Hi,
My entire /usr directory is completely empty! It was working, I did a reset and on restart, I started to get all sorts of errors, the drives didn't mount and on checking, /usr is devoid of absolutely everything!
Does anyone know if there is anyway back from this catastrophe? Everything else is there, just completely nothing in /usr!
TTFN
Paul
On Fri, Nov 26, 2004 at 07:33:34PM +0000, Paul wrote:
My entire /usr directory is completely empty! It was working, I did a reset and on restart, I started to get all sorts of errors, the drives didn't mount and on checking, /usr is devoid of absolutely everything! Does anyone know if there is anyway back from this catastrophe? Everything else is there, just completely nothing in /usr!
Did you have that as a separate partition?
On Fri, Nov 26, 2004 at 07:33:34PM +0000, Paul wrote:
My entire /usr directory is completely empty! It was working, I did a reset and on restart, I started to get all sorts of errors, the drives didn't mount and on checking, /usr is devoid of absolutely everything!
Is /usr empty because it didn't mount. Don't panic until you've checked the partition tables and fsck
Hi,
On Fri, Nov 26, 2004 at 07:33:34PM +0000, Paul wrote:
My entire /usr directory is completely empty! It was working, I did a reset and on restart, I started to get all sorts of errors, the drives didn't mount and on checking, /usr is devoid of absolutely everything!
Is /usr empty because it didn't mount. Don't panic until you've checked the partition tables and fsck
Everything other than /usr is fine. The only thing which doesn't look right is that in fstab, the line for user is LABEL=/usr, whereas everything else is /var (and so on).
How do I check the partition tables? Remember, I have nothing in /usr
TTFN
Paul
-- Get your free @ukpost.com account now http://www.ukpost.com/
On Fri, 2004-11-26 at 21:46 +0000, Paul F. Johnson wrote:
Hi,
On Fri, Nov 26, 2004 at 07:33:34PM +0000, Paul wrote:
My entire /usr directory is completely empty! It was working, I did a reset and on restart, I started to get all sorts of errors, the drives didn't mount and on checking, /usr is devoid of absolutely everything!
Is /usr empty because it didn't mount. Don't panic until you've checked the partition tables and fsck
Everything other than /usr is fine. The only thing which doesn't look right is that in fstab, the line for user is LABEL=/usr, whereas everything else is /var (and so on).
How do I check the partition tables? Remember, I have nothing in /usr
All the tools you would need in this sort of situation are in /sbin and /bin. Check to see what partitions should be mounted where:
/bin/cat /etc/fstab
And then look and see what you actually have mounted:
/bin/cat /proc/mounts
What you discover will define what you need to do next...
Keith.
On Fri, Nov 26, 2004 at 09:46:42PM +0000, Paul F. Johnson wrote:
Everything other than /usr is fine. The only thing which doesn't look right is that in fstab, the line for user is LABEL=/usr, whereas everything else is /var (and so on).
That looks fairly sane. fdisk /dev/hda will show you the partition tables where swap is and which one is likely /usr
On Friday 26 November 2004 16:46, Paul F. Johnson wrote:
Hi,
On Fri, Nov 26, 2004 at 07:33:34PM +0000, Paul wrote:
My entire /usr directory is completely empty! It was working, I did a reset and on restart, I started to get all sorts of errors, the drives didn't mount and on checking, /usr is devoid of absolutely everything!
Is /usr empty because it didn't mount. Don't panic until you've checked the partition tables and fsck
Everything other than /usr is fine. The only thing which doesn't look right is that in fstab, the line for user is LABEL=/usr, whereas everything else is /var (and so on).
How do I check the partition tables? Remember, I have nothing in /usr
TTFN
Paul
This sounds as if you had it setup with labels, but are not using an initrd now, which is required to run labels as opposed to direct pointers such as /dev/hda7. If you know which partition was the one that has the /usr on it, and you can do that with experimental mounts to /mnt/someplace, (after a mkdir /mnt/someplace) then mount -t ext3 /dev/hda1 /mnt/someplace, then do an ls on it and see if thats /usr, if not, umount it, and try /dev/hda2 etc until you find the /usr partition. When you find it, get rid of that LABEL= crap in your fstab and use that instead. When its fixed, reboot.
