DSM-G600, DNS-3xx and NSA-220 Hack Forum

quekky wrote:

perhaps after writing the partition table you may want to verify it is stored as Linux not FAT. Maybe your thumbdrive auto-rename it back to FAT

Code:

#fdisk -l

you can run this on another telnet session while you are formating

Interesting.  The DNS-323 seems to ignore the partition type.  My fdisk partition  was still set to FAT even though it was formatted and mounted as EXT2.  You can update the partition type without reformatting the by using the "t" option of fdisk.  (not that it seems to matter on 1.04 or 1.05)

Cheers!
k

halfsoul, glad to hear you solved your problem--I was gone all weekend, so I didn't see your post until today.  I added this thread to my instructions since it provides yet more symptoms of the weird compatibility issue with the USB port on the DNS-323.

SilentException wrote:

thanks. i did something similar using your first script as a template..

then, with help from uf0un from irc channel i modified fun_plug that mounts the usb partition using UUID instead of fixed device name.

attached is my current fun_plug

One question on the use of UUID. DOes it make it more robust from the point of view of locating the right stick (in case you have many attached)? Also, if I were to use your script, do I need to change the UUID to something else? If so, how do I find out what UUID to use for my stick? TIA

Thanks for the reply. I managed to figure out the blkid command after looking through your script. One more question though... I was under the impression that after ffp is moved to the USB stick, I should then be able to umount HD_a2 once I shutdown all services (except ssh of cuz) right? I can't seem to get to that stage for my box. FYI, I need to unmount it so that I can run e2fsck on it. The cleanboot script doesn't seem to work for me and I was hoping that by moving ffp to the USB stick, I will no longer have problems with the umounting of HD_a2!

SilentException wrote:

try reading here and forward if you'd like to chk your disks:

http://dns323.kood.org/forum/p14563-200 … tml#p14563

i've quickly made a small pack for booting into e2fsck mode. hope it helps!

Yeah, I've been using that for running e2fsck on HD_a2. but booting into an alternative mode is a pain that I'm trying to avoid. (yeah, I'm lazy! )

bspvette86 wrote:

bfg100k wrote:

SilentException wrote:

try reading here and forward if you'd like to chk your disks:

http://dns323.kood.org/forum/p14563-200 … tml#p14563

i've quickly made a small pack for booting into e2fsck mode. hope it helps!

Yeah, I've been using that for running e2fsck on HD_a2. but booting into an alternative mode is a pain that I'm trying to avoid. (yeah, I'm lazy! )

bfg100k,
Are you sure your USB-storage support is starting from someplace like /tmp instead of /mnt/HD_a2?

ie:
cp /mnt/HD_a2/lnx_bin/usb-storage.ko /tmp
insmod /tmp/usb-storage.ko

If it is being loaded from /mnt/HD_a2, it will prevent the FS from being unmounted.  Also, what does a "ps -ef" show?

Cheers!
bspvette

yeah, if you take a look at my fun_plug, it copies the module to ram drive, then loads it

usbmod_loaded=no
#copy usb-storage module
if [ -e $FFP_HD/lib/modules/$(uname -r)/usb-storage.ko ]; then
        echo "Found usb-storage.ko module. Copying..."
        mkdir -p /lib/modules/$(uname -r)/
        cp $FFP_HD/lib/modules/$(uname -r)/usb-storage.ko /lib/modules/$(uname -r)/
        #load usb-storage module
        insmod /lib/modules/$(uname -r)/usb-storage.ko
        echo "insmod usb-storage.ko"
        sleep 2
        usbmod_loaded=yes
fi

i made sure *nothing* runs on the hdd as i'm alergic to disk spin-ups

you should be able to umount hdds. if you want to debug, get a copy of lsof for arm (i could send you one) and do:

#lsof |grep HD_
(this will show you open files on HDDs)

also, before umount try this:
#swapoff /dev/sda1
#swapoff /dev/sdb1

not exactly sure what happened last night but after a reboot today, i'm able to unmount the partition and run e2fsck successfully... new (but expected) problem though... my usb stick is not able to unmount properly on shutdown/reboot (even after applying cleanboot)

bfg100k wrote:

not exactly sure what happened last night but after a reboot today, i'm able to unmount the partition and run e2fsck successfully... new (but expected) problem though... my usb stick is not able to unmount properly on shutdown/reboot (even after applying cleanboot)

Same problem here, cleanboot does not work on my CH3SNAS at all.
My current work around: After reboot I kill the process mt-daapd and then i am able to remount the USB stick as read only. Then I run e2fsck to correct any errors...

