I'm upgrading my network front end from etch to lenny, now that security support is being dropped in a few days. As I've done before (from sarge to etch, for example) I copied the entire partition containing my system using tar/untar (and it *is* all in one partition). edited the copied / etc/fstab, checked that the copy booted, rebooted the original system, chrooted to the copy, and started upgrading.
Why this convoluted scheme? Because this way I still have an unupgraded working system incase the upgrade fails. (And it has failed in the past). I upgrade inside the chroot so the unupgraded system can go on serving my users while the upgrade happens. This time I encountered a new problem: dpkg-preconfigure: unable to open stdin Now I presume this is because /dev is absent from the chroot and it's trying to use /dev/fd/0 -- the device that acts like stdin. What do I do to set things up so that I have a valid /dev? The /dev outside the chroot is set up by udev. Is there some way of mounting it from inside the chroot? Is there some way of duplicate-mounting it from outside so it's visible inside? And there are probably similar problems with /proc, though I haven't hit them yet. -- hendrik -- To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
