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

Re: Making Linux disaster recovery user friendly

Previously Ron O'Hara wrote:
> As a longtime Slackware user,

I'll refrain from commenting on that ;)

Anyway on your suggestion (I've passed it on to debian-devel so they
can discuss this, so this is purely a personal opinion): it won't
work. There is no way you can guarantee that a user error, disk crash or
something else will leave a safe kernel in the same place and intact.

You also seem to suggest that the `safe' option should do more then just
boot a known-safe kernel but also but in a special rescue mode, so your
propsed change doesn't really help.

There is only one method to have a rescue method that always works: make
a rescue disk (or use the installation-CD) and boot from that if the
normal system fails. Then ask the user to write-protect that disk and
store it safely. 

And it so happens that most (I think, if not all) Linux distributions
give the user exactly that option during system installation.

On your thoughts about upgrading: why would it matter what kernel you
use during upgrading? Or how you boot your system? Upgrading should be a
matter if having a good tool to do the upgrade; it has nothing to do
with what kernel you use or how you boot your system.

With Debian I can upgrade from an old a.out system to a state-of-the-art
glibc2 system without rebooting, no matter what kernel (as long as it
supports the ELF binary format of course). I don't know enough about
the other distributions to be able to say if this holds for them as
well, but it does show that a special Upgrade-boot method is not


This combination of bytes forms a message written to you by Wichert Akkerman.
E-Mail: wakkerma@cs.leidenuniv.nl
WWW: http://www.wi.leidenuniv.nl/~wichert/

Attachment: pgp_5U6QJmv5A.pgp
Description: PGP signature

Reply to: