N5200PRO read only after rebuild.

[Kinesia]

Hi!

I had a disk fail on my 5200pro. I replaced it and rebuilt the array.
This worked fine and I didn't see any errors.
It now says the disks are all healthy and everything is fine.

There's just one problem... I can no longer write to any of my shares.
I can view them all fine and the data is all there but I can't edit anything.

I've tried going into the ACL screen for a folder and changing the ACLs but it times out when I apply the change (well, about an hour later).

I don't know if the ACLs are the real problem though.

Is anything like this known?


Thanks,
Kinesia.

[peterfu]

Usually the RAID switches to read only in a degraded state, but should switch back to normal operation after rebuild.
On which FW are You?
br
Peter

[Kinesia]

I can't check at the moment.

I thought of something else to try, a disk check, and that's come up with some errors and is taking a very long time to finish.
I hadn't thought of doing it before because it says the array was healthy

[Kinesia]

I'm on 2.0.15, I've been trying to upgrade to the latest (not beta) but it gives

System firmware upgrade fail
Check Sum : e8734a50dfb59f324b29e5c9f4635796


After the disk check finished and fixed a few things it let me create some directories, so it's not acting _completely_ read only now, but trying to copying anything onto it fails. That makes sense because it's almost full...
But it won't let me delete anything or move anything off it.

How do I stop it being full if it needs space to delete or move things? It's all very odd. I don't know what to try next.

[philhu]

Raid units do NOT go into readonly when they are degraded.

That would just about make them useless for a production shop.

[deluxx]

Did you reboot the NAS? I just recreated our raid and when I tried to go to the other links on the admin page, it told me that I didn't have a raid built. So I did a soft reboot and everything seem to be fine now.

[Kinesia]

I'm having to reboot it about once a day, I completely lose both Samba and Web access to it until I restart it using the power switch.
Everything reads fine but it's still read only, even after both a rebuild and a disk check.
The ACLs show all my users and groups in the far left column. I move them to write access but it looks like the action never completes, that webpage stays in "working" mode constantly.

[Kinesia]

The ACLs are definitely bad on some directories and it won't let me change them.

I tried installing the SSHD server so I could login and do it that way but it does not update the module list.
I tried deleting some other modules and they won't be removed from the list either.

It looks like it won't let me update almost anything at all, it's very frustrating. Does anyone have any ideas on how to fix it?

[peterfu]

Seems something is serious damaged on Your box, so I would backup all data as first step, then remove the raid from the box, do a factory reset via webinterface or front panel, make the basic configurations, update FW to the newest one and rebuild everything from scratch.
If the FW update does not work, the try to upgrade the FW with the USB method - You find the needed steps here: http://www.futterknecht.at/index.php...-usefull-stuff (check the factory_test.sh section) It#s only available now for FW 2.00.15, but it will in any way replace then the installed FW.

If this is also not working then I'm afraid Your flash is broken and has to be replaced by Your dealer or Yourself (an image of the 2.00.15 FW for a flash is also available at the above link)

br
Peter

[Kinesia]

I dont have the space to back up the data unfortunately, this is it.

It accepted the beta flash, no idea why it wouldn't take the stable one.

I fixed the permissions using instructions at http://dkprojects.wordpress.com/2008...cus-n4100-nas/
Looking at them they'd definitely gotten corrupted somehow.

Some subdirectories still exhibit problems but I haven't had time to look at them individually yet.

I dearly wish I could've bought another box and copied everything off and start again! Oh for money!

[Kinesia]

I have now worked out (from using sshd and sysuser) that my problem is that the file system is literally being mounted as read-only.
I don't know why or how to fix it. Normally I'd think it was damaged and use fsck but I've already done that and now it's clean...

[peterb]

As a tempoary measure you could manually remount the array as rw

mount -o rw -a should do it, although if you want to be certain, have a look at /etc/fstab to see what ythe system is mounting and where, and applie the mount command accordingly.

I should add that I am not a Thecus user - these are just generic *nix commands.

Once you have got the file system mounted RW you can then do some tidying up and back up etc before you get down to the longer term fix.

It is worth checking the fstab file anyway.

One other thing is that *nix systems tend to mount the FS in read only mode as part of the boot up, and then remount them in rw, so a look at the boot log file might give you some clues.

[Kinesia]

It looks like the raid system isn't in the fstab itself.

/dev/hda5 /app ext2 defaults 0 0
/dev/ram0 / ext2 defaults 0 0
proc /proc proc defaults 0 0
none /dev/pts devpts gid=5,mode=620 0 0


There's a raid0.conf.new file that describes some of the rest but doesn't give the mount point
raiddev /dev/md0
raid-level 0
nr-raid-disks 3
chunk-size 64
persistent-superblock 1
device /dev/sda2
raid-disk 0
device /dev/sdb2
raid-disk 1
device /dev/sdc2
raid-disk 2


It looks like other people have had the problem in the past, but no solution:
http://www.thecusforum.eu/showthread.php?p=7456

(That it works on first boot sometimes is probably why I was confused to begin with. Sometimes it'd play up and others it seemed ok.)

[Kinesia]

Thecus are refusing to give me support because I have SSHD installed, you all should be aware of that.
Even though it uses their mod framework they consider it a "Crack".

So I'm royally stuck.

[Kinesia]

My fstab and mtab are:

root@127.0.0.1:/# more /etc/fstab
/dev/hda5 /app ext2 defaults 0 0
/dev/ram0 / ext2 defaults 0 0
proc /proc proc defaults 0 0
none /dev/pts devpts gid=5,mode=620 0 0


root@127.0.0.1:/# more /etc/mtab
rootfs / rootfs rw 0 0
/dev/root / ext2 rw 0 0
/dev/hda5 /app ext2 rw,noatime 0 0
tmpfs /tmpfs tmpfs rw 0 0
/dev/hda6 /logfs ext2 rw 0 0
proc /proc proc rw,nodiratime 0 0
sysfs /sys sysfs rw 0 0
none /dev/pts devpts rw 0 0
/dev/cloop0 /lib/modules ext2 ro 0 0
/dev/cloop1 /opt ext2 ro 0 0
/dev/cloop2 /img/www ext2 ro 0 0
/dev/cloop3 /img/bin ext2 ro 0 0
/dev/cloop4 /lib/security ext2 ro 0 0
/dev/cloop5 /usr/share/zoneinfo ext2 ro 0 0
sunrpc /tmpfs/var/lib/nfs/rpc_pipefs rpc_pipefs rw 0 0
nfsd /proc/fs/nfsd nfsd rw 0 0
/dev/vg0/lv0 /raid0/data ext3 ro,noatime 0 0
/dev/vg0/syslv /raid0/sys ext3 rw 0 0

[Kinesia]

mount -o remount rw -a /raid0/data
seems to work in the very short term but as soon as you copy over a few k of data it goes ro again and there are no logs on the system to say what's going on.
It's so frustrating.