FAI Stable release 3.4.4 (was: FAI Stable release 3.4.3 available)

[Michael Tautschnig]

> * Michael Prokop <mika at grml.org> [Sun Oct 10, 2010 at 10:48:14PM +0200]:
> > * Michael Prokop <mika at grml.org> [Thu Oct 07, 2010 at 11:59:43PM +0200]:
> 
> > > And finally we have to verify the hostname<->HOSTNAME issue, see
> > > https://lists.uni-koeln.de/pipermail/linux-fai/2010-October/008400.html
> 
> > I think I've news regarding this issue.
> > Can someone please do me a favour and test the following fix?
> 
> > Just disable the 'export NOHOSTS="Yes"' in conf/live.conf
> > and drop the NOHOSTS occurences in script make-fai-nfsroot.
> > Finally rebuild the initramfs using update-initramfs and this
> > should bring hostname=... back then without breaking anything.
> 
> > Anyone willing to help us in testing this?
> 
> Ok, no one showed up in time, invested further time on my own...
> 
> Sadly this is a very annoying situation for us: The reworked
> live-initramfs layout of live-boot/live-config (which is the stuff
> present in squeeze) breaks assumptions for FAI. As FAI provides its
> own /etc/init.d/rcS and doesn't go through normal runlevels
> specifying live-config=hostname doesn't help and the (dropped)
> NOHOSTS="Yes" isn't considered at all.
> 
> Reworking FAI to support old live-initramfs as well as newer
> versions requires more effort and unless I can come up with a
> solution those days I'll provide 3.4.4 without any changes WRT
> hostname<->HOSTNAME handling ASAP.
> 

I have also pointed that out in private mail, but maybe it's worth restating it
in public, just in case someone else does find a bit of time for testing this:

Would it be the single most important step to call live-config, as Karl
suggested? If this is the case, we might fairly quickly come to both a clean and
simple solution!?

> I'm considering the suggestions by michael log for inclusion into
> 3.4.4, would be great if anyone else could take a closer look at it
> as well.
> 

I hope not to have missing anything -- I have reviewed all three messages
suggesting improvements; after a review two of them already made it into the
experimental branch, one still requires further discussion (see my earlier
post).

Hope this helps,
Michael

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
Url : http://lists.uni-koeln.de/pipermail/linux-fai-devel/attachments/20101012/f160ef82/attachment.bin