I tried this on the x86_64 list, but no one was able to help me and I'm not having any luck with google - I suspect it's a terminology problem. I had a bad hard drive crash last weekend. KDE froze to the point that I needed to do a hard reboot (the problem turned out to be a memory stick that was loose). Tried to repair it using a SuSE boot disk and the repair option under install. This made things worse. 2 partitions - /dev/hda1 is a linux swap and /dev/hda2 is a bootable reiserfs (kicking myself now that I didn't make home a separate partition). My current state is that if I try to boot that drive, I get a Grub error 16. I can't get into Grub to edit the boot setup SuSE recognizes the partitions but fails when I try mounting the partition on another drive, repair using the tools under installation or update. I've used a rescue cd to check the disk with fdisk and gnu-parted - they agree on the partitions but gnu-parted is reporting that I have 2 TB of space used on a 163GB partition. I have no clue what to do next. Should I try resizing the partition and hope it "deletes" the extra used space or is there some other way to force a recheck or reindex of the partition. System Details: X86_64 on a MSI K8T Neo FIS2R motherboard. 1024kb ram now properly installed and being recognized. The drive in question is made by Hitachi but I don't have the details handy. However, I don't think any of this should matter since the problem doesn't seem to be hardware related. Thanks Hilary -- Hilary L. Hertzoff Young Adult Librarian Mamaroneck Public Library hhertzof@gmail.com
Can you run the install CD , go installation, repair installed system, and do a manual repair of the reiserfs? If you have bad HARD sectors, you can try your luck with ddrescue. Adam Hilary Hertzoff wrote:
I tried this on the x86_64 list, but no one was able to help me and I'm not having any luck with google - I suspect it's a terminology problem.
I had a bad hard drive crash last weekend. KDE froze to the point that I needed to do a hard reboot (the problem turned out to be a memory stick that was loose).
Tried to repair it using a SuSE boot disk and the repair option under install. This made things worse.
2 partitions - /dev/hda1 is a linux swap and /dev/hda2 is a bootable reiserfs (kicking myself now that I didn't make home a separate partition).
My current state is that if I try to boot that drive, I get a Grub error 16. I can't get into Grub to edit the boot setup
SuSE recognizes the partitions but fails when I try mounting the partition on another drive, repair using the tools under installation or update. I've used a rescue cd to check the disk with fdisk and gnu-parted - they agree on the partitions but gnu-parted is reporting that I have 2 TB of space used on a 163GB partition.
I have no clue what to do next. Should I try resizing the partition and hope it "deletes" the extra used space or is there some other way to force a recheck or reindex of the partition.
System Details: X86_64 on a MSI K8T Neo FIS2R motherboard. 1024kb ram now properly installed and being recognized. The drive in question is made by Hitachi but I don't have the details handy. However, I don't think any of this should matter since the problem doesn't seem to be hardware related.
Thanks Hilary
Hi Hilary, On Fri, 26 Aug 2005, Hilary Hertzoff wrote:
I tried this on the x86_64 list, but no one was able to help me and I'm not having any luck with google - I suspect it's a terminology problem.
I had a bad hard drive crash last weekend. KDE froze to the point that I needed to do a hard reboot (the problem turned out to be a memory stick that was loose).
So, essentially, the freeze and ensuing reboot was triggered by a memory problem. I wouldn't call this a hard drive crash.
Tried to repair it using a SuSE boot disk and the repair option under install. This made things worse.
Normally, when you switch off a hard drive it will do a disk check (in your case reiserfsck) before it starts up again. This may take some time depending on how large the drive is. Still it should have come up and run fine afterwards. I am wondering what you did in order to "repair" the problem to make it worse in your opinion.
2 partitions - /dev/hda1 is a linux swap and /dev/hda2 is a bootable reiserfs (kicking myself now that I didn't make home a separate partition).
What are the error messages if you try to mount and it does not succeed?
My current state is that if I try to boot that drive, I get a Grub error 16. I can't get into Grub to edit the boot setup
Not sure what grub error 16 really means but perhaps your MBR was trashed somewhere in the process. I would try to rewrite the boot loader information using yast from the rescue install.
SuSE recognizes the partitions but fails when I try mounting the partition on another drive, repair using the tools under installation or update. I've used a rescue cd to check the disk with fdisk and gnu-parted - they agree on the partitions but gnu-parted is reporting that I have 2 TB of space used on a 163GB partition.
Did you try to repair the partition with "reiserfsck --fix-fixable"? That is the gentle way of approaching the problem. If that doesn't help you may want to try to rebuild the superblock or the tree, depending on the error message you get from running reiserfsck with the fix-fixable option.
I have no clue what to do next. Should I try resizing the partition and hope it "deletes" the extra used space or is there some other way to force a recheck or reindex of the partition.
