[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

unexpected changes - live-bottom/13swap running by default



For the last 2-3 weeks, I've been working on a step-by-step document to get 
paella up and running for people.  I have decided to use live-helper to make 
the nfsroot installer, since my self-made live nfs installer was really 
hacky, and there was not any decent alternative.  The document I'm working on 
is located here.  http://paella.berlios.de/quickstart-vbox.html  This 
document might make a good example of how to test the "net" images with 
virtualization, and may be hacked to fit on the 
http://wiki.debian.org/DebianLive/Virtualization page.  I may do this myself, 
when I'm done with getting paella up and running again.

Recently, with the updates, I've noticed a peculiar change that really 
surprised me.  It seems that the new live-initramfs package searches for swap 
partitions on the system that's being booted, and automatically enables them.  
I would've probably never noticed this on another system, or if I was booting 
from a live cd.  My installer, however, noticed immediately, as I couldn't 
repartition a disk due to a swap partition already being enabled.  After a 
bit of searching, I found where the problem was, and how to disable it.  I 
wound up putting 'noswap' in the LH_BOOTAPPEND_LIVE variable.  It also seems 
that the live-bottom/13swap script has been there for a while, so maybe the 
actual change happened somewhere in live-helper.  I'm not sure, and I don't 
want to spend time tracking this further, as I'm not as familiar with these 
scripts as y'all are.

My concern is that 'noswap' isn't enabled by default (if what I'm describing 
isn't a bug).  I've been under the impression that a live system shouldn't 
write to any of the drives on the system by default.  I can't find any 
documentation that states whether the policy of a live system should write to 
the existing drives or not.  I think that it's possible that you're not aware 
of this change, since the script has been in live-initramfs for a while, but 
I haven't had this problem occur yet.

On a side note,  I've been having problems with the /etc/resolv.conf file.  In 
my setup, mentioned above, the system that builds the live image is running 
bind, and it's resolv.conf uses "nameserver 127.0.0.1".  I can't seem to 
replace this file using the local-includes.  My solution was to execute 
dhclient in the /etc/rc.local on the live system.  I would rather not replace 
the /etc/resolv.conf during the build process, anyway.  From what I 
understand, the dhclient keeps the interface from being deconfigured if there 
is an nfsroot running from it (this is bad english, but it's late).  I am 
curious as to why the interface is set to manual instead of dhcp in 
the /etc/network/interfaces file.  Also, I would like to know the best way to 
enable dhcp on the live system.  I don't really like putting the dhclient 
command in the /etc/rc.local file.

Here's my configuration if you're interested (I haven't committed the 'noswap' 
option yet).  

http://paella.berlios.de/netinst/paella/live-config/

or

svn://svn.berlios.de/paella/website/netinst/paella/live-config

Thanks for making it easier to concentrate on other things! :)

-- 
Thanks:
Joseph Rawson

Attachment: signature.asc
Description: This is a digitally signed message part.


Reply to: