1 |
On March 11, 2012 at 11:16 PM Walter Dnes <waltdnes@××××××××.org> wrote: |
2 |
|
3 |
> Not exactly your typical "remote machine", but the principle is the |
4 |
> same. I have a dedicated HTPC machine next to my 50" plasma, connected |
5 |
> by 50 feet of ethernet cable to my computer den. I use the TV as a |
6 |
> monitor when running NHL GameCenter Live. |
7 |
> |
8 |
> I have Lilo set up to "dual boot" between a "production" and an |
9 |
> "experimental" kernel. The first (i.e. default) boot option is the |
10 |
> "production" kernel. When I set up a new kernel, I try to always run it |
11 |
> as experimental. Even if the kernel panics, I don't<G>. I boot back |
12 |
> into the production kernel, and try again. Once the experimental kernel |
13 |
> has run for a couple of weeks without problems, I copy it over the |
14 |
> production kernel. |
15 |
> |
16 |
> One problem... if I build a new kernel, is there a way to get the |
17 |
> "remote machine" to boot to the non-default experimental kernel just |
18 |
> once? Any future boots to default to production (unless its a restart |
19 |
> from hibernate). |
20 |
> |
21 |
> -- |
22 |
> Walter Dnes <waltdnes@××××××××.org> |
23 |
> |
24 |
|
25 |
Unless I misunderstand you, after you issue "lilo" to write to the MBR, |
26 |
then issue: |
27 |
lilo -R experimental |
28 |
|
29 |
where experimental is the name of the kernel image you want to boot. The R |
30 |
creates a one time command which it will use the next time you boot, then |
31 |
it will be erased. |
32 |
|
33 |
And give the kernel an append statement: |
34 |
append="panic=10" |
35 |
|
36 |
so that if the kernel does not boot, you get automatically rebooted back |
37 |
into the good kernel. |
38 |
|
39 |
-- |
40 |
Happy Penguin Computers >`) |
41 |
126 Fenco Drive ( \ |
42 |
Tupelo, MS 38801 ^^ |
43 |
662-269-2706; 662-491-8613 |
44 |
support at happypenguincomputers dot com |
45 |
http://www.happypenguincomputers.com |