D-Link Forums
The Graveyard - Products No Longer Supported => D-Link Storage => DNS-343 => Topic started by: dcmwai on March 03, 2010, 03:28:34 AM
-
Dear All,
I've a WD Caviar Green 1.5 TB
Model: WD15EARS
As shown below.
http://www.wdc.com/en/products/products.asp?driveid=772
It is recognized by the NAS but the performance is horrible.
It will intermittently pause when I'm accessing the folder or files.
After some checking I found that it was a Linux issue with the 4K Physical Sector.
Anyone have any idea on how can this be fix?
Thank You
-
http://forum.dsmg600.info/viewtopic.php?id=5468 (http://forum.dsmg600.info/viewtopic.php?id=5468)
for optimal performance, you would need to create the partitions on your own
fdisk or parted
... see below
mkswap /dev/sdd1
mke2fs -j -m 0 -T largefile4 /dev/sdd2
mke2fs -j /dev/sdd4
hd_verify -w
Disk /dev/sdd: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sdd1 64 1060287 530112 82 Linux swap / Solaris
Partition 1 does not end on cylinder boundary.
/dev/sdd2 3164800 3907029167 1951932184 83 Linux
Partition 2 does not end on cylinder boundary.
/dev/sdd4 1060288 3164799 1052256 83 Linux
Partition 4 does not end on cylinder boundary.
Partition table entries are not in disk order
(parted) unit s
(parted) print
Model: WDC WD20EARS-00S8B1 (scsi)
Disk /dev/sdd: 3907029168s
Sector size (logical/physical): 512B/512B
Partition Table: msdos
Number Start End Size Type File system Flags
1 64s 1060287s 1060224s primary linux-swap
4 1060288s 3164799s 2104512s primary ext3
2 3164800s 3907029167s 3903864368s primary ext3
Dear All,
I've a WD Caviar Green 1.5 TB
Model: WD15EARS
As shown below.
http://www.wdc.com/en/products/products.asp?driveid=772
It is recognized by the NAS but the performance is horrible.
It will intermittently pause when I'm accessing the folder or files.
After some checking I found that it was a Linux issue with the 4K Physical Sector.
Anyone have any idea on how can this be fix?
Thank You
-
How can you do that?
I'm even having issue on unmounting the sda4 and sdb4 (raid1)
Please advise.
Thank You
-
are you using fun_plug?
what are the command that you are sending?
How can you do that?
I'm even having issue on unmounting the sda4 and sdb4 (raid1)
Please advise.
Thank You
-
Yes I do have fun_plug 0.5 installed
I try to use fdlink but it will crash with segment fault
So I'm using parted to make the fs change..
But I cannot remove /dev/HD_{a,b,c,d}4
As it say it is still using.
Although it have been umount...
-
yes, fdisk have some problem, try to use parted
http://www.inreto.de/dns323/fun-plug/0.5/packages/parted-1.8.8-1.tgz (http://www.inreto.de/dns323/fun-plug/0.5/packages/parted-1.8.8-1.tgz)
you should stop smb and nfs
you should also turn off the swap on the disk that you want to partition
smb stop
nfs stop
swapoff /dev/sd?1
Yes I do have fun_plug 0.5 installed
I try to use fdlink but it will crash with segment fault
So I'm using parted to make the fs change..
But I cannot remove /dev/HD_{a,b,c,d}4
As it say it is still using.
Although it have been umount...
-
Oh Thank for the help.
Let me try later tonight.
I've a question is that would my telnetd be kill when the ffp have been formated?
And also my driver is a raid 1 array.
Yes, I can copy the HD_{abcd}4 out to a place temporary and later replace it.
But the problem is the raid array how can I remount it and format that?
Thank You
-
i am not sure about the commands for raid setup, i am using "single disk" setup
for "single disk" setup, if you don't touch /dev/sda you should be fine
only 1 of my new disk is having 4k sector and i am using it as back up
Oh Thank for the help.
Let me try later tonight.
I've a question is that would my telnetd be kill when the ffp have been formated?
And also my driver is a raid 1 array.
Yes, I can copy the HD_{abcd}4 out to a place temporary and later replace it.
But the problem is the raid array how can I remount it and format that?
Thank You
-
I'm having the same issue as I bought two of the WD20EARS drives and want to use them in a RAID 1. My linux isn't that strong, so I'm leary of trying to make all these changes. What do you think about setting the jumper on the drives that is recommended for Windows XP? It supposedly offsets the sectors by one so that even though the OS thinks it starts on sector 63 its really starting at 64 and the sector bounaries would be ok?
