After successfully extending a LVM volume like WhatsHisName sugested (see end) but skipping the e2fsck part i had a problem when trying to write something to the volume I always got Read only file system. So i figured that e2fsck war really needed. The first attempt was interrupted by a power failure. Now is been almost 4 days, the e2fsck seems to be running but it doesn't seem to end.

Should I interrupt it?

The command was

USER TTY FROM LOGIN@ IDLE JCPU PCPU WHAT
root tty1 - Wed16 3days 7:30m 7:30m e2fsck -y /dev/mapper/disks-movies


--- Volume group ---
VG Name disks
System ID
Format lvm2
Metadata Areas 4
Metadata Sequence No 12
VG Access read/write
VG Status resizable
MAX LV 0
Cur LV 1
Open LV 1
Max PV 0
Cur PV 4
Act PV 4
VG Size 795.41 GB
PE Size 4.00 MB
Total PE 203626
Alloc PE / Size 203626 / 795.41 GB
Free PE / Size 0 / 0
VG UUID ubjXVi-Q3Wo-rXOI-kh4b-Dl6v-ZQdl-yMS1dP

--- Logical volume ---
LV Name /dev/disks/movies
VG Name disks
LV UUID qVsMoG-Uodk-FkPb-gDyg-1To1-MknY-KMGmIi
LV Write Access read/write
LV Status available
# open 1
LV Size 795.41 GB
Current LE 203626
Segments 4
Allocation inherit
Read ahead sectors 0
Block device 253:2

It's a FC3 on athlon 2000+
Linux ikaros.esties.duth.gr 2.6.9-1.667 #1 Tue Nov 2 14:41:25 EST 2004 i686 athlon i386 GNU/Linux

the VG consist of 2 X 200GB WD 1 X 200GB Matror and 1 X 250GB WD SATA

Thanks




WhatsHisName Wrote

=============
1) After creating the new partition of type 8e (Linux LVM), convert it to an LVM physical volume using the command:

pvcreate /dev/hdb1 # Use the actual partition, not hdb1

The partitioning and creation of the physical volume can be done while running FC.

2) The easiest way to work with the unmounted LVM is to boot into linux rescue mode using the first installation CD (boot: linux rescue), but don’t search for or mount the FC installation. That’s the point: Keep the filesystem(s) unmounted.

From the rescue command prompt, you may need to activate the LVM with the following commands:

lvm vgscan
lvm lvscan
lvm vgchange -a y

Note that when using the lvm commands from FC rescue mode, you need to precede them with “lvm ” for them to work. Now extend the volume group:

lvm vgextend VolGroup00 /dev/hdb1 # Use the actual partition designation

Now see how may free physical extents (PE) and how much free space are available in the volume group:

lvm vgdisplay VolGroup00 # Use the actual VG name

As an example, assume that the Free PE from vgdisplay was 1280 (40GB using 32MB physical extents). Now add that many new physical extents to the logical volume.

lvm lvextend -l +1280 /dev/VolGroup00/LogVol00 # Use the actual # PE and VG/LV names

where the “-l +#” means you want to increase the size by “#” PE’s. Now enlarge the ext3 filesystem to fill the logical volume:

resize2fs /dev/VolGroup00/LogVol00 # Use the actual VG/LV names

You will get a message telling you to run e2fsck, so do what the message says and then run resize2fs...

That’s it! You’re done!

Four days???!!! Damn, you're a patient sort, aren't you?! I would have killed it days ago.

How are you mounting this volume? Let us see your /etc/fstab or the command line you are running.

# This file is edited by fstab-sync - see 'man fstab-sync' for details
/dev/VolGroup00/LogVol00 / ext3 defaults 1 1
LABEL=/boot /boot ext3 defaults 1 2
none /dev/pts devpts gid=5,mode=620 0 0
none /dev/shm tmpfs defaults 0 0
none /proc proc defaults 0 0
none /sys sysfs defaults 0 0
/dev/VolGroup00/LogVol01 swap swap defaults 0 0

/dev/mapper/disks-movies /disks ext3 defaults 0 0

4 days now

So, do you have additional disks beyond the four that make up voulme group "disks" that make up another volume group "VolGroup00" where you have your root filesystem and swap? That's what I'm reading in your fstab. Just wondering. You must be pushing terabyte range with your setup.

Your fstab looks fine to me. I was wanting to check that you weren't simply mounting the filesystem readonly by accident.

You said you skipped the e2fsck part after you initially extended your LV. Then were you able to run resize2fs? Did you skip the resize2fs part as well? On my systems, resize2fs will balk and refuse to do anything until AFTER you've run e2fsck when you're extending an LV. Was the "movies" LV mounted when you tried to run e2fsck? When you were running the e2fsck command for those four days, was it outputting anything to the console? Did that look like normal e2fsck output or was it reporting errors?

ok there is a "VolGroup00" where you have your root filesystem and swap ARE but they r ok. I did not mount the file system readonly by accident. when i skipped the e2fsck part after I initially extended my LV I foced the resize2fs (resize2fs -f) and as you see that worked. in df all was normal. the e2fsck IS running still prining out on the console some numbers about inode blocks which defrent every time( It seems to work). of course I first unmounted the /movies before i fsck it.
is there any way to see the progress of it?
how can i redirect the output of tty1 to an ssh terminal pts/0?

Good. But you'd be surprised how often people do this. Just search these forums for posts talking about not being able to write a mounted filesystem.
I was not even aware of the -f option to resize2fs. Not that I think I'd want to use it.
Sorry, I didn't know this. I didn't see any df output posted that would indicate you actually resized you filesystem. The LV, yes (the lvdisplay output shows this), but the filesystem, no.
I don't know what it's doing. I would not expect anything useful at this point.
That's good. But these questions get asked because we can't always tell if the person asking the questions has any experience with Linux or is a new user and doesn't know things that seem obvious to others. So the question gets asked.
No way that I'm aware of. I don't know of anything you could find out now that would say "IProcessing is 87% complete" or anything like that.

ok thnks haertig I'll let run for one more day and then i will interrup it and try something diffrent.

Ok I found it.
the e2fsprog1.35 have a bug. as soon as i upgrated to 1.37 everything worked fine!!!

I hate it when things like this (weird bugs) happen. I just checked ... I'm running 1.37 at work and 1.38 at home. Hopefully I'm immune to the problem (fingers crossed!)

Thanks for following up. I appreciate it when people put a forum post to rest, once they find the solution.