Reiserfs file system problems [solved]

[ianw1974]

I built a system with reiserfs filesystems, as follows:

 

/dev/hda = 19GB partition "root" or "/".

/dev/hda = 1GB partition "swap".

/dev/hdb = 160GB partition "home".

 

No Windows partitions exist whatsoever, is completely Linux. What happens is when I shut my system down, it appears to be normal, the drives are powered down and then the system turns itself off.

 

However, when I turn it back on again, it seems to think it had a dirty shutdown, and is keeps checking the disks and wanting to replay logfiles etc. I'm not prompted for anything however. But it doesn't list any transaction logs to replay, always says 0 transaction logs to replay.

 

I need to find out how I can get my system shutdown cleanly or at least get rid of these errors so that it doesn't keep attempting to playback a transaction log that doesn't exist. Has anyone any ideas?

[theYinYeti]

Isn't that the correct behaviour (0 transactions to replay is a good point)? What exactly seems wrong?

 

Yves.

[ianw1974]

Well, when it's going through the boot process, it's attempting to replay them. I'm at work now, but I'll post the error message a bit later. I just need to obtain the log files. I'm assuming it'll be in boot.log or dmesg depending on where in the boot process.

 

My other systems don't do the same, which is why I thought it was odd. I'll capture the text, so you can see what it's actually doing.

[scarecrow]

Isn't that the correct behaviour (0 transactions to replay is a good point)? What exactly seems wrong?

<{POST_SNAPBACK}>

Yes, this is the normal ReiserFS 3.X behaviour. Have in mind that Reiser does not keep "lost+found" records- it's a self-healing FS. Compared to ext3, you will have more usable HD space right out of the box due to that FS peculiarity. Since you (ianw) are rather new to Reiser, it's rather natural to wonder, but it seems nothing is wrong with your system.

[ianw1974]

OK, I'm not used to seeing it appear between all the green "OK" as each service is loading.

 

I'll post what it says a bit later when I get home, because it seems to show a lot more information than one of my other installations, which was why I think something wasn't quite right.

[ianw1974]

OK, I've checked the system, and I brought part of the log file to read on my machine, but I cannot read it (Windows at work you see).

 

I formatted the disk using:

 

fdformat /dev/fdd0

 

and copied the file to the disk, and could read it no problems on my Linux system, but cannot in Windows. I guess there must be something else I need to do.

 

Anyway, what happens is it does the usual message for /dev/hda which I expect. And for /dev/hdb it says Replaying Log File, but then says no logfile to replay? And it keeps doing this no matter how many times I reboot, and I just want to reset this so that it doesn't keep attempting to do it.

 

Any ideas?

[ianw1974]

Here's the log file from the machine in question relating to a normal reiserfs entry, and that of the problem I have with /dev/hdb:

 

Nov 14 17:43:16 ianlinux fsck: Filesystem seems mounted read-only. Skipping journal replay.
Nov 14 17:43:16 ianlinux fsck: Checking internal tree..
Nov 14 17:43:17 ianlinux fsck: finished
Nov 14 17:43:17 ianlinux fsck: Reiserfs super block in block 16 on 0x301 of format 3.6 with standard journal
Nov 14 17:43:17 ianlinux fsck: Blocks (total/free): 4624704/4068261 by 4096 bytes
Nov 14 17:43:17 ianlinux fsck: Filesystem is clean

 

and the problematic one:

 

Nov 14 17:43:46 ianlinux fsck: Replaying journal..
Nov 14 17:43:48 ianlinux fsck: Reiserfs journal '/dev/hdb1' in blocks [18..8211]: 0 transactions replayed
Nov 14 17:43:57 ianlinux fsck: Checking internal tree..
Nov 14 17:43:57 ianlinux fsck: finished
Nov 14 17:43:57 ianlinux fsck: Reiserfs super block in block 16 on 0x341 of format 3.6 with standard journal
Nov 14 17:43:57 ianlinux fsck: Blocks (total/free): 39072080/37362273 by 4096 bytes
Nov 14 17:43:57 ianlinux fsck: Filesystem is clean

 

It attempts to reply every single time I restart my system, and I believe it shouldn't be doing this. Has anyone experienced, and how can I clear it from attempting to replay a journal?

[Steve Scrimpshire]

What do your respective /etc/fstab lines look like?

[ianw1974]

This is how my /etc/fstab lines look:

 

/dev/hda1 / reiserfs notail 1 1
/dev/hdb1 /home reiserfs notail 1 2

 

now I thought that since they were different, that it should actually read:

 

/dev/hdb1 /home reiserfs notail 1 1

 

so I changed it to this, but when I rebooted, it still attempted to replay a log file that doesn't exist as per my previous post with the info from the log file.

[ianw1974]

Well, I've decided to give up on reiserfs. First, this error kept occurring, and sometimes I had to choose yes to playback the logfile that didn't exist.

 

Just done a clean installation of Mandriva 2006 this morning, with xfs as the file system, and everything seems to be working perfectly fine. Or at least so far :P

 

Many thanks for your help though, just a shame I couldn't get it sorted to use reiserfs.