-
unfortunatlly that is not going to help as the HDD in DNS-343 have 3 partation.
And the pin offset will just safed partation 1 the swap...
So not much helpful...
By the way I think I fail to do that on DSN-343.
the changes must be done on another linux box... too bad :)
-
yes correct, it will add 1 to the sector,
i.e. start from 64 (which is aligned),
it's only useful if you create only ONE partition
unfortunatlly that is not going to help as the HDD in DNS-343 have 3 partation.
And the pin offset will just safed partation 1 the swap...
So not much helpful...
By the way I think I fail to do that on DSN-343.
the changes must be done on another linux box... too bad :)
-
Ok, so assuming I am willing to go through the work to move the sectors to the proper bounderies. Is it possible to do this and end up with a raid 1 configuration?
-
i am using the following to umount ALL partitions and run e2fsck,
i hope if works for you to run fdisk
i believe you best bet is to
- use the GUI to create raid volumes
* setup telnet (follow the link)
* smb stop
* nfs stop
* umount (ALL partitions)
* swapoff (ALL partitions)
- use fdisk to delete and create partition 1(swap) and 2(data), for all your disks
(you can probably leave 4 alone as it's small and contains some admin data)
- please post your success story!
Note: i used the steps marked * before running e2fsck
http://wiki.dns323.info/howto:telnet (http://wiki.dns323.info/howto:telnet)
#
# Minimal fun_plug, only enables telnet
#
# Requires: /mnt/HD_a2/starttelnet.sh
# /mnt/HD_a2/lnx_bin/busybox3
# /mnt/HD_a2/lnx_bin/utelnetd
#
#
# Uncomment next line if you want "proof of concept".
# A filem dmesg.out will be located on \\NAS\HDD_1 or
# \\NAS\Volume_1 depending on firmware version.
#
# dmesg > /mnt/HD_a2/dmesg.out
#
# Start the telnet deamon
#
/mnt/HD_a2/starttelnet.sh
# improved starttelnet.sh, enabling the filesystems to be unmounted and checked, if need be.
# copy the provided components to a directory on the ramdisk
# the ramdisk is regenerated with every boot, so the copy has no lasting effects at all
# simply copy the two files over
cp /mnt/HD_a2/lnx_bin/utelnetd /sbin/utelnetd
cp /mnt/HD_a2/lnx_bin/busybox3 /bin/busybox3
# create the terminal device as usual
/bin/busybox3 mknod /dev/ptyp0 c 2 0
/bin/busybox3 chmod 0666 /dev/ptyp0
/bin/busybox3 mknod /dev/ttyp0 c 3 0
/bin/busybox3 chmod 0666 /dev/ttyp0
# make a shell link on the ramdisk
mkdir /bin/busybox3.dir/
PATH="$PATH:/bin/busybox3.dir"
ln -s /bin/busybox3 /bin/busybox3.dir/sh
# and start the Telnet service from the ramdisk as well
/sbin/utelnetd -l /bin/busybox3.dir/sh -d
Ok, so assuming I am willing to go through the work to move the sectors to the proper bounderies. Is it possible to do this and end up with a raid 1 configuration?
you would need this also, because the original fdisk gives segmentation fault
modify starttelnet.sh to also copy this to /tmp
http://forum.dsmg600.info/viewtopic.php?id=3375 (http://forum.dsmg600.info/viewtopic.php?id=3375)
-
I'm getting closer, but need some more help. I've got funplug on it and added starttelnet.sh to the start directory. I even downloaded the alternative fdisk and copied it to /tmp. Then I unmount everything and try to use the fdisk in /tmp but when say I try to list the existing partitions with 'fdisk -l' I get this error "cannot open /proc/partitions".
Any ideas what I'm doing wrong?
-
i am just guessing, maybe raid is in the way,
have you tried, raidstop?
if i am not wrong you have
-raidstop
-fdisk
-raidstart
-mke2fs
I'm getting closer, but need some more help. I've got funplug on it and added starttelnet.sh to the start directory. I even downloaded the alternative fdisk and copied it to /tmp. Then I unmount everything and try to use the fdisk in /tmp but when say I try to list the existing partitions with 'fdisk -l' I get this error "cannot open /proc/partitions".
