reiserfs partition corrupted

hi guys, happy new year, it's been a while...

i have a root reiserfs partition abt 13 GB, that seem to have been totally corrupted (unable to boot it anymore).

it began with a mysterious disk-space issue, i had about 500 MB free,
using df -h, I noticed that it would rapidly decrease from abt 500 to a few KBs, without apparent reason.

now, i tried to fix it using reiserfsck, then reiserfsck --rebuild-tree, where fixed some things, but it stopped past PASS 1 b/c insufficient diskspace, at that point i was unable to mount/boot it anymore.

does anyone know of a possible way to still rescue it? any program/procedure ?

any help is appreciated.

OS:
slackware 11

Did you have / and /swap as the only partitions? If /var is not on it's own partition, it CAN take up and fill the root drive pretty quick, but it shouldn't do 300 MBs like that unless you were running a server.

Have you had a look at /var/log?

I know it's possible to fill quick, I set up Slackware 13.0 on an OLD machine as a secondary FTP server and transferred over 50 GBs to it in an hour.

yeah, /var was in the same partition, and im starting to get afraid that /home was there too :S since i can't locate it in my backups, but it was like in seconds, the whole partition gets full, i mean it got to be a problem with the filesystem.

however, my main problem now is try find a way to repair it other that reiserfsck stuff,

cheers

EDIT:no it's not a server just for personal use. that's weird

Quote:

---

Originally Posted by HackerzMaster

however, my main problem now is try find a way to repair it other that reiserfsck stuff,

cheers

---

Again.... are you running a server or have a Media Center setup? You never told us what you're running. Could be a file permission problem. If you have scheduled recordings setup with the DVR setup, delete all those high-def mpegs. Those could suck the life out your rig.

Try the following:
1. boot with a livecd; login at the console
2. Try to mount the suspected partition; 'fsck' to repair
3. after mount, delete huge files in your /tmp in root and do the same for each account you have setup. Using the filemanager in GUI can quickly find those files by name and date using the search feature.
4. Surgery: In your home directory, view hidden files. Delete everything in .Trash -- guessing your setup.
5. You can regain some file space by deleting .gz or .zip files inside /var/logs

[5] Keep your home directories off the system partition.

Hmmmm,

A Slackware 11 root partition of 13Gb..............that sounds rather excessive to me, what else is in there? I don't know Slackware, but I can run Windows 2000 quite happily on a 2.1Gb HDD?

Have you tried booting with a live CD and see if you could get into the partition? If you can, you might try moving stuff into another partition to create more space. Or just temporarily save it onto a DVD or USB drive.

I guess the first thing I would do in these circumstances is get the HDD manufacturer's diagnostic tool and run that. When strange things like this happen right out of the blue it frequently means that the HDD is on the way out.

Quote:

---

Originally Posted by nihil

I guess the first thing I would do in these circumstances is get the HDD manufacturer's diagnostic tool and run that. When strange things like this happen right out of the blue it frequently means that the HDD is on the way out.

---

Keep it simple. If that HDD is S.M.A.R.T. it can tell right away if it's dieing. For the OP....on the commandline 'ls /usr/bin | grep smart'.

-rwxr-xr-x 1 root root 1230268 Apr 7 2009 gsmartcontrol*
-rwxr-xr-x 1 root root 2871 Apr 7 2009 gsmartcontrol-root*

gsmartcontrol can give you a status on all your active HDD's

hi there,
it's not a file permission problem the partition is messsed up.
used a live cd, and no mounting, and as i said im now only concerned about recovering it.
the thing happened when rebuild-tree froze in the middle, (i was told this is a serious thing)

EDIT: im running now on a windows partition on the same hd, have similar issues, with disk-space.
but what i need now is any advice on recovering the reiserfs partition.

EDIT: no media centre or anything, just regular use, couple of movies that's all
cheers

Quote:

---

Originally Posted by HackerzMaster

what i need now is any advice on recovering the reiserfs partition.

cheers

---

Try this from the liveCD

reiserfstune --no-journal-available <disk>

allows reiserfstune to continue when the current journal's block device is no longer available. This might
happen if a disk goes bad and you remove it (and run fsck).


Hope you get your stuff back

will try this and get back,

Before you pull the life support on that HDD, search for a low-level disk tool that allows you to flash or reset all stored "bad" blocks it think it has. I forgot what you call those programs but I know they exist. You probably need to check the drive company's download section for that drive model.

You might try this:

http://www.roadkil.net/program.php?ProgramID=29

It will recover anything it can find on a drive and attempt a repair.

What version of Windows and how much space?

Have you run a hard drive diagnostic yet?

thanks for your help
here is what i got so far:

reiserfstune --no-journal-available /dev/sda4

output:
reiserfstune: Filesystem looks not cleanly umounted, check the consistency first .

reiserfsck /dev/sda4 gives:

Bad root block 0. (--rebuild-tree did not complete)

Aborted

Well, I personally don't hold out much hope of recovery, perhaps it would be wise to consider what caused the problem?

1. Slackware 11 is a 2006 distro which suggests that your HDD is older than that?

2. It is not a question of if a HDD will fail it is just a question of when. I always check the hardware first in situations like this.............the test only takes minutes compared with hours messing with OSes and Apps, only to find that the HDD is past it and you are going to have to start over anyway ;)

3. You have been having problems with both your Slackware and Windows partitions due to lack of filespace? That tells me you are filling them too full..........I personally try to allow 20&#37; freespace on a drive/partition to give tools some space in which to operate. OK, you don't need that much if it is just a data storage partition or drive, but I still wouldn't like to go under 10%

4. Do you have access to a Windows machine that would accept a slaved drive? I seem to recall that there is a file recovery utility that will run under Windows that will fix reiserfs problems?

5. How much RAM does your machine have?

EDIT:

Quote:

---

Keep it simple. If that HDD is S.M.A.R.T. it can tell right away if it's dieing. For the OP....on the commandline 'ls /usr/bin | grep smart'.

---

Unfortunately SMART isn't................it has a detection rate of less than 60%. Anyway, manufacturers diagnostics check the SMART logs as well. Might as well go the full mile IMO.

Quote:

---

Work at Google on over 100,000 drives has shown little overall predictive value of S.M.A.R.T. status as a whole, but suggests that certain sub-categories of information which some S.M.A.R.T. implementations track do correlate with actual failure rates – specifically, in the 60 days following the first scan error on a drive, the drive is, on average, 39 times more likely to fail than it would have been had no such error occurred.

---

Please remember that SMART is predictive, whereas the manufacturer's diagnostic is analytical.

I think the drive is trying to commit suicide. Don't take it the wrong way, but if TWO OSs, BOTH are having these problems, you've already done one part by testing it with another OS entirely.

I mean, had you booted Windows and had no issue, then OK, maybe there was a partition problem with your "/" file system, however, booting another OS up and seeing a similar issue, there MIGHT be a small chance you can try to recover things, but I would very seriously consider buying a new disk, as this one might be on it's last legs unless for some reason you were holding a Bass Amp near it.

Sound can damage a disk and so can heat, but filling a disk, I don't think I've even heard of this before, so, I'm thinking the disk is going. Have you noticed any strange noises? Like a grinding sound? Lots of disk activity when there shouldn't be?

I've only seen a disk go once in my 10 years where it physically was grinding, and it was a Windows machine of my Mom's, and She had told me there was this weird problem and "It wasn't working right and it sounded funny like a WRRRRRRRRR" lol.

Two days later it wouldn't even boot, and I finally heard the sweet sound of a hardware device begging lol. anyway, long story short, I pulled the drive out, which was a crap 9 GB "Fireball" or whatever the hell they called it, I hadn't even heard of this brand before, and the thing was grinding like a Stick shift with an idiot at the wheel.

So anyway, if you have another disk, I'd say pop it in, and start over, and possibly trying to mount the old disk and recover some of your stuff, but don't assume it will, as a physical error on a drive isn't exactly going to just stop.

I saw someone saying to try low level, and I agree, you may be able to try and use it from that, but the data is most likely gone. Do you have back ups somewhere else? I take back ups pretty seriously ever since I got totally screwed once where I lost all my personal files, and had really no back ups except a few things on floppy disks, and it was hard, I was pissed, and from that day, I started making sure I had at least a ZIP Drive, which at the time worked fine, and I then, lost everything again....

Here is how stupid I was this particular day:

I was planning on doing a tri-boot scenario. I had Windows 98 SE, and Linux, and planned on putting FreeBSD on, and made the partitions, and I don't know why, but I for some reason decided "Well I can erase these old back ups and make new ones... AFTER I install an OS, like FreeBSD, and that should be fine!".....

I felt like a moron, and I was. The drive stopped booting do to HD errors, and I had deleted my back ups, and for the second time, lost my stuff.

Now I have a nice FTP server set up on my old desktop, with an extra disk on there for storage, and, I've started using an old machine (My Mom's, who's disk pissed on itself) and using that for the very much needed data so I can back those up in two machines, and now I have two machines with back ups, and a separate external USB HD, where I have back ups again, + CDs I burned of the really needed stuff. That's like 4 different locations just in case. And on top of that I still have my ZIP Drive, so some of the under 100 MB files are still on there.

This might seem like way too much for anyone, but I have good back ups for the most part, and a lot of them, which makes sure I won't lose all my stuff again from a disk drive either stopping, or from a whoops format / rm -rf /* / del /S /Q *.* or whatever else lol.

Sorry for the length and oddness of my post. It's almost 8 PM right now, and I went to bed Monday morning at 4:30 AM, and got up at 9:30 AM, and then was awake until 10:40 AM, and an hour later I was awake again, and I'm still up... Wee bit tired. Wanted to help though.

Quote:

---

Originally Posted by nihil

Please remember that SMART is predictive, whereas the manufacturer's diagnostic is analytical.

---

The Smart feature has alot of error codes to present to you at anytime. If your supplier starts acting up.....stop them in their tracks and give them a random smart code. They will shut up and summerize an error report. For example......error code 203 means ECC errors. After giving the supplier that code what else can they say to you? The supplier works for you under contract. You don't have time to defrag, do this and do that. Make them work.

They will end up giving you a free tech person to replace all those harddrives or give you priority over everyone else when their supply is low. As a bonus, it will stop the clock until the new unit arrives 2-3 days later. If your customers complain, give them the RMA number and tell them to STFU. **This is a good answer to give an interviewer if you are looking for a tech job and hardware failure questions come up. It separates the men from the little boys.**

The person working at Google didn't tell you the whole story. This is how you get things done and make people jump. That code saved your workload and made you look good to the boss when the customers are complaining to him/her. You right....it is predictive in this case. :)

hey guys,
i tried this diagnostic thing, called pc-doctor that came preinstalled, and strangely enough, it passed all tests (both simple & extended).
i dunno, maybe its some bad sectors, cuz i remember dropping it once (it's a laptop, no biggie just from about a feet or so).

and i think the large size ( 13GB / ) is due to having /home, /var within same partition, which is stupid i know.

im looking through my portable hd backups, hopefully i'll find some old stuff..

Having a bi / partition isn't a bad thing. The only time I really go for making customized partitions is when I'm setting up a box for something specific, which is rare for me.

I set up a FreeBSD machine not to long ago, and everything was set up on it's own slice more or less, which is what FreeBSD uses for that, and I was doing it basically as a test machine, and as a test server, so I made very small root and so on partitions, and then the rest was for /usr.

I'm setting up a machine right now, with Slackware, where I'm going to be using it for multiple things.

I'm setting up a swap space of one gig, and then, the rest is going as one big / partition. No separate /home, no separate /var, nothing. I have a 120 GB HD in that machine, and so 1 gig for swap and the rest as a / partition, will do fine, since I'm going to be using it mostly for testing out as a new FTP server.

I'm running low on disk space on my main Server which has served me so well, but now I have a lot of stuff I need to have backed up, and some isn't going to fit even though that machine, which came with a 43 GB HD (Yes, 43) which I've dedicated to /storage and allowed all users to read and write to it, and then, I installed a 160 GB HD in it, switched the plugs around, and made it the master drive with swap and one big huge / partition, and i's getting full from me backing up my CD collection on it, as MP3s, and also from me storing all my stuff on it. I have like 10 user accounts on there so that people who come over can use it too, but they don't store as much as I do.

