[Resolvconf-devel] Bug#504409: Bug#504409: resolvconf runs too early in bootsequence - can't mkdir
Peter T. Breuer
ptb at inv.it.uc3m.es
Tue Nov 4 12:53:52 UTC 2008
"Also sprach Petter Reinholdtsen:"
> > Anyway, it runs too early. It was linked in at S07. I moved to S36
> > (after S35mountall). In rcS.d, I mean.
>
> Please provide the output from 'ls /etc/rcS.d/'. Do you use
% ls /etc/rcS.d
README S36mountall-bootclean.sh@
S01glibc.sh@ S36resolvconf@
S01mountkernfs.sh@ S37apparmor@
S02hostname.sh@ S37mountoverflowtmp@
S02readahead@ S37udev-finish@
S03splashy@ S39ifupdown@
S05bootlogd@ S39readahead-desktop@
S05keymap.sh@ S39ufw@
S06keyboard-setup@ S40netscript@
S07hdparm@ S40networking@
S07linux-restricted-modules-common@ S40shorewall@
S08hwclockfirst.sh@ S43portmap@
S08loopback@ S44nfs-common@
S09scsitools-pre.sh@ S45mountnfs.sh@
S10udev@ S45open-iscsi@
S11hwclock.sh@ S45waitnfs.sh@
S11mountdevsubfs.sh@ S46mountnfs-bootclean.sh@
S12mtab.sh@ S47lm-sensors@
S13pcmciautils@ S49console-setup@
S15module-init-tools@ S55dns-clean@
S17procps@ S55pppd-dns@
S18ifupdown-clean@ S70nviboot@
S20checkroot.sh@ S70screen-cleanup@
S22scsitools.sh@ S70x11-common@
S25brltty@ S75sudo@
S25mdadm-raid@ S80bootmisc.sh@
S26cryptdisks-early@ S80slashem-common@
S28cryptdisks@ S85urandom@
S30checkfs.sh@ S90console-screen.sh@
S31hibernate@ S90mountdebugfs@
S34fuse@ S99stop-bootlogd-single@
S35mountall.sh@
%
(you forgot to tell me to set my locale to whatever, but I assume you
can cope with this particular alphanumeric ordering ;).
> dependency based boot sequencing?
Que? Como? Was? Huh? Warryouonabaht? My boot sequence runs a typical
unix sysvinit and in consequence it executes the stuff in /etc/rcS.d
in the order you see above! Dependencies on WHAT, do you mean? Would
you like the bit from inittab where it defines the order?
# /etc/init.d executes the S and K scripts upon change
# of runlevel.
#
# Runlevel 0 is halt.
# Runlevel 1 is single-user.
# Runlevels 2-5 are multi-user.
# Runlevel 6 is reboot.
l0:0:wait:/etc/init.d/rc 0
l1:1:wait:/etc/init.d/rc 1
l2:2:wait:/etc/init.d/rc 2
l3:3:wait:/etc/init.d/rc 3
l4:4:wait:/etc/init.d/rc 4
l5:5:wait:/etc/init.d/rc 5
l6:6:wait:/etc/init.d/rc 6
# Normally not reached, but fallthrough in case of emergency.
z6:6:respawn:/sbin/sulogin
Maybe you are asking if my scripts do some miraculous semi-dynamic
auto-reordering based on some declared dependencies of something on
something. Maybe debian package dependencies. That might be a neat
trick. But I don't know of any such thing being installed.
> If so, what is the output from
> /usr/share/insserv/check-initd-order?
Never heard of it, but since you ask:
% /usr/share/insserv/check-initd-order
LSB header missing in /etc/rcS.d/S07linux-restricted-modules-common
LSB header missing in /etc/rcS.d/S08loopback
LSB header missing in /etc/rcS.d/S10udev
Incorrect order checkroot at 20 > /etc/rcS.d/S15module-init-tools15
Incorrect order mdadm at 25 > /etc/rcS.d/S28cryptdisks28
Incorrect order mdadm at 25 > /etc/rcS.d/S31hibernate31
LSB header missing in /etc/rcS.d/S37udev-finish
Incorrect order console-screen at 90 > /etc/rcS.d/S49console-setup49
LSB header missing in /etc/rc2.d/S10powernowd.early
LSB header missing in /etc/rc2.d/S10xserver-xorg-input-wacom
LSB header missing in /etc/rc2.d/S20transproxy
%
That looks like it may be checking the debian header/comments in some
of these files, and comparing with the symlinks.
Oh .. BTW S08loopback is in the wrong place too. It attempts to do a
mkdir /var/run/network yet runs before /var is mounted too.
I've modded it to not complain but fallback to ifconfig lo 127.0.0.1
silently. If I hadn't done that I would have moved it to S036 too, but
then I'd have had to see what else needed lo and taken that into
account ... so I didn't.
> Happy hacking,
> --
> Petter Reinholdtsen
>
Peter
More information about the Resolvconf-devel
mailing list