Any ideas what I'm doing wrong?
-
Just to update some test result...
Format the RAID in other linux can't seem to be see or display on the system.
Even I've copy over the HD_a4/b4 completely...
Hum... damn hard
-
So, in getting telnet access there seems to be two different ways. The first I followed and worked. That is installing the full .5 version of Fonz Fun Plug (FFP). The problem with that one was I couldn't unmount the drives. The other methoed I've seen mentioned on here, are the instructions listed on http://wiki.dns323.info/howto:telnet. This seems to be a much more minimalist version which has you download utelnetd, busybox3 and a version of starttelnet.sh. However, I cannot get that version to run with the updated starttelnet.sh which places the files into a ram drive. When I do that, and telnet I get something like "c0 s3, ..." on the telnet screen and then it disconnects. Any ideas what might be wrong there? And can someone tell me if once I get this working, there is a way to add parted to the minimal install or a way to get the full FFP in a ram disk?
-
sorry guys (previously, i was replying from memory),
i forgot that the sh module of busybox3 had some problem running on DNS-343
(i don't know how to fix it)
i am using /sbin/sh from the firmware
please modify your starttelnet.sh
<the path you copied to>utelnetd -l /bin/sh -d
you will not get a prompt unless you type 5784468
read the following for details
http://wiki.dns323.info/howto:ffp#the_root_user
-
Ok That make telnet running from RAM Disk.
But still the fdisk don't run...
I'm using the fdisk from this thread
http://forum.dsmg600.info/viewtopic.php?id=3375
# ./fdisk -l /dev/sda
./fdisk: No such file or directory
It seem that we have to use the original fdisk :(
It support some input and that might be what the vendor have modified it (mostly the cause of the segment fault as well)
something like
sh -c fdisk /dev/sdc n 4 83 1024 w
Create a new partation "4"
Type 83
Size 1024
I don't know what it meant for the w
Maybe for "write"
To remove a patation
sh -c fdisk /dev/sda d -1 w
I think this will be for delete all (-1)
But things is I donno how they
calculate the size....
-
did you try chmod? maybe there is no execute permission when you unpack your fdisk...
chmod +x fdisk
it should work, i am using if,
filesize and cksum as follows
root@DNS-343:~# ls -l fdisk
-rwxr-xr-x 1 root root 181440 Mar 1 02:28 fdisk
root@DNS-343:~# cksum fdisk
4117949009 181440 fdisk
sda is the new WDC EARS
sdb is the old WDC EADS
note: i am not using raid, just using rsync to mirror the disks,
because i want sdb to be asleep most of the time unless i need to restore data or accidental deletion
root@DNS-343:~# ./fdisk -ul
Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sda1 64 1060287 530112 82 Linux swap / Solaris
Partition 1 does not end on cylinder boundary.
/dev/sda2 3164800 3907029167 1951932184 83 Linux
Partition 2 does not end on cylinder boundary.
/dev/sda4 1060288 3164799 1052256 83 Linux
Partition 4 does not end on cylinder boundary.
Partition table entries are not in disk order
Disk /dev/sdb: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sdb1 63 1060289 530113+ 82 Linux swap / Solaris
/dev/sdb2 3164805 3907024064 1951929630 83 Linux
/dev/sdb4 1060290 3164804 1052257+ 83 Linux
Partition table entries are not in disk order
-
That was the really strange part. Size is correct but I don't have the cksum things...
# ls -al fdisk
-rwxr-xr-x 1 root root 181440 Mar 10 08:35 fdisk
#
# ./fdisk -ul
./fdisk: No such file or directory
something really wrong...
-
sharks, i just realized,
the fdisk that guy compiled needs some libs from ffp
and i have already installed ffp, hence i am not getting the error you got
root@DNS-343:~# ldd fdisk
libc.so.0 => /ffp/lib/libc.so.0 (0x4000e000)
ld-uClibc.so.0 => /ffp/lib/ld-uClibc.so.0 (0x40000000)
i have uploaded the libs here,
http://forum.dsmg600.info/viewtopic.php?pid=35163#p35163 (http://forum.dsmg600.info/viewtopic.php?pid=35163#p35163)
there should be enough space to put these 2 libs in /
drwxr-xr-x admin/500 0 2010-03-10 16:14:29 /ffp/
drwxr-xr-x admin/500 0 2010-03-10 16:14:40 /ffp/lib/
-rwxr-xr-x admin/500 21168 2010-03-10 16:14:09 /ffp/lib/ld-uClibc.so.0
-rw-r--r-- admin/500 322696 2010-03-10 16:14:09 /ffp/lib/libc.so.0
That was the really strange part. Size is correct but I don't have the cksum things...