So I had to figure out where to store everything else, and decided on trying out a new version. The Slackware I have installed on the server is 12.0 and I'm using 13.0 on the old machine I talked about before which only has an 80 GB HD in it, so I'm like OK, well, I need a machine that can handle a bit more and has a better processor and more RAM.

The machine I had set up, with the 80 GB HD, has a crap Celeron Processor that's 433 MHz, and it only has 192 MBs of RAM. So, this machine, with a 120 GB HD, and 512 MBs RAM, has an AMD Athlon XP 2600+, so I'm using that to test how Slackware 13.0 works as a Server.

Anyway, point is, the / partition you made is fine. I don't think there is any problem at all. When I make a /home partition, I generally make that the biggest, but I don't like small root partitions either. So I decided whenever I made them, to make sure they had enough room for things like patches.

I generally store patches on the /root of the machine so I can keep track of them, which means I use a little extra space usually.

Quote:

---

Originally Posted by HackerzMaster

same partition, which is stupid i know.

im looking through my portable hd backups, hopefully i'll find some old stuff..

---

I'm not really sure but it sounds like you are having a good day. It's good that you are keeping backups.

Quote:

---

Originally Posted by gore

Anyway, point is, the / partition you made is fine. I don't think there is any problem at all. When I make a /home partition, I generally make that the biggest, but I don't like small root partitions either. So I decided whenever I made them, to make sure they had enough room for things like patches.

---

This forum is for people just like you. Machines break. Look up this post when your "master" HDD resets during a write session. Look up this post when you experienced a real "brown-out" power outage. Look up this post when you can mount a drive but all your directories turn into meaningless numbers.

Why do you think dos\Ntfs file systems suck and real journaling systems like JFS rule? Why do you think RAID's exist? Why do you think they ever came up with the idea of partitions in the first place?

I'm sure Antionline does not keep everything about it on 1 big harddrive. Good luck with your toys dude... it's obvious you never lost anything important. I can still paper chase money from emails from '2001.

You just told the world you are not a real admin but a microcrap testbook memorizer. I stole many jobs from them.

Quote:

---

Originally Posted by Linen0ise

This forum is for people just like you. Machines break. Look up this post when your "master" HDD resets during a write session. Look up this post when you experienced a real "brown-out" power outage. Look up this post when you can mount a drive but all your directories turn into meaningless numbers.

Why do you think dos\Ntfs file systems suck and real journaling systems like JFS rule? Why do you think RAID's exist? Why do you think they ever came up with the idea of partitions in the first place?

I'm sure Antionline does not keep everything about it on 1 big harddrive. Good luck with your toys dude... it's obvious you never lost anything important. I can still paper chase money from emails from '2001.

You just told the world you are not a real admin but a microcrap testbook memorizer. I stole many jobs from them.

---

Wtf are you talking about? I use ReiserFS and Ext4... And how is me saying I make my root file system bigger got anything to do with Windows?

Your posts are usually pretty hard to follow but how the hell you managed to call me a Windows test book person and not an admin, which, I've never done as a job anyway, from me saying that I have 4 back ups of everything, is beyond me.

How the hell did you come to a conclusion that I've never lost anything, when one of my posts pointed out not only how I lost everything twice, but that it was the main reason I set up a server to make back ups easier, and then started talking about my back up policy? Seriously I've read your post twice just to try and figure out if it was ME reading it wrong, but I'm coming up blank because you said that I'm a Microsoft weenie, because I said how I generally just use one root partition? How in the hell does this make any sense at all?

I'm not a sys admin, I'm a Bastard, the title kind of pointed it out.
I slept through my MCSE class and still passed. A also passed my security and Linux certs.

EDIT:

I finally found something in your post I missed!

Where do you see on this thread that his root file system, which he was trying to fix, was on a server? I saw him saying he had Slackware and Windows on that machine. NO ONE would set up a server with Windows and Slackware on it, unless it was a toy project. And, where exactly did I say my server only had one big partition? I didn't, I said it had multiples and that on TEST machines, I made one big one. I was trying to make him feel better since he seemed to think a root partition should be small. FreeBSD is thinking about changing their default sizes for that file system. Are the people who make FreeBSD Windows admins too?