-- Get your free @ukpost.com account now http://www.ukpost.com/
On Friday 26 November 2004 16:46, Paul F. Johnson wrote:
Hi,
On Fri, Nov 26, 2004 at 07:33:34PM +0000, Paul wrote:
My entire /usr directory is completely empty! It was working, I did
a reset and on restart, I started to get all sorts of
errors, the drives didn't mount and on checking, /usr is devoid of
absolutely everything!
Is /usr empty because it didn't mount. Don't panic until you've checked the partition tables and fsck
Everything other than /usr is fine. The only thing which doesn't look right is that in fstab, the line for user is LABEL=/usr, whereas everything else is /var (and so on).
How do I check the partition tables? Remember, I have nothing in /usr
TTFN
Paul
This sounds as if you had it setup with labels, but are not using an initrd now, which is required to run labels as opposed to direct pointers such as /dev/hda7. If you know which partition was the one that has the /usr on it, and you can do that with experimental mounts to /mnt/someplace, (after a mkdir /mnt/someplace) then mount -t ext3 /dev/hda1 /mnt/someplace, then do an ls on it and see if thats /usr, if not, umount it, and try /dev/hda2 etc until you find the /usr partition. When you find it, get rid of that LABEL= crap in your fstab and use that instead. When its fixed, reboot.
No, there seems to be something that happened with yesterdays rawhide update. I just rebooted and got the message that my home dir did not exist. I went to a virtual term and looked to see what ws up. This is what I got:
Filesystem Size Used Avail Use% Mounted on /dev/hda1 97M 18M 75M 20% /boot none 244M 184K 244M 1% /dev/shm /dev/hda5 97M 18M 75M 20% /home
Well that is weird.
This is my fstab: # This file is edited by fstab-sync - see 'man fstab-sync' for details LABEL=/ / ext3 defaults 1 1 LABEL=/boot /boot ext3 defaults 1 2 none /dev/pts devpts gid=5,mode=620 0 0 none /dev/shm tmpfs defaults 0 0 LABEL=/home /home ext3 defaults 1 2 none /proc proc defaults 0 0 none /sys sysfs defaults 0 0 /dev/hda3 swap swap defaults 0 0 /dev/hdc /media/cdrecorder auto pamconsole,fscontext=system_u:object_r:removable_t,ro,exec,noauto,managed 0 0 /dev/scd0 /media/cdrecorder1 auto pamconsole,fscontext=system_u:object_r:removable_t,ro,exec,noauto,managed 0 0
A ls of /home showed that it was empty. So I mounted it. and it was there.
But I did not see anything in /boot. /boot is hda1, but / is hda2. The size for /boot is correct, but when I went to /boot it was empty. Looking at / showed that it was mounted as /boot. I mounted /boot ok, and the files are there, but the latest df shows:
[bpm]$ df -h Filesystem Size Used Avail Use% Mounted on /dev/hda1 97M 18M 75M 20% /boot none 244M 184K 244M 1% /dev/shm /dev/hda5 26G 6.2G 19G 26% /home /dev/hda1 97M 18M 75M 20% /boot [bpm]$ cd / [bpm]$ df -h . Filesystem Size Used Avail Use% Mounted on - 9.7G 5.2G 4.0G 57% /
So I'd say something is rotten.
Hi,
A ls of /home showed that it was empty. So I mounted it. and it was there.
But I did not see anything in /boot. /boot is hda1, but / is hda2. The size for /boot is correct, but when I went to /boot it was empty. Looking at / showed that it was mounted as /boot. I mounted /boot ok, and the files are there, but the latest df shows:
[bpm]$ df -h Filesystem Size Used Avail Use% Mounted on /dev/hda1 97M 18M 75M 20% /boot none 244M 184K 244M 1% /dev/shm /dev/hda5 26G 6.2G 19G 26% /home /dev/hda1 97M 18M 75M 20% /boot [bpm]$ cd / [bpm]$ df -h . Filesystem Size Used Avail Use% Mounted on
9.7G 5.2G 4.0G 57% /
So I'd say something is rotten.
I've altered fstab now so that /dev/hda5 is mounted as /usr and chnaged LABEL=/ to /dev/hda. Made no difference. It looks like fstab is just not being loaded.
Anything held on another partition or drive other than /dev/hda1 or /dev/hda2 is being ignored.
Which package do I need to roll-back (I imagine it will be initscripts) so my machine will fire up properly?
TTFN
Paul
-- Get your free @ukpost.com account now http://www.ukpost.com/
On Friday 26 November 2004 19:05, Paul F. Johnson wrote:
Hi,
A ls of /home showed that it was empty. So I mounted it. and it was there.
But I did not see anything in /boot. /boot is hda1, but / is hda2. The size for /boot is correct, but when I went to /boot it was empty. Looking at / showed that it was mounted as /boot. I mounted /boot ok, and the files are there, but the latest df shows:
[bpm]$ df -h Filesystem Size Used Avail Use% Mounted on /dev/hda1 97M 18M 75M 20% /boot none 244M 184K 244M 1% /dev/shm /dev/hda5 26G 6.2G 19G 26% /home /dev/hda1 97M 18M 75M 20% /boot [bpm]$ cd / [bpm]$ df -h . Filesystem Size Used Avail Use% Mounted on
9.7G 5.2G 4.0G 57% /
So I'd say something is rotten.
I've altered fstab now so that /dev/hda5 is mounted as /usr and chnaged LABEL=/ to /dev/hda. Made no difference. It looks like fstab is just not being loaded.
Thats patently incorrect, you must give it the partition number and even if /dev/hda was all one partition, it would still be /dev/hda1 if a filesystem has been made on it.
Anything held on another partition or drive other than /dev/hda1 or /dev/hda2 is being ignored.
Which package do I need to roll-back (I imagine it will be initscripts) so my machine will fire up properly?
NDI. Since no one else has chimed in with a similar problem, what I described earlier is still the method I would use.
TTFN
Paul
-- Get your free @ukpost.com account now http://www.ukpost.com/
On Fri, Nov 26, 2004 at 05:34:15PM -0600, Brian Millett wrote:
No, there seems to be something that happened with yesterdays rawhide update. I just rebooted and got the message that my home dir did not exist. I went to a virtual term and looked to see what ws up. This is what I got:
Filesystem Size Used Avail Use% Mounted on /dev/hda1 97M 18M 75M 20% /boot none 244M 184K 244M 1% /dev/shm /dev/hda5 97M 18M 75M 20% /home
This does not mean very much. What you will see if you will do 'cat /proc/mounts'?
For some time initscripts are seriously broken ( https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=139642 ) and /etc/mtab is messed up throwing out of kilter 'mount', 'df' and who knows what else.
It is somewhat strange that this things remains broken for so long time.
Michal
I think what OP meant. I just applied recent rawhide updates and _none_ of disk based file systems was mounted except of /. I happen to have here four of these plus two bind mounts.
Typing 'mount -a' at the root prompt immediately restored a resemblance of a sanity but I do not know yet what actually is messed up. Possibly this was a "fix" for that bug report I mentioned previously or maybe some other "improvement" is responsible. :-)
Michal
Hi,
I think what OP meant. I just applied recent rawhide updates and _none_ of disk based file systems was mounted except of /. I happen to have here four of these plus two bind mounts.
Latest initscripts are to blame. Drop back down a version and all returns to sanity, well, mostly all...
TTFN
Paul
On Sat, Nov 27, 2004 at 12:49:34AM +0000, Paul wrote:
I think what OP meant. I just applied recent rawhide updates and _none_ of disk based file systems was mounted except of /. I happen to have here four of these plus two bind mounts.
Latest initscripts are to blame. Drop back down a version and all returns to sanity, well, mostly all...
Agreed. The previous version of initscript was seriously messed up too but this one really takes the cake. How come you did not notice that various services are not starting up before you even see a shell? Quite likely easy to work around.
To add an icing to the cake it look that the current version of xorg-x11, i.e. 6.8.1-19, is seriously broken as well. At least on my Radeon it simply locks up everything with no trace of a picture. I will have look at logs once I will get to them. :-) The previous xorg-x11 was running just fine.