# ls -al fdisk
-rwxr-xr-x 1 root root 181440 Mar 10 08:35 fdisk
#
# ./fdisk -ul
./fdisk: No such file or directory
something really wrong...
-
I've been successfully doing the fdisk things
But the raid is the problem..
I try startraid and it automatically format the new /dev/md0
But when I reboot the system...
I cannot see any volume any more...
I really have no idea what do to next...
Other then waiting for new firmware (if any)...
-
could you post the commands and responds which you tried...
I've been successfully doing the fdisk things
But the raid is the problem..
I try startraid and it automatically format the new /dev/md0
But when I reboot the system...
I cannot see any volume any more...
I really have no idea what do to next...
Other then waiting for new firmware (if any)...
-
I don't kept the log...
What I do is that I create 3 partation.
P 1 -> 512MB (Swap)
P 2 -> 1395G (Raid1)
P 4 -> 1G
And do the same for the 2nd drive.
mkswap /dev/sda1
mkswap /dev/sdb1
mke2fs -j /dev/sda4 (Fail as it say the partation was being use)
mke2fs -j /dev/sdb4
However I've kill most of the application on the device like FTP, UPNP and the iTunes server
raidstart and it will start to format the /dev/md0 --> (dev/sda2 + dev/sdb2) and etc
It is quite a long process.
I try smbstart to see if I can read the volume without reboot but unfortunatelly no.
So I reboot and it is over....
-
if you format partition 4 you will lose the config files,
you will need to backup the config files somewhere
and copy them back (which will make the process much more tedious)
my original suggestion was to only recreate partition 1 and 2 and leave 4 alone
since partition 2 is the most (and probable the only 1 used, it should be sufficient
mke2fs -j /dev/sda4 (Fail as it say the partation was being use)
mke2fs -j /dev/sdb4
-
hum...
Never think of that...
I just thinking that because of Partation 4 is also affected and thus need to re-creat as well...
Maybe that was the issue...hum...
TimeZlicer do you have any idea on the size of partation 1 and partation 4?
because when I try to recreate it on another linux system and put in back the HDD it said that it is a new HDD (Partation 4 is populated with the old configuration...) any idea?
Thank You
-
partition 1(type:82 linux swap) is about 512MB
partition 4(type 83:linux) is about 1GB
i don't think D-Link firmware checks the size,
it just check the existence based on the admin data in partition 4
here's a command sequence on how to resize partition 1 and 2 while leaving 4 alone
fdisk -u /dev/sdb
The number of cylinders for this disk is set to 243201.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)
Command (m for help): p
Disk /dev/sdb: 2000.4 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sdb1 63 1060289 530113+ 82 Linux swap / Solaris
/dev/sdb2 3164805 3907024064 1951929630 83 Linux
/dev/sdb4 1060290 3164804 1052257+ 83 Linux
Partition table entries are not in disk order
Command (m for help): d
Partition number (1-4): 1
Command (m for help): n
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): 1
First sector (63-3907029167, default 63): 64
Last sector, +sectors or +size{K,M,G} (64-1060289, default 1060289):
Using default value 1060289
Command (m for help): d
Partition number (1-4): 2
Command (m for help): n
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): 2
First sector (63-3907029167, default 63): 3164808
Last sector, +sectors or +size{K,M,G} (3164808-3907029167, default 3907029167):
Using default value 3907029167
Command (m for help): t
Partition number (1-4): 1
Hex code (type L to list codes): 82
Changed system type of partition 1 to 82 (Linux swap / Solaris)
Command (m for help): p
Disk /dev/sdb: 2000.4 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sdb1 64 1060289 530113 82 Linux swap / Solaris
/dev/sdb2 3164808 3907029167 1951932180 83 Linux
/dev/sdb4 1060290 3164804 1052257+ 83 Linux
Partition table entries are not in disk order
Command (m for help): w
-
Strange...
My fdisk can only start with 63 and not configurable...
I found that..
My single enter in putty will make it two...
Something wrong?