Anyway, maybe you missed the part where he said it wasn't a server at all? So why are you talking about Servers? I brought them up to say I set up a PC Server, as an FTP service so I could make back ups from every machine. How would you bring up the AO Servers, when someone is talking about a small home based one purely for back ups? It's not even close! Learn to read next time. The AO Servers, and my little Pentium 3, are not even in the same area code in the server world.

Quote:

---

The Smart feature has alot of error codes to present to you at anytime. If your supplier starts acting up.....stop them in their tracks and give them a random smart code. They will shut up and summerize an error report. For example......error code 203 means ECC errors. After giving the supplier that code what else can they say to you? The supplier works for you under contract. You don't have time to defrag, do this and do that. Make them work.

---

You obviously live on a different planet? In Europe you run the manufacturer's diagnostics first. This will read any SMART logs and will conduct other tests that SMART doesn't get near. It also checks to see if the item is still under warranty.

It will then produce an error code that the manufacturer knows, understands and trusts............ and automagically provide an RTB document that you just send with the defective drive.

Quote:

---

After giving the supplier that code what else can they say to you?

---

They will tell you to run their diagnostics :D

This is particularly true of small/medium businesses, SOHO and private domestic users who have no special relationship with their suppliers and whose only contractual rights are those granted by the manufacturer's warranty and local commercial law.

Please remember that Europe is made up of a number of countries who all speak different languages, and that long distance telephone calls are obscenely expensive by US standards. The manufacturers diagnostics are the way to go to get a speedy resolution, or any resolution at all.

Another thing, is that here OEMs typically supply a system with a 1 year warranty. HDDs frequently have a 3 year warranty. Strictly speaking, your legal rights are 1 year for ALL components, but the HDD manufacturer doesn't check that, if they get their own RTB document..........they just send you a replacement.............I have pulled that stunt more than once ;)
You certainly don't want to make contact with them other than through their RTB system because you will be asked what machine it was running in and that would be game over.

Legally your contract is with the OEM and their contract is with the HDD supplier.

Quote:

---

This is how you get things done and make people jump. That code saved your workload and made you look good to the boss when the customers are complaining to him/her. You right....it is predictive in this case.

---

SMART is predictive and premptive. You are seriously suggesting you would tell your boss that a drive you were responsible for had been giving SMART errors and you had done nothing about it? That sounds distinctly career limiting to me :eek:

Because the manufacturer's software is analytical and diagnostic there is no suggestion that you might have been asleep on your watch :cool:

Anyways, how many end users have the faintest idea what SMART is? They bring the drive to you when it is already playing up, and the SMART data are an irrelevancy in the past. Show them the manufacturer's RTB voucher and they are happy because they are going to get a replacement, or the diagnostics log and they know that they are going to have to buy one because it is out of warranty.

Quote:

---

Originally Posted by nihil

SMART is predictive and premptive. You are seriously suggesting you would tell your boss that a drive you were responsible for had been giving SMART errors and you had done nothing about it? That sounds distinctly career limiting to me :eek:

---

I was trained and told by the BOSS himself to go this route. That site was a global Headquarters for computer repairs and returns. You have limited number of qualified techs but 4-days worth of computer problems. A tech was made sure not to spend too much time on 1 stupid computer. The name of the game was to stop the clock long enough to fix the simple things.

BTW, why is it that I post solutions and it's always the same TROLLS thinking they attacking me. It is funny.

Quote:

---

I was trained and told by the BOSS himself to go this route. That site was a global Headquarters for computer repairs and returns.

---

Yeah, as I said, it is a question of scale............I am afraid you were coached by someone who technically knew jack sh1t.

I am just firing up another machine to see exactly how long it takes to run manufacturer diagnostics...............you don't have to watch it..........you can go to the PBX for a beer :D it will work whilst you are away.........trust me ;)

And real life isn't the bloody USN...............

Try having to charge a real customer for your work.......that is one hell of a lot different.......believe me!

The OP is a private user with a single machine...........or don't you bother to read the requirements?

EDIT:

<60 seconds to launch,< 10 minutes to run the full test suite