If there is a better solution, I would be more happy!

I am trying to understand what happens when i remove the usb stick. I thought that in that case the /ffp on HD_a2 would load. This /ffp is in my case the same as the /ffp on the usb stick because i copied it over.
But the logfile shows that /ffp on a usb stick is found?

**** fun_plug script for DNS-323 (2008-04-13 tp@fonz.de) ****
Thu Jul  3 15:14:06 GMT 2008
Found usb-storage.ko module. Copying...
insmod usb-storage.ko
Waiting for sdc1 (up to 30 seconds) ...
Failed. Did not find sdc1 in /proc/partitions.
* Found FFP on USB device
ln -snf /mnt/HD_a2/Datagb/sdc1/ffp /ffp

Is this normal behavior? I'm using the funplug file provided by Kylek

just had our new USB stick die.. maybe "worn out" ? I wonder if it explains a few NFS streaming errors that we experienced over the weekend ? hmmmmm., Anyways now we get these lovely errors in the dmesg output and the majority of the FS on the USB drive is toast. 

usb-storage: *** thread awakened.
usb-storage: Bad target number (7:0)
usb-storage: scsi cmd done, result=0x40000
usb-storage: *** thread sleeping.
usb-storage: device scan complete
scsi2 (0:0): rejecting I/O to dead device
scsi2 (0:0): rejecting I/O to dead device
scsi2 (0:0): rejecting I/O to dead device
scsi2 (0:0): rejecting I/O to dead device
Buffer I/O error on device sdc1, logical block 0
lost page write due to I/O error on sdc1
EXT2-fs error (device sdc1): ext2_readdir: bad page in #2
scsi2 (0:0): rejecting I/O to dead device
scsi2 (0:0): rejecting I/O to dead device
Buffer I/O error on device sdc1, logical block 0
lost page write due to I/O error on sdc1
EXT2-fs error (device sdc1): ext2_readdir: bad page in #8146
scsi2 (0:0): rejecting I/O to dead device
scsi2 (0:0): rejecting I/O to dead device
Buffer I/O error on device sdc1, logical block 0
lost page write due to I/O error on sdc1
EXT2-fs error (device sdc1): ext2_get_inode: unable to read inode block - inode=2, block=123
scsi2 (0:0): rejecting I/O to dead device
scsi2 (0:0): rejecting I/O to dead device
Buffer I/O error on device sdc1, logical block 0
lost page write due to I/O error on sdc1
EXT2-fs error (device sdc1): ext2_get_inode: unable to read inode block - inode=8145, block=32891

#######################################
#              HD0 awake now !        #
#######################################
scsi2 (0:0): rejecting I/O to dead device
scsi2 (0:0): rejecting I/O to dead device
Buffer I/O error on device sdc1, logical block 0
lost page write due to I/O error on sdc1
EXT2-fs error (device sdc1): ext2_get_inode: unable to read inode block - inode=8146, block=32891
scsi2 (0:0): rejecting I/O to dead device
scsi2 (0:0): rejecting I/O to dead device
scsi2 (0:0): rejecting I/O to dead device

KyleK wrote:

Yes, the script should fall back to a HDD-based fun_plug should it not find the USB stick.
Looking at your log, it seems like you use the same path for the hdd funplug and the USB funplug, which might confuse the script.

Make sure that your mount point for the USB stick is not a directory on one of your disks. I use /mnt/USB for this. (You have to create the directory before something can get mounted to it).

Thanks, that solved it. I made another mount with a different name to the HD_a2 for easier access.

halfsoul wrote:

SilentException wrote:

i made sure *nothing* runs on the hdd as i'm alergic to disk spin-ups

I don't understand something in the script.  After mounting the usb drive and before the symbolic link, why is there a second SUID BUSYBOX script, and why does it seem to call the HD?  Or am I misunderstanding something?

Code:

    # suid busybox on USB
    if [ -x $FFP_USB/bin/busybox ]; then
        chown root.root $FFP_HD/bin/busybox
        chmod 0755 $FFP_HD/bin/busybox
        chmod u+s $FFP_HD/bin/busybox
    fi

thats my bug. it should say $FFP_USB in each of chown and chmod commands.

btw, those commands are there just to make sure busybox has right permissions and that it's suid root. in 99.99% cases they don't even get executed and have nothing to do with disk spinning up or down