I would not try to do that. In fact depending on how important the data in your /home directory is, I strongly suggest to use a disk recovery tool like dd_rescue to build an image of the partition someplace else. You can then mount that image with the loopback device and try to recover your data. This would be the preferred way of doing things if the disk is really damaged. Incidentally, I am currently doing exactly this on a Hitachi drive that gave up on me a couple of days ago. dd_rescue and the very valuable dd_rhelp script can be downloaded here: http://www.garloff.de/kurt/linux/ddrescue/ . They even supply a SuSE binary rpm. The simplest way of using it is via dd_rhelp which is explained here: http://www.kalysto.org/utilities/dd_rhelp/index.en.html . Hope this helps. Best regards, Alex.
System Details: X86_64 on a MSI K8T Neo FIS2R motherboard. 1024kb ram now properly installed and being recognized. The drive in question is made by Hitachi but I don't have the details handy. However, I don't think any of this should matter since the problem doesn't seem to be hardware related.
Hilary Hertzoff wrote:
SuSE recognizes the partitions but fails when I try mounting the partition on another drive, repair using the tools under installation or update. I've
is your mbr was damaged, take a look at my "Partition Rescue HOWTO" to rebuild a correct partition table. (available anywhere, look at google or on my web site) jdd -- pour m'écrire, aller sur: http://www.dodin.net http://valerie.dodin.net http://arvamip.free.fr
jdd sur free wrote:
Hilary Hertzoff wrote:
SuSE recognizes the partitions but fails when I try mounting the partition on another drive, repair using the tools under installation or update. I've
is your mbr was damaged, take a look at my "Partition Rescue HOWTO" to rebuild a correct partition table.
(available anywhere, look at google or on my web site)
Grub error 16 is a stage-2 error; that is hardly possible if the mbr is damaged. You just quoted where she said that SuSE recognizes the partitions, hardly possible if the partition table is damaged.
Darryl Gregorash wrote:
jdd sur free wrote:
Hilary Hertzoff wrote:
SuSE recognizes the partitions but fails when I try mounting the partition on another drive, repair using the tools under installation or update. I've is your mbr was damaged, take a look at my "Partition Rescue HOWTO" to rebuild a correct partition table.
(available anywhere, look at google or on my web site)
Grub error 16 is a stage-2 error; that is hardly possible if the mbr is damaged. You just quoted where she said that SuSE recognizes the partitions, hardly possible if the partition table is damaged.
safe suse recognize a stupid partitioning theme (terabytes...) so the partition table must be damaged... jdd -- pour m'écrire, aller sur: http://www.dodin.net http://valerie.dodin.net http://arvamip.free.fr
jdd sur free wrote:
Darryl Gregorash wrote:
jdd sur free wrote:
Hilary Hertzoff wrote:
SuSE recognizes the partitions but fails when I try mounting the partition on another drive, repair using the tools under installation or update. I've is your mbr was damaged, take a look at my "Partition Rescue HOWTO" to rebuild a correct partition table.
(available anywhere, look at google or on my web site)
Grub error 16 is a stage-2 error; that is hardly possible if the mbr is damaged. You just quoted where she said that SuSE recognizes the partitions, hardly possible if the partition table is damaged.
safe suse recognize a stupid partitioning theme (terabytes...) so the partition table must be damaged...
jdd
No, she said she was seeing "2 TB of space used on a 163GB partition." That is a file system error.
Darryl Gregorash wrote:
No, she said she was seeing "2 TB of space used on a 163GB partition." That is a file system error.
may be you are right. Should see a fstab report. If the partitioning is normal I can't help :-( jdd -- pour m'écrire, aller sur: http://www.dodin.net http://valerie.dodin.net http://arvamip.free.fr
It sounds like I haven't given enough detail to properly diagnose the problem. First of all, I should have said system crash rather than hard drive crash. Nothing I've seen indicates a hardware problem. What I did to make things worse: The original problem was that it was failing when it tried to load kde so lacking a 9.3 disk (which was at my parents in Ohio) I tried to fix it using a 9.0 disk and the automatic repair via installation. This is what caused the current problems. I should have waited for my dad to send me the correct disk. I've done some experimenting (making notes this time) and these are the results: Using the automatic repair under the SuSE 9.3 installation - it recognizes and mounts the /dev/hda1 swap drive then tries to repair the filesystem on /dev/hda2. It tries twice, claiming it's repaired both times but then cannot mount the filesystem to continue the checking. I then moved on to the manual tools. Using the install new bootloader comes up with a "can't read /etc/fstab" message and fails. Trying the option to find lost partitions comes up with no valid partition table. Using edit partitions comes up with a partition table and a message that the partition table cannot be edited. Attempting to mount via the rescue system comes up with a can't read superblock message. I've got an old drive running in the same computer at the moment for internet access, but I haven't been able to get the printer working - the cups driver keeps dying when I try to set it up and I don't want to spend too much time on fixing this. So I have to keep shutting down and switching the cable when I need to look something up. dd_rescue is not an option at the moment - I don't have another drive that's large enough. I am currently feeling like an idiot because if I had stuck home in a separate partition, I could have installed a new system with no problem. Off to try a few more tests and make more notes Hilary -- Hilary L. Hertzoff Young Adult Librarian Mamaroneck Public Library hhertzof@gmail.com
On August 27, 2005 6:07 pm, Hilary Hertzoff wrote:
It sounds like I haven't given enough detail to properly diagnose the problem.
