added |
| 47 |
to the "make" script of the kernel-sources and as such, make this more |
| 48 |
specific? |
| 49 |
|
| 50 |
Another issue arrises where some of the tools are updated that are also in the |
| 51 |
initramfs. Will we then still need to remember to also update the initramfs if |
| 52 |
these are needed? |
| 53 |
|
| 54 |
My server currently uses mdadm raid1 for /, /boot and swap and raid1+lvm for |
| 55 |
the rest. This works without the need of an initramfs. |
| 56 |
|
| 57 |
Will this still work? Or will I need to be using dracut instead? |
| 58 |
|
| 59 |
Please note, I'm not the only one using this option as it was taking directly |
| 60 |
from the Gentoo guides: |
| 61 |
http://www.gentoo.org/doc/en/gentoo-x86+raid+lvm2-quickinstall.xml |
| 62 |
|
| 63 |
-- |
| 64 |
Joost |
Archives