** Reply to message from Carl Hartung <suselinux@cehartung.com> on Wed, 04 Jan 2006 00:15:07 -0500

Shut down. Boot into rescue mode using the installation DVD/CD. Mount the partition. Make backups, as needed. Unmount the partition. Run reiserfsck to diagnose/repair the problem(s).

That worked. It eventually required reiserfsck --rebuild-tree, which has scared me in the past, but it went well. However, I did the backup with find and cpio. The backup runs so well, I think I will use that partition. Ed Harrison, Registered Linux User #199533 SuSE 10.0, Kernel 2.6.15 PolarBar Mailer 1.26

** Reply to message from Carl Hartung <suselinux@cehartung.com> on Wed, 04 Jan 2006 00:15:07 -0500

Hi Ed,

I'm amazed no one's responded yet! Since it's a ReiserFS error message, I think it's a safe bet you should shut down, back up everything important on that partition and undertake diagnostics/repairs of the filesystem on it.

The basic procedure is:

Shut down. Boot into rescue mode using the installation DVD/CD. Mount the partition. Make backups, as needed. Unmount the partition. Run reiserfsck to diagnose/repair the problem(s).

It must not be a reiserfs problem alone. I booted into hda5 and did a reiserfsck --rebuild-tree on /dev/hdb6 (original partition with messages). When I booted into hda5, with a new reiserfs format, I got similar messages with dmesg. I booted into hdb6 and did a rebuild-tree on hda5. Both are showing no corruptions and no error messages from reiserfsck. The messages now are slightly different, but the interval is still less than 1 minute between entries. ReiserFS: hda5: warning: vs-2100: add_save_link:search_by_key ([-1 298547 0x1 IND]) returned 1 ReiserFS: hda5: warning: vs-2100: add_save_link:search_by_key ([-1 298547 0x1 IND]) returned 1 ReiserFS: hda5: warning: vs-2100: add_save_link:search_by_key ([-1 298547 0x1 IND]) returned 1 ReiserFS: hda5: warning: vs-2100: add_save_link:search_by_key ([-1 298547 0x1 IND]) returned 1 ReiserFS: hda5: warning: vs-2100: add_save_link:search_by_key ([-1 298547 0x1 IND]) returned 1 ReiserFS: hda5: warning: vs-2100: add_save_link:search_by_key ([-1 298547 0x1 IND]) returned 1 Ed Harrison, Registered Linux User #199533 SuSE 10.0, Kernel 2.6.15 PolarBar Mailer 1.26

** Reply to message from Carl Hartung <suselinux@cehartung.com> on Wed, 04 Jan 2006 13:26:34 -0500

You can download (may have on CD already?) diagnostic utilities from your drive manufacturer that can query and test the drive. Of course, the easiest thing to do is replace the entire drive and reinstall... I would. I think that drive is about to *fail*.

I've gotta run now. So, I'll catch up with you on this later today or this evening.

Good luck!

If you noticed, I referenced hda5 and hdb6. The 2 installs are on partitions on separate hard disks, both of which are practically new WD 80Gb disks. I tend to lay this on something software-related, or kernel-related, because they are so new. Until I installed SuSE 10 fresh on hdb6, the only thing in regular use on that drive was a 1.6 Gb swap partition. By the way, I am booted into hda5 at present and the messages to console 10 have returned to 2 every 29 or 30 seconds. Ed Harrison, Registered Linux User #199533 SuSE 10.0, Kernel 2.6.15 PolarBar Mailer 1.26

** Reply to message from Carl Hartung <suselinux@cehartung.com> on Wed, 04 Jan 2006 18:44:20 -0500

I haven't looked over the other responses, yet, but I seem to recall someone else having an issue with 10.0 defaulting to UDMA100 but only having cables capable of supporting UDMA66 or 33. He got better/correct IDE cables and his problems went away.

Both of mine are udma5 with new matching cables. I am fairly certain they are ATA100. Ed Harrison, Registered Linux User #199533 SuSE 10.0, Kernel 2.6.15 PolarBar Mailer 1.26

** Reply to message from "Hylton Conacher (ZR1HPC)" <hylton@global.co.za> on Fri, 06 Jan 2006 15:41:44 +0200

What about investigating SMARTCTL to see if the drives are failing and assuming the drives are S.M.A.R.T. enabled?

HIH

They both report that they are health puppies. The problem was reiserfs. I changed the problem partition to XFS, and no more error messages. Ed Harrison, Registered Linux User #199533 SuSE 10.0, Kernel 2.6.15 PolarBar Mailer 1.26

On Wednesday 04 January 2006 12:07 pm, Ed Harrison wrote:

It must not be a reiserfs problem alone.

I booted into hda5 and did a reiserfsck --rebuild-tree on /dev/hdb6 (original partition with messages). When I booted into hda5, with a new reiserfs format, I got similar messages with dmesg.

I booted into hdb6 and did a rebuild-tree on hda5.

Both are showing no corruptions and no error messages from reiserfsck.

There are those who insist that Reiser is infallible and the problem is in your harddrives. I have had similar problems with my Reiser partitions on different machines doing flakey stuff and requiring the rebuild-tree thing to get right for awhile. My solution to the problem was to no longer use reiser. I've had Zero problems with drives since then. AND, no loss of data or hangs since. Reiserfs may be perfect but I'm perfectly happy not using it anymore. The choice is yours. I'm out of the debate over whether Reiser is perfect or not. Richard

** Reply to message from Stan Glasoe <srglasoe@comcast.net> on Wed, 04 Jan 2006 15:14:36 -0600

My free advice is _if_ you are experiencing lots of Reiser errors and must run reiserfsck --rebuild-tree often, try switching to XFS and see what happens.

Stan

I switched the problem partition to XFS, and no error messages of any kind in 1 hour, 20 minutes. Thanks for the free advice--it will be worth far more than I paid for it :>). Ed Harrison, Registered Linux User #199533 SuSE 10.0, Kernel 2.6.15 PolarBar Mailer 1.26