Michal
I wrote before that also xorg-x11 seems to be goner after recent updates. A closer examination reveals that this is not precisely true. A check in logs shows that because of a confusion of mounts created by initscripts X cannot find any fonts so it does not start. The only problem is that instead of returning with a failure X holds to a keyboard and sits there doing nothing. Courtesy if initscripts sshd also did not start, even if had a network up, so I could not reach buggered out machine over a network as well.
Booting to level 1, cleaning up all that mess done by a startup, and then doing 'telinit 5' does work. Until you shovel out that junk things like 'mount' or 'df' have no chance to work correctly.
I guess that I will submit a bugzilla report tommorow unless somebody will be faster (which I do not mind at all :-).
Michal
On Sat, Nov 27, 2004 at 12:49:34AM +0000, Paul wrote:
Latest initscripts are to blame. Drop back down a version and all returns to sanity, well, mostly all...
Previous version of initscripts is also broken even if consequences are not so dire.
I did some investigations. In /etc/rc.d/rc.sysinit there are lines which say:
# Enter mounted filesystems into /etc/mtab mount -a -f
Put directly before that 'mount -a' (without quotes, of course).
This does not solve all troubles but at least makes a system bootable (even if you will notice such "details" like possibly a missing entry for '/' in /etc/mtab and that also means that '/' will not show up in an output of 'df'; this can be fixed too but I do not see a "one-liner").
Michal
On Friday 26 November 2004 18:34, Brian Millett wrote:
On Friday 26 November 2004 16:46, Paul F. Johnson wrote:
Hi,
On Fri, Nov 26, 2004 at 07:33:34PM +0000, Paul wrote:
My entire /usr directory is completely empty! It was working, I did
a reset and on restart, I started to get all sorts of
errors, the drives didn't mount and on checking, /usr is devoid of
absolutely everything!
Is /usr empty because it didn't mount. Don't panic until you've checked the partition tables and fsck
Everything other than /usr is fine. The only thing which doesn't look right is that in fstab, the line for user is LABEL=/usr, whereas everything else is /var (and so on).
How do I check the partition tables? Remember, I have nothing in /usr
TTFN
Paul
This sounds as if you had it setup with labels, but are not using an initrd now, which is required to run labels as opposed to direct pointers such as /dev/hda7. If you know which partition was the one that has the /usr on it, and you can do that with experimental mounts to /mnt/someplace, (after a mkdir /mnt/someplace) then mount -t ext3 /dev/hda1 /mnt/someplace, then do an ls on it and see if thats /usr, if not, umount it, and try /dev/hda2 etc until you find the /usr partition. When you find it, get rid of that LABEL= crap in your fstab and use that instead. When its fixed, reboot.
No, there seems to be something that happened with yesterdays rawhide update. I just rebooted and got the message that my home dir did not exist. I went to a virtual term and looked to see what ws up. This is what I got:
Filesystem Size Used Avail Use% Mounted on /dev/hda1 97M 18M 75M 20% /boot none 244M 184K 244M 1% /dev/shm /dev/hda5 97M 18M 75M 20% /home
Well that is weird.
This is my fstab: # This file is edited by fstab-sync - see 'man fstab-sync' for details LABEL=/ / ext3 defaults 1 1 LABEL=/boot /boot ext3 defaults 1 2 none /dev/pts devpts gid=5,mode=620 0 0 none /dev/shm tmpfs defaults 0 0 LABEL=/home /home ext3 defaults 1 2 none /proc proc defaults 0 0 none /sys sysfs defaults 0 0 /dev/hda3 swap swap defaults 0 0 /dev/hdc /media/cdrecorder auto pamconsole,fscontext=system_u:object_r:removable_t,ro,exec,noauto,ma naged 0 0 /dev/scd0 /media/cdrecorder1 auto pamconsole,fscontext=system_u:object_r:removable_t,ro,exec,noauto,ma naged 0 0
A ls of /home showed that it was empty. So I mounted it. and it was there.
But I did not see anything in /boot. /boot is hda1, but / is hda2. The size for /boot is correct, but when I went to /boot it was empty. Looking at / showed that it was mounted as /boot. I mounted /boot ok, and the files are there, but the latest df shows:
[bpm]$ df -h Filesystem Size Used Avail Use% Mounted on /dev/hda1 97M 18M 75M 20% /boot none 244M 184K 244M 1% /dev/shm /dev/hda5 26G 6.2G 19G 26% /home /dev/hda1 97M 18M 75M 20% /boot [bpm]$ cd / [bpm]$ df -h . Filesystem Size Used Avail Use% Mounted on
9.7G 5.2G 4.0G 57% /
So I'd say something is rotten.
