repeated ntfs errors ?

Started by Metgod, November 08, 2006, 01:46:45 AM

Previous topic - Next topic
Ok... I know next to nothing about NTFS. Hell, I know next to nothing about pretty much every file system. Anyway, the other day I noticed on one of my win machines that the boot was delayed (I was in the other room). So I checked the logs (event viewer, windows 2k sp 4) and see this:

Event Type:   Error
Event Source:   Ntfs
Event Category:   Disk
Event ID:   55
Date:      11/5/2006
Time:      8:23:24 AM
User:      N/A
Computer:   PSYCHOTIC
Description:
The file system structure on the disk is corrupt and unusable. Please run the chkdsk utility on the volume C:.
Data:
0000: 0c 00 00 00 02 00 4e 00   ......N.
0008: 02 00 00 00 37 00 04 c0   ....7..À
0010: 00 00 00 00 32 00 00 c0   ....2..À
0018: 18 00 00 00 00 00 00 00   ........
0020: 00 00 00 00 00 00 00 00   ........

Never mind the computer name, it has nothing to do with me. At least not directly :)

Anyway, it appeared to have fixed the problem. Of course, I did copy files on that drive that were important to me that hadn't been backed up. But today, I noticed another entry (same entry even up to the hexdump, different date).

The problem/concern is that I recently lost a drive to mechanical errors. This doesn't seem like that sort of nature... but it is a bit concerning to me. I have the files safe, but does anyone have an idea what this error on multiple occasions could mean ?

I have some thoughts (like making sure I have all files backed up and deleting all partitions, recreating and formatting the drive again) but not sure how to best approach it or if they're even 100% right...

Any ideas thrown my way could be useful.

Btw, drive is a seagate with capacity of 200g (so more like 187 or so after file system information and such).
"My Terminal is my Soul"

So after I posted this and was getting ready for bed, I had a thought on this.

I think the problem was this:

When my boot drive died on me, Windows had to obviously be reinstalled. Then, because my secondary drive was already partitioned, it seemed to support all data. Of course, with reinstalling windows on the boot drive, the 48 bit lba support was not enabled (forgot about this). According to MS's site, any partition that spans past the limit can get corrupted ....

So the fix is to enable big LBA support.

Curious if anyone else has other thoughts/comments though.
"My Terminal is my Soul"

Our major fileserver here stores about 8TB of data for around 6000 users, so it gets pounded all day.  We've got it running Windows right now, the filesystem being NTFS, of course.  We've run into so many problems, it gets reset once a week and every time it comes back up the drives are corrupted somehow.  I think NTFS might just suck.  It's lightyears past FAT, but it seems to have constant corruption problems.

We're switching over to a SANS soon, and we're going with xfs.
Build a man a fire, and he'll be warm for a day. Set a man on fire, and he'll be warm for the rest of his life.

Hrm.. yeah could be it (NTFS) just sucks. Could also be a lot of stress on the partition ... I use stress loosely, though. How big are the partitions on your disks (saw something about this during my short research of this problem I had) ? Must suck to have ntfs partitions to become corrupt almost every time the server reboots ... not to mention that it is just sad.

Anyway, thanks for the feedback. For now things seem ok.. just hope it stays that way. I get the impression it's not the drive since it's only one partition. Of course, could also be bad sectors, but it seems really unlikely since it's not an old drive (maybe a year old or so from what I remember). That's what I'm telling myself anyway...
"My Terminal is my Soul"

I believe the partitions are the whole disk, 300GB.  It has been my experience that NTFS is both faster and less corruptible with smaller partitions.  I've read that if you're doing alot of writing and reading, 12GB is the max partition size you shoud use under NTFS.  So, on one of the machines I configured for a recording studio, it's got a 250GB hard drive with about 23 partitions and I must say, it gets the job done.  They've never had any problems with it.
Build a man a fire, and he'll be warm for a day. Set a man on fire, and he'll be warm for the rest of his life.

Quite interesting at only 12GB per. Seems to make sense that smaller partitions would work better with a lot of stress on the partition, though.

I have had more trouble today though. At first I thought (or more like hoped) the registry issue was the problem. But today I booted up again and it was a different partition. And I would have been okay with just removing all partitions and formatting them (smaller than before) and then restore data. Doing that anyway but something else is concerning me:

This morning as Windows was loading (just after my partition was corrected by chkdsk), the screen started fading out, as in darkening in colour. It over the next 10 seconds went completely black... and it required a reboot to fix it (monitor cables seemed okay which were the first thing I checked). The thing that concerns me is is my power supply messing things up (this was suggested to me by a long time friend and also I've suspected the PSU was having trouble for a while - maybe I should have just replaced it). Of course, I also recently (say 3 weeks ago) had another problem that I think was power related; it wouldn't boot up (just blank screen like it can sometimes happen when a hdd dies). Thing is, I opened up the case and secured the connection to the drive and then all was ok. Problem also is that the cable seems really short so it seems like it's a bit stressed reaching the HDD from the PSU (perhaps my fault there.. don't know).

Anyway, I'm no hardware guru.. (or more specifically I hate dealing with inside of computers).

Any thoughts on my predicament ? Could use any help/ideas here. Mostly on the fading screen and power supply (or other possibilities of this kind of behaviour with the patitions and the screen fading).

Anyway, thanks again for the feedback.

Hope all is going well over there...
"My Terminal is my Soul"

It doesn't sound like a power supply issue to me.  However, the power supply might be aiding all these faults.  Normally, power supply issues turn up on a cold boot--because to start up, everything needs alot of juice to get going.  A normal hard drive like you're talking about needs about 100-120 watts during boot, then the power needed goes way down.  So, normally it would go on the fritz when it draws all this power.

I have seen three things similar to what you've discribed.  One was because of bad RAM (I have no idea why, RAM errors are so weird), one was the result of a bad hard drive, and one was the result of a nasty virus.  I do have a second suggestion, though.  One of the guys in the department just bought 20 200GB Seagate drives and 8 of them went bad.  Another guy in the department bought 2 and both went bad.  Apparently, the plant that manufactured their model had some issues.  I'm not sure if it's the same one that you have, but maybe you should check into that.
Build a man a fire, and he'll be warm for a day. Set a man on fire, and he'll be warm for the rest of his life.

Actually, you're right. Had the PSU tested and all was ok. What I also discovered is that the data cable on the drive wasn't in (connected) all the way. Furthermore, a fan (a pci card fan) recently died (this I knew but forgot). So I replaced that and did some maintance on the drive (see below). All seems ok so far.

The maintance would be this:

Remove all partitions, recreate and then format then restore data. I think another problem was (besides the fact that the data cable wasn't entirely secure) of course that Windows wasn't configured to support drives > 137GB initially. I imagine that the cable combined with the lack of that registry value, are the problems (or potential problems) here. So hopefully all should be fixed now.

That sucks about all those Seagate drives dying. Hopefully everything was backed up. I've only lost data once for insufficient backups.. it so happened that the one drive that died on me recently was the only one I had that wasn't backed up. Lost a lot of email that was rather important (well nothing urgent but important to me in some ways). Oops...

And yeah memory errors are weird. I'd be interested to know why all the symptoms of bad memory are that way. I only know of random crashes (often sigsegv/11 iirc) and also corrupt data. Those are what I've seen but I know there are more.

Anyway, thanks for the suggestions/feedback in this thread.  I'll post back here if I have any more problems.
"My Terminal is my Soul"

SMF spam blocked by CleanTalk