What I did to make things worse: The original problem was that it was failing when it tried to load kde so lacking a 9.3 disk (which was at my parents in Ohio) I tried to fix it using a 9.0 disk and the automatic repair via installation. This is what caused the current problems. I should have waited for my dad to send me the correct disk.
I did something similar. I tried to apt-get upgrade from 9.0 to 9.3. It would not boot afterwards from the rescue kernels. AT the time my thought was that there were changes in the startup betweeen the new 2.6 kernels and the older 2.4 kernels. I suspect you are facing the same problem in reverse. A repair with the 9.3 disk should fix it. Run of of the rescue system until it arrives.
Using the automatic repair under the SuSE 9.3 installation - it recognizes and mounts the /dev/hda1 swap drive then tries to repair the filesystem on /dev/hda2. It tries twice, claiming it's repaired both times but then cannot mount the filesystem to continue the checking.
I found that boot.scsidev was mounting root before boot.localfs which caused it to fail. I just manually remounted it. localfs is supposed to start scsidev but on the 9.0 system this was not happening.
Using the install new bootloader comes up with a "can't read /etc/fstab" message and fails. if root is not mounted it will
Attempting to mount via the rescue system comes up with a can't read superblock message.
did you try the spares?
I am currently feeling like an idiot because if I had stuck home in a separate partition, I could have installed a new system with no problem.
That was the one thing I did right -- Collector of vintage computers http://www.ncf.ca/~ba600
And the culprit turned out to be the new memory that I had added. Because all of the check tools I'd tried had failed, I though the damage was worse than it was. When attempting to install SuSE 9.3 on a new drive crashed several times I realized that the new memory wasn't working well with the old memory, so I removed it. I was then able to reboot to a rescue system, run reiserfsck to fix the superblock and repair the tree - both of which completed properly this time instead of stalling, and then run the automatic repair and reboot. Problem solved. All my data seems to be present and accounted for. Thanks for all the tips. Hilary -- Hilary L. Hertzoff Young Adult Librarian Mamaroneck Public Library hhertzof@gmail.com
jdd sur free wrote:
Hilary Hertzoff wrote:
SuSE recognizes the partitions but fails when I try mounting the partition on another drive, repair using the tools under installation or update. I've
is your mbr was damaged, take a look at my "Partition Rescue HOWTO" to rebuild a correct partition table.
(available anywhere, look at google or on my web site)
jdd
What version of SuSE did you say you were using? file:///usr/share/doc/howto/en/html/Partition-Rescue/index.html It is even the most recent version, which I could not find on your website ;)
Darryl Gregorash wrote:
What version of SuSE did you say you were using?
fdisk will do with any
file:///usr/share/doc/howto/en/html/Partition-Rescue/index.html
It is even the most recent version, which I could not find on your website ;)
in fact, my site is mostly in french, but google gives 58 200 results for "partition rescue howto" :-) howaver I don't know if it will be of any use on the present case ? jdd
-- pour m'écrire, aller sur: http://www.dodin.net http://valerie.dodin.net http://arvamip.free.fr
Hilary Hertzoff wrote:
I had a bad hard drive crash last weekend. KDE froze to the point that I needed to do a hard reboot (the problem turned out to be a memory stick that was loose).
This is hardly a hard drive crash -- your drive is still working.
Tried to repair it using a SuSE boot disk and the repair option under install. This made things worse.
What exactly did you do, and what were the reported errors when you attempted the repair?
My current state is that if I try to boot that drive, I get a Grub error 16. I can't get into Grub to edit the boot setup
SuSE recognizes the partitions but fails when I try mounting the partition
Then there is nothing wrong with the partition table; grub error 16 is an inconsistent file system structure. Under no circumstances should you attempt to resize a partition with a corrupted file system. At most you should be trying to restore the file system using reiserfsck -- first with no parameters, then again following the instructions that will be returned by reiserfsck. You will be told to run reiserfsck either with the --rebuild-tree, the --fix-fixable, or the --rebuild-db options. As per 'man reiserfsck', normally the last option is only used if mount reports "read_super_block: can't find a reiserfs file system" and you are sure that a Reiserfs file system is there. If nothing above works, try --rebuild-tree again, using the -S option to force reiserfsck to scan the entire partition. Don't try this until you have exhausted all other options; if you haven't got a mountable partition after this, then as far as I know it is lost, though others may know of other things you may try.
participants (6)
-
Adam Vazquez Kb2Jpd -
Alex Angerhofer -
Darryl Gregorash -
Hilary Hertzoff -
jdd sur free -
Mike