# ./fdisk -u /dev/sda
The number of cylinders for this disk is set to 182401.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)
Command (m for help): Command (m for help): p
Disk /dev/sda: 1500.3 GB, 1500301910016 bytes
255 heads, 63 sectors/track, 182401 cylinders, total 2930277168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x5c25e06f
Device Boot Start End Blocks Id System
/dev/sda1 63 1060289 530113+ 82 Linux swap / Solaris
/dev/sda2 3164805 2930272064 1463553630 83 Linux
/dev/sda4 1060290 3164804 1052257+ 83 Linux
Partition table entries are not in disk order
Command (m for help): Command (m for help): d
Partition number (1-4): 1
Command (m for help): Command (m for help): d
Partition number (1-4): 2
Command (m for help): Command (m for help): p
Disk /dev/sda: 1500.3 GB, 1500301910016 bytes
255 heads, 63 sectors/track, 182401 cylinders, total 2930277168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x5c25e06f
Device Boot Start End Blocks Id System
/dev/sda4 1060290 3164804 1052257+ 83 Linux
Command (m for help): Command (m for help): p
Disk /dev/sda: 1500.3 GB, 1500301910016 bytes
255 heads, 63 sectors/track, 182401 cylinders, total 2930277168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x5c25e06f
Device Boot Start End Blocks Id System
/dev/sda4 1060290 3164804 1052257+ 83 Linux
Command (m for help): Command (m for help): n
Command action
e extended
p primary partition (1-4)
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): Value out of range.
Partition number (1-4): 1
First sector (63-2930277167, default 63): Using default value 63
Last sector, +sectors or +size{K,M,G} (63-1060289, default 1060289):
Using default value 1060289
-
tip#1: don't just press enter, you can type 64 yourself
you don't have to use the default which is not aligned
see the transcript in my previous post, the default is 63 and i punch in 64 when i create partition 1
First sector (63-2930277167, default 63): Using default value 63
-
That was the issue.
What happen was (See the duplicate on the input?)
The utelnetd and my putty/telnet seem to be sending double enter per input.
Command (m for help):
Command (m for help): d
Partition number (1-4):
Partition number (1-4): 1
Command (m for help):
Command (m for help): n
Command action
e extended
p primary partition (1-4)
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4):
Partition number (1-4): 1
First sector (63-3907029167, default 63): 64
-
it might be your client, what telnet client are you using?
windoze telnet? http://www.computerhope.com/telnethl.htm (http://www.computerhope.com/telnethl.htm)
That was the issue.
What happen was (See the duplicate on the input?)
The utelnetd and my putty/telnet seem to be sending double enter per input.
anyway, extra local_echo should not affect the partitions that you create or delete,
it just looks irritating, that's all
-
I'm using both putty and telnet (from win XP)
and try to set/unset the LOCAL_ECHO
That don't seem to help.
Still it is 2 x ## when I do anything and press enter.
-
i am not sure why you are getting double echo...
there is a ssh server in D-Link firmware, you can give it a try
/usr/sbin/sshd
if it works you can ssh to your DNS-343 instead of telnet and hopefully solve your double echo issue
try starting it from the telnet session and see if it works, then change your fun_plug script
I'm using both putty and telnet (from win XP)
and try to set/unset the LOCAL_ECHO
That don't seem to help.
Still it is 2 x ## when I do anything and press enter.
-
How do you start the SSHD on DNS-343?
Mine ended like this.
# /usr/sbin/sshd start
/usr/local/etc/sshd_config: No such file or directory
If I try to run ssh-keygen
# ssh-keygen
Generating public/private rsa key pair.
Enter file in which to save the key (/mnt/HD_a2/Anonymous_ftp/.ssh/id_rsa):
Could not create directory '/mnt/HD_a2/Anonymous_ftp/.ssh'.
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
open /mnt/HD_a2/Anonymous_ftp/.ssh/id_rsa failed: No such file or directory.
Saving the key failed: /mnt/HD_a2/Anonymous_ftp/.ssh/id_rsa.
#
Strange...
-
try this
/usr/sbin/ssh_daemon start
How do you start the SSHD on DNS-343?
-
Hum Still no luck.
# /usr/sbin/ssh_daemon start
$Generating SSH2 DSA host key: $DSA key generation
Starting sshd:/etc/ssh/sshd_config line 13: Unsupported option GSSAPIAuthentication
/etc/ssh/sshd_config line 14: Unsupported option GSSAPICleanupCredentials
/etc/ssh/sshd_config line 15: Unsupported option UsePAM
Privilege separation user sshd does not exist
#
-
did your telnet (double echo) caused you any problem when you try to fdisk?
i would think double echo is just a visual thing and should not affect...
