Large file corruption on XFS

Bug #190839 reported by gsyoungblood
6
Affects Status Importance Assigned to Milestone
linux (Ubuntu)
Fix Released
Undecided
Unassigned
linux-source-2.6.22 (Ubuntu)
Won't Fix
Undecided
Unassigned

Bug Description

Summary:

Large files seem to have problems on XFS partition with Ubuntu 7.10 (Gutsy).

How to Repeat:
1. Home directory mounted on XFS partition.
2. Download kubuntu 7.10 DVD and save to ~/Desktop/.
3. Copy file to new name (i.e cp kubuntu-7.10-dvd-i386.iso kubuntu-7.10-dvd-i386-test.iso).
4. Run md5sum on the files 2 or 3 times (i.e. md5sum kubuntu*.iso)

Description:

I downloaded the Kubuntu DVD to try it out on another machine. After the download completed, I burned the image to a blank DVD. During the install, an error was reporting during a copy operation, noting that the problem is usually due to a bad media/disc. In the process of burning another DVD, I happened to do an md5 sum on the original file a second time and noticed the values were different. I did a couple of additional tests and confirmed md5sum was reporting different values each time.

Hardware:

Compaq sr1520nx (AMD based system w/Sempron 3100+)
1 gb RAM
ST3160812A 160gb hard drive (UDMA/100 PATA)
TSSTcorpCD/DVDW CD/DVD RW

Kernel: 2.6.22-14-generic
Modules: libata: ata_generic,pata_sis
Distro: Ubuntu 7.10 (Gutsy), updated to current patches (no outstanding patches)

Hardware is working fine. No other hardware issues. Was running Solaris 10 x86 a couple of weeks ago without any problems (including with large files like dvd images).

Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

Hi gsyoungblood,

The Hardy Heron Alpha series is currently under development and contains an updated version of the kernel. It would be helpful if you could test the latest Hardy Alpha release: http://www.ubuntu.com/testing . You should be able to then test the new kernel via the LiveCD. If you can, please verify if this bug still exists or not and report back your results. We'll keep this open against the actively developed kernel but against 2.6.22 this will be closed. Thanks.

Changed in linux:
status: New → Incomplete
Changed in linux-source-2.6.22:
status: New → Won't Fix
Revision history for this message
gsyoungblood (greg-tcscs) wrote :

Downloading beta hardy desktop - will test this weekend.

Revision history for this message
Rob van der Linde (robvdl) wrote :

I have been affected by this bug as well, and seeing this bug has had no closure yet, I felt it was my duty to run some tests.

First of all, let me explain my experiences with XFS corruption. I have had 3 completely different machines, all ran Gutsy, some 64 and some 32 bit. Over time, these machines started experiencing numerous problems with file corruption and it wasn't because of bad shutdowns. First it was minor things such as my firefox bookmarks.html got corrupt for no reason, then things started getting worse like another machines apt status file got corrupt, so I could no longer install packages or updates, it rendered the machine useless and I had to reinstall. Another friends machine also started experiencing numerous signs of corruption and I had to reinstall, I had then switched to JFS instead.

Since Hardy has been released and it appears some work has been done on XFS (I found a page on the net that mentioned the bug had been looked at and it was marked with a green line (indicating it's fixed?), but can't find the page now), I decided to test it. I have setup an entire server on XFS in software LVM (two identical disks) mode and have tested and retested copying large ISO's over and over again, even copying two in parallel. I must have at least tried it 10 or more times. Every time the MD5 sums came up perfect.

I am pretty confident this bug is resolved in Hardy and only exists in Gutsy. I will continue running this system and see how it goes. I will report again in the future if the machine has been going fine, but I am pretty confident it is fixed as I ran the tests heaps of times.

Revision history for this message
Leann Ogasawara (leannogasawara) wrote :

Hi Rob,

Thanks so much for the testing and the feedback. From your comments it sounds promising that this issue is resolved with Hardy. I'm going to go ahead and mark this "Fix Released" against Hardy for now. Obviously if this is not the case feel free to reopen this report by changing the status back to "New". Thanks again.

Changed in linux:
status: Incomplete → Fix Released
To post a comment you must log in.
This report contains Public information  
Everyone can see this information.

Other bug subscribers

Remote bug watches

Bug watches keep track of this bug in other bug trackers.