It would appear so, and I don't have any other ideas.
On Fri, Nov 26, 2004 at 08:03:05PM -0500, Gene Heskett wrote:
On Friday 26 November 2004 18:34, Brian Millett wrote:
So I'd say something is rotten.
It would appear so, and I don't have any other ideas.
Do for the moment 'ln -sf /proc/mounts /etc/mtab'. It is not really correct but should tide you over.
If you have some file systems which appear to be mounted but still have nothing to show, say /usr to be specific, do
umount /usr ; mount -a
without paying attention to error messages. Most likely it will help. Repeat as long as needed.
Michal
On Fri, Nov 26, 2004 at 08:03:05PM -0500, Gene Heskett wrote:
On Friday 26 November 2004 18:34, Brian Millett wrote:
So I'd say something is rotten.
It would appear so, and I don't have any other ideas.
Do for the moment 'ln -sf /proc/mounts /etc/mtab'. It is not really correct but should tide you over.
If you have some file systems which appear to be mounted but still have nothing to show, say /usr to be specific, do
umount /usr ; mount -a
without paying attention to error messages. Most likely it will help. Repeat as long as needed.
Ok, just for laughs, I rebooted into single user and cat /proc/mounts rootfs / rootfs rw 0 0 /proc /proc proc rw,nodiratime 0 0 none /dev tmpfs rw 0 0 /dev/root / ext3 rw 0 0 none /dev tmpfs rw 0 0 none /selinux selinuxfs rw 0 0 /proc /proc proc rw,nodiratime 0 0 /proc/bus/usb /proc/bus/usb usbfs rw 0 0 /sys /sys sysfs rw 0 0 none /dev/pts devpts rw 0 0 none /proc/sys/fs/binfmt_misc binfmt_misc rw 0 0
then I mounted /boot and /home. then
cat /proc/mounts rootfs / rootfs rw 0 0 /proc /proc proc rw,nodiratime 0 0 none /dev tmpfs rw 0 0 /dev/root / ext3 rw 0 0 none /dev tmpfs rw 0 0 none /selinux selinuxfs rw 0 0 /proc /proc proc rw,nodiratime 0 0 /proc/bus/usb /proc/bus/usb usbfs rw 0 0 /sys /sys sysfs rw 0 0 none /dev/pts devpts rw 0 0 none /proc/sys/fs/binfmt_misc binfmt_misc rw 0 0 /dev/hda5 /home ext3 rw 0 0 /dev/hda1 /boot ext3 rw 0 0 sunrpc /var/lib/nfs/rpc_pipefs rpc_pipefs rw 0 0
Then I
cat /etc/mtab /dev/hda1 /boot ext3 rw 0 0 none /dev/pts devpts rw,gid=5,mode=620 0 0 none /dev/shm tmpfs rw 0 0 /dev/hda5 /home ext3 rw 0 0 none /proc proc rw 0 0 none /sys sysfs rw 0 0 none /proc/sys/fs/binfmt_misc binfmt_misc rw 0 0 /dev/hda5 /home ext3 rw 0 0 /dev/hda1 /boot ext3 rw 0 0 sunrpc /var/lib/nfs/rpc_pipefs rpc_pipefs rw 0 0
Ok, so it is wrong. Am going to try the ln -sf /proc/mounts /etc/mtab and see what happens.
Then I'll back grade initscripts.
On Fri, 2004-11-26 at 18:10 -0700, Michal Jaegermann wrote:
On Fri, Nov 26, 2004 at 08:03:05PM -0500, Gene Heskett wrote:
On Friday 26 November 2004 18:34, Brian Millett wrote:
So I'd say something is rotten.
It would appear so, and I don't have any other ideas.
Do for the moment 'ln -sf /proc/mounts /etc/mtab'. It is not really correct but should tide you over.
And won't work if you have things on different filesystems. initscripts is broken, downgrade to 7.98-1 and things work again.
TTFN
Paul