-
But it caused the "using default 1st sector 63"
That options...
-
can you post the output of
printenv
-
# printenv
printenv: No such file or directory
#
# printenv
printenv: No such file or directory
But env will show this
# env
USER=root
OLDPWD=/
HOME=/
TERM=vt102
PATH=/usr/bin:/bin:/usr/sbin:/sbin:/bin/busybox3.dir
SHELL=/bin/sh
PWD=/mnt/HD_a2/lnx_bin
-
try the following
export TERM=xterm
and
export TERM=linux
-
ok what happen after that?
-
it changes the terminal emulation, try both (1 by 1) and see if they get rid of your double echo
ok what happen after that?
-
Er.. sorry I don't quite get what you meant...
any guide?
Thank You
it changes the terminal emulation, try both (1 by 1) and see if they get rid of your double echo
-
enter
export TERM=xterm
and press enter a few times and see if you still getting double echo
then enter
export TERM=linux
and press enter a few times and see if you still getting double echo
Er.. sorry I don't quite get what you meant...
any guide?
Thank You
-
Still the same...
-
Finally got it working...
Under putty --> Telnet
Disable "Retuen Key sends. Telnet New Line instate of ^M"
That will fix the prompt things and I can do the fdisk now..
Formatting now...
let see what happen next.
-
great to hear that! post your success story!
btw i have never got your double echo problem in windows and linux,
are you using mac?
Finally got it working...
Under putty --> Telnet
Disable "Retuen Key sends. Telnet New Line instate of ^M"
That will fix the prompt things and I can do the fdisk now..
Formatting now...
let see what happen next.
-
Hum... No success story...
After I reboot ...
The Volume is gone...
I really have no idea why it happen....
I just do the fdisk.
mkswap /dev/sda1
mke2fs -j /dev/sda2
that all :(
I'm on windows xp.
-
when you do the resize on part 1,2 if you have two disks setup in raid 1 will that maintain your Raid setup? it looks like it will and I'm in the same boat as dcmwai with the same drives -- not that I have an issue re-formating if required but am I correct in assuming you can just do a resize of the partitions instead of formating?
-
could you post the output of hd_verify?
[root@DNS-343 ~]$ hd_verify
Hum... No success story...
After I reboot ...
The Volume is gone...
I really have no idea why it happen....
I just do the fdisk.
mkswap /dev/sda1
mke2fs -j /dev/sda2
that all :(
I'm on windows xp.
-
I'm actually remove the raid and using single disk now...
But still have that issue...
Really wonder why it cannot mount.
Format it wrongly?
hum...
-
i think you should read http://forums.dlink.com/index.php?topic=11631.msg69234#msg69234 (http://forums.dlink.com/index.php?topic=11631.msg69234#msg69234) again
I'm actually remove the raid and using single disk now...
But still have that issue...
Really wonder why it cannot mount.
Format it wrongly?
hum...
-
Thank TimeZlicer,
Finally get it working.
Ah what a long way.
-
cool! you can finally post your success story! ;)
Thank TimeZlicer,
Finally get it working.
Ah what a long way.
-
TimeZlicer thanks for your help on this. I have a WD20EARS and DNS-323.
I tried doing the partition myself using the steps you laid out but my 4th partition's numbers didn't match. I didn't think anything of it and comitted the changes and rebooted. Upon reboot the drive light is showing up purple, so looks like it has corruption. Not a big deal since no data on it, but what do you think I'm doing wrong. Here's my fdisk results.
The number of cylinders for this disk is set to 243201.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
(e.g., DOS FDISK, OS/2 FDISK)
Command (m for help): p
Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sda1 63 1060289 530113+ 82 Linux swap / Solaris
/dev/sda2 2088450 3907024064 1952467807+ 83 Linux
/dev/sda4 1060290 2088449 514080 83 Linux
Partition table entries are not in disk order
Command (m for help): d
Partition number (1-4): 1
Command (m for help): n
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): 1
First sector (63-3907029167, default 63): 64
Last sector, +sectors or +size{K,M,G} (64-1060289, default 1060289):
Using default value 1060289
Command (m for help): d
Partition number (1-4): 2
Command (m for help): n
Command action
e extended
p primary partition (1-4)
p
Partition number (1-4): 2
First sector (63-3907029167, default 63): 3164808
Last sector, +sectors or +size{K,M,G} (3164808-3907029167, default 3907029167):
Using default value 3907029167
Command (m for help): t
Partition number (1-4): 1
Hex code (type L to list codes): 82
Changed system type of partition 1 to 82 (Linux swap / Solaris)
Command (m for help): p
Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sda1 64 1060289 530113 82 Linux swap / Solaris
/dev/sda2 3164808 3907029167 1951932180 83 Linux
/dev/sda4 1060290 2088449 514080 83 Linux
Partition table entries are not in disk order
Command (m for help): w
The partition table has been altered!
