At least for the nonce, I got the /usr problem 'licked'... 42.529749] XFS (sdc3): Mounting Filesystem [ 42.598586] XFS (sdc3): Ending clean mount Kernel logging (ksyslog) stopped. Kernel log daemon terminating. Boot logging started on /dev/tty1(/dev/console) at Sun Feb 17 11:38:57 2013 <notice -- Feb 17 11:38:57.733378000> service boot.sysctl start <notice -- Feb 17 11:38:57.742041000> service boot.udev start <notice -- Feb 17 11:38:57.747212000> service boot.usr-mount start Applying sysctl settings* Applying /lib/sysctl.d/-sysctl.conf- ... net.ipv4.icmp_echo_ignore_broadcasts = 1 net.ipv4.conf.all.rp_filter = 1 net.ipv4.conf.default.promote_secondaries = 1 net.ipv4.conf.all.promote_secondaries = 1 fs.inotify.max_user_watches = 65536 kernel.sysrq = 176 dev.cdrom.autoclose = 0 * Applying /etc/sysctl.conf ... net.ipv4.ip_forward = 0 done <notice -- Feb 17 11:38:57.933526000> service boot.sysctl done Starting udevd: done Loading drivers, configuring devices: <notice -- Feb 17 11:38:57.956211000> service boot.usr-mount done <<<<-------- udevd[195]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /etc/udev/rules.d/50-euvccam.rules:1 udevd[195]: SYSFS{}= will be removed in a future udev version, please use ATTR{}= to match the event device, or ATTRS{}= to match a parent device, in /etc/udev/rules.d/50-euvccam.rules:2 -------------- Now WHY you guys couldn't figure that out, I dunno. BUT it creates a problem. My system boots too fast for LVM.... And systemd isn't paying attention to the Sxxboot orders -- though I know it doesn't use those scripts it should at least mimic their order damnit! Those orders are there for a reason. I have on the boot scripts a desired order: S01boot.sysctl S01boot.udev S01boot.usr-mount S03boot.rootfsck S04boot.clock S05boot.device-mapper S05boot.loadmodules S06boot.cgroup S06boot.lvm S09boot.localfs S11boot.cleanup S11boot.klog S11boot.localnet S11boot.swap S11setserial S12boot.isapnp S12boot.ldconfig S12boot.sysstat (at least that's how chkconfig ordered them based on the dependencies. yeah, /usr-mount happens before rootfsck -- because it's supposed to be mounted before anything happens like root-fsck... or at least that's what i'd like to say, but I didn't plan it that way -- I gave them equal priority, since my system doesn't need to run fsck's on the disks (they are a test to see if the device is there...and if the root device isn't there, then how would code to run the test have been loaded? ;-) But the problem comes later: My notes "in line with the log"....... share/virtualbox/VBoxCreateUSBNode.sh 189 2 00': No such file or directory udevd[516]: failed to execute '/usr/share/virtualbox/VBoxCreateUSBNode.sh' '/usr/share/virtualbox/VBoxCreateUSBNode.sh 189 3 00': No such file or directory -------------- Sorry, but /usr/share isn't required at boot I'm told. Someone should tell the VBOX people ... not that I'm running their product anyway -- I couldn't get it to restart after creating one -- tried multiple times -- but always had to recreate the machine... that and never could get networking to work. Sigh. done <notice -- Feb 17 11:39:00.38615000> service boot.udev done <notice -- Feb 17 11:39:00.39151000> service boot.loadmodules start <notice -- Feb 17 11:39:00.40231000> service boot.rootfsck start ---- Remember in the init boot script order, rootfsck comes at #3 and load modules shouldn't happen until #5. But AFAIK, this isn't problematic... Loading required kernel modules done <notice -- Feb 17 11:39:00.61188000> service boot.loadmodules done Activating swap-devices in /etc/fstab... doneroot file system (/) is NOT being checked. <notice -- Feb 17 11:39:00.107218000> service boot.rootfsck done <notice -- Feb 17 11:39:00.111265000> service boot.cgroup start <notice -- Feb 17 11:39:00.120248000> service boot.clock start mounting cgroup file systems... cpuset debug cpu cpuacct devices freezer blkio net_priodone <notice -- Feb 17 11:39:00.257108000> service boot.cgroup done Set System Time to the current Hardware Clockdone <notice -- Feb 17 19:39:00.334317000> service boot.clock done <notice -- Feb 17 19:39:00.334838000> service boot.device-mapper start <notice -- Feb 17 19:39:00.340342000> service boot.localfs start Activating device mapper...done <notice -- Feb 17 19:39:00.352882000> service boot.device-mapper done ===================================== Problem -- device mapper started (step 5 in initscripts... obviously not in systemd) BUT before it gets to 'done' boot.localfs starts. that's not good.... Waiting for /dev/Home+Space/Home /dev/Home+Space/Share /dev/Home+Space/Home.diff /dev/Media/Media /dev/Home+Space/Squid_Cache /dev/Backups/Backups /dev/Home+Space/Media_Back .............................. timeout! --- local fs shouldn't have started berfore device mapper had ***settled** but we don'd see that message till later.... Checking file systems... fsck from util-linux 2.20.1 donedone Mounting local file systems... mount: /dev/sdc6 already mounted on /usr /dev/sdc2 on /var type xfs (rw,noatime,nodiratime,swalloc,largeio,logbsize=256k) /dev/sdc3 on /boot type xfs (rw,noatime,nodiratime,swalloc,largeio,allocsize=1m) /var/rtmp on /tmp type none (rw,bind) mount: special device /dev/Home+Space/Home does not exist mount: special device /dev/Home+Space/Share does not exist mount: special device /dev/Home+Space/Home.diff does not exist mount: special device /dev/Media/Media does not exist mount: special device /dev/Home+Space/Squid_Cache does not exist mount: special device /dev/Backups/Backups does not exist mount: mount point /backups/Media does not exist mount: special device /home/share does not exist /Media on /Media type none (rw,bind) failed <notice -- Feb 17 19:39:32.829649000> service boot.localfs done ============================================================================ GREAT... local fs is done,.... but all the lvm volumes failed: wonder why?..... hmmmm... the local-fs-target wants? fsck-root.service@ remount-rootfs.service@ var-lock.mount@ media.mount@ systemd-remount-api-vfs.service@ var-run.mount@ But no lvm--- the only references I find to it SOUNDS like it should have worked...(but obviously I have little experience with systemd).... Ishtar:/lib/systemd/system> grep -r lvm * fsck@.service:After=systemd-readahead-collect.service systemd-readahead-replay.service %i.device lvm.service md.service dmraid.service local-fs-pre.target:After=md.service lvm.service dmraid.service I see it should have happened BEFORE the localfsck and it is in a local-fs-pre-target, but that seems to be getting ignored. <notice -- Feb 17 19:39:32.830172000> service boot.cleanup start <notice -- Feb 17 19:39:32.842134000> service boot.klog start <notice -- Feb 17 19:39:32.845642000> service boot.localnet start <notice -- Feb 17 19:39:32.846799000> service boot.lvm start ============================================================================== OH LOOKIE AT THAT!!! lvm is starting now... it was S05 in the boot script order and localfs was in the S09 -- so can't systemd get a clue...ok...expecting too much, sorry. But people.. the reason you couldn't figure out how to do user before everything else, is because systemd does EVERYTHING at the same time and ignores all ordering--- AT LEAST , that appears to be what it is doing. ===============Notice...about a page further down... <notice -- Feb 17 19:39:32.852691000> service setserial start Configuring serial ports... /dev/ttyS0 at 0x03f8 (irq = 4) is a 16550A /dev/ttyS1 at 0x02f8 (irq = 3) is a 16550A Configured serial ports done <notice -- Feb 17 19:39:32.913597000> service setserial done <notice -- Feb 17 19:39:33.81793000> service boot.isapnp start <notice -- Feb 17 19:39:33.107182000> service boot.isapnp done Using boot-specified hostname 'Ishtar' Setting up hostname 'Ishtar'done Setting up NIS domainname 