Calling ioctl() to re-read partition table.
WARNING: Re-reading the partition table failed with error 16: Device or resource busy.
The kernel still uses the old table.
The new table will be used at the next reboot.
Syncing disks.
-
did you mkswap partition 1, format partition 2 and run hd_verify before you reboot?
if not, surely you will get some error,
because the data in the original partition 4 is assuming that partition 1 and 2 are already formated
(i assumed you have chosen ext3, if not drop the -j option in mke2fs
mkswap /dev/sda1
mke2fs -j -m 0 -T largefile4 /dev/sda2
hd_verify -w
TimeZlicer thanks for your help on this. I have a WD20EARS and DNS-323.
I tried doing the partition myself using the steps you laid out but my 4th partition's numbers didn't match. I didn't think anything of it and comitted the changes and rebooted. Upon reboot the drive light is showing up purple, so looks like it has corruption. Not a big deal since no data on it, but what do you think I'm doing wrong. Here's my fdisk results.
in your case you can use 2088456 for partition 2
ceiling(2088450 / 8) * 8 = 2088456
-
You're right, I didn't format the swap and 2nd partition. I was able to successfully execute the mkswap command but the mke2fs (yes, ext2 so no -j option) yields that the partition is still mounted and I can't format. How do I go about formatting it? I tried the -F (force) option but that didn't work.
-
if you have created partition 2 correctly using fdisk,
this should be able to format partition 2 as ext2
mke2fs -m 0 /dev/sda2
if you still face problem, please post some transcript,
or it would be difficult to help you
You're right, I didn't format the swap and 2nd partition. I was able to successfully execute the mkswap command but the mke2fs (yes, ext2 so no -j option) yields that the partition is still mounted and I can't format. How do I go about formatting it? I tried the -F (force) option but that didn't work.
-
TimeZlicer,
After I format, how can I tell if the drive (WD15EARS) is properly aligned?
This was my parted output prior to formating /dev/sdb
Number Start End Size Type File system Flags
1 64s 1060287s 1060224s primary linux-swap
2 2088450s 2930272064s 2928183615s primary ext2
Also. I tried to create/change the partitions with parted, but it complained that the disk signature was invalid and wouldn't do it so I jumped to the mke2fs steps. Is this something to be concerned about?
-
you can tell before you format, during partition creation
in your case,
partition 1 is ok (64 is divisible by 8 )
partition 2 is not ok (2088450 is not divisible by 8 )
could you post more details on the error message from parted?
have you tried fdisk?
mke2fs only creates the filesystem, alignment is only done during partition creation
reason for divisible by 8 is that the physical sector is 4096, however the drive is reporting 512,
hence partition should start on a number that is divisible by 8 to be aligned (8x512 = 4096)
TimeZlicer,
After I format, how can I tell if the drive (WD15EARS) is properly aligned?
This was my parted output prior to formating /dev/sdb
Number Start End Size Type File system Flags
1 64s 1060287s 1060224s primary linux-swap
2 2088450s 2930272064s 2928183615s primary ext2
Also. I tried to create/change the partitions with parted, but it complained that the disk signature was invalid and wouldn't do it so I jumped to the mke2fs steps. Is this something to be concerned about?
-
Thanks TimeZlicer. I figured out the divisible by 8 after googling for a bit. So I need to solve the parted issue.
These are the errors I see (I forgot to copy the signature error previously):
(parted) rm 2
Error: Partition /dev/sdb2 is being used. You must unmount it before you modify it with Parted.
(parted) rm 1
Error: Invalid partition table - recursive partition on /dev/sdb.
parted: invalid token: 1
I tried fdisk, but it crashes.
-
Replicated the signature error again with parted
rm 2
Error: Partition /dev/sdb2 is being used. You must unmount it before you modify it with Parted.