'sc.tlinx.org'/etc/init.d/boot.localnet: line 52: domainname: command not found done Setting up loopback interface done <notice -- Feb 17 19:39:33.418268000> service boot.localnet done <notice -- Feb 17 19:39:33.442066000> service boot.cleanup done <notice -- Feb 17 19:39:33.442210000> service boot.sysstat start <notice -- Feb 17 19:39:33.612498000> killproc: kill(165,29) adding: boot-130217.1939 (deflated 79%) test of /var/log/boot.msg.zip OK done <notice -- Feb 17 19:39:33.613730000> service boot.klog done Running sadc done Cannot open font file lat1-08.psfu.gz /etc/init.d/boot.sysstat: line 55: /bin/unicode-start: No such file or directory FATAL: Module bond0 not found. ----------------------------------- Where does this come from?^^^ it needs to be fixed -- if it looked in the loaded modules /sys/modules BEFORE trying to load it, it would find that it is already in the kernel?!?! But this is what I really would have like to have seen above: Waiting for udev to settle... Scanning for LVM volume groups... Reading all physical volumes. This may take a while... Found volume group "Media" using metadata type lvm2 Found volume group "Backups" using metadata type lvm2 Found volume group "Home+Space" using metadata type lvm2 Activating LVM volume groups... 1 logical volume(s) in volume group "Media" now active <notice -- Feb 17 19:39:33.819432000> service boot.sysstat done 1 logical volume(s) in volume group "Backups" now active 20 logical volume(s) in volume group "Home+Space" now active done <notice -- Feb 17 19:39:37.442276000> service boot.lvm done ----------------------------- Why oh WHY did it have to try to start localfs before lvm finished? The rest is fairly uninteresting.... So do I have to hand port over all the boot scripts with interlocks to get systemd to pay attention to boot order? Or are the above references to lvm not enough? I don't know -- I didn't set them up ... just found them like that. there doesn't seem to be an "lvm.service".... maybe that's part of the problem. But init would have failed to boot right then and there and stopped.. and good or bad -- lvm... actually this is a bit disturbing for anyone who like reliability -- the fsck for the local LVM volumes didn't fail and cause a stop in single user ==> LVM wasn't up yet. But the system came up just fine aside from the above, as all the system disks were mounted -- though some services failed (dependent on the lvm volumes that didn't quite make it...)... I was able to restart those by hand... So wazup with the ordering?...shouldn't there be lvm services to correspond to the boot.lvm and boot.lvm_monitor? <notice -- Feb 17 19:39:37.449508000> service boot.swap start Activating remaining swap-devices in /etc/fstab... done <notice -- Feb 17 19:39:37.475149000> service boot.swap done <notice -- Feb 17 19:39:37.490474000> service boot.ldconfig start <notice -- Feb 17 19:39:37.503860000> service boot.ldconfig done System Boot Control: The system has been set up System Boot Control: Running /etc/init.d/boot.local Cannot open font file lat1-08.psfu.gz /etc/init.d/boot.local: line 29: unicode-start: command not found failed INIT: Entering runlevel: 3 Master Resource Control: previous runlevel: N, switching to runlevel: 3 Master Resource Control: Running /etc/init.d/before.local done <notice -- Feb 17 19:39:37.759743000> service earlysyslog start <notice -- Feb 17 19:39:37.770688000> service acpid start <notice -- Feb 17 19:39:37.783250000> service dbus start <notice -- Feb 17 19:39:37.794894000> service cpufreq start <notice -- Feb 17 19:39:37.821673000> startproc: execve (/sbin/acpid) [ /sbin/acpid.... ---------------------------------------------------------------------------------------- and rest LOOKS like output of scripts being run by init instead of systemd. -- To unsubscribe, e-mail: opensuse-factory+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-factory+owner@opensuse.org