(parted) unit s
(parted) print
Model: WDC WD15EARS-00Z5B1 (scsi)
Disk /dev/sdb: 2930277168s
Sector size (logical/physical): 512B/512B
Partition Table: msdos
Number Start End Size Type File system Flags
2 2088450s 2930272064s 2928183615s primary
(parted) mkpart primary linux-swap 64 1060287
(parted) print
Model: WDC WD15EARS-00Z5B1 (scsi)
Disk /dev/sdb: 2930277168s
Sector size (logical/physical): 512B/512B
Partition Table: msdos
Number Start End Size Type File system Flags
2 2088450s 2930272064s 2928183615s primary
(parted) print
Error: Invalid partition table on /dev/sdb -- wrong signature 0.
-
you must umount the partitions before you run parted
Error: Partition /dev/sdb2 is being used. You must unmount it before you modify it with Parted.
swapoff /dev/sdb1
umount /dev/sdb2
umount /dev/sdb4
please read my previous post again on minimal fun_plug for telnet
-
I did unmount everything. And mount reported that nothing was mounted prior to parted running. At this point, I've given up on parted. I ran the fdisk executable that you recommended earlier in the thread and this is working. Each partition begins on a sector divisible by 8. I'm formatting now and will know shortly if this works. Will update everyone on my results (hopefully positive).
-
fdisk worked and after reboot my drive (sdb) is properly aligned. I'm now backing everything up on /dev/sda up and will try the same technique. Hopefully, everything will unmount from /dev/sda.
-
cool! congrats!
fdisk worked and after reboot my drive (sdb) is properly aligned. I'm now backing everything up on /dev/sda up and will try the same technique. Hopefully, everything will unmount from /dev/sda.
for me, properly aligned EARS writes marginally faster than EADS
writing speed for non-properly aligned EARS is on the average 20% slower than EADS
EARS is much cooler that EADS, 5-8C in most cases and 10C in rare occasions
BTW, cp is faster than rsync when you try to copy from disk to disk
(i assume you have optware installed, the firmware cp don't have the -u option)
cp -auv /mnt/HD_b2/AAAAAAA /mnt/HD_a2/
is the same as
rsync -av /mnt/HD_b2/AAAAAAA/ /mnt/HD_a2/AAAAAAA/
-
/mnt/HD_a2 is proving trickier to unmount than /dev/sdb. Any tips on how I can unmount it cleanly? lsof doesn't report any files being active on that mount point after I've stopped smb, nfs and turned off swapping. I suspect its remants of ffp that are still attached to that mount point.
Is there any way to launch ffp from /dev/sdb and not /mnt/HD_a2?
-
Solved my problem. Updated funplug to use /mnt/HD_b2 and was able to unmount /dev/sda and reformat on aligned partitions.
Thanks for the cp tip. Will use it for my nightly backups.
-
(i assume you have optware installed, the firmware cp don't have the -u option)
Actually, I don't have optware installed right now. What package is it part of in optware?
-
Package: coreutils
Version: 8.4-1
Section: core
Architecture: arm
Maintainer: NSLU2 Linux <nslu2-linux@yahoogroups.com>
MD5Sum: b3efb1df534aaf86228ba255c9ab3dc8
Size: 1985358
Filename: coreutils_8.4-1_arm.ipk
Source: http://ftp.gnu.org/pub/gnu/coreutils/coreutils-8.4.tar.gz
Description: Bunch of heavyweight *nix core utilities
Actually, I don't have optware installed right now. What package is it part of in optware?
-
TimeZlicer,
I was able to successfully unmount /dev/sda2 and ran fdisk. Everything worked until I got to the mke2fs command. I still tried hd_verify and rebooted but drive is coming up with pink again and will need to format. Any idea what's going on? I'm using ext2.
Command (m for help): p
Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sda1 64 1060289 530113 82 Linux swap / Solaris
/dev/sda2 3164808 3907029167 1951932180 83 Linux
/dev/sda4 1060290 2088449 514080 83 Linux
Partition table entries are not in disk order
Command (m for help): w
The partition table has been altered!
Calling ioctl() to re-read partition table.
Syncing disks.
# mkswap /dev/sda1
Setting up swapspace version 1, size = 542830592 bytes
# mke2fs -m 0 /dev/sda2
*** Device name for mke2fs: /dev/sda2
mke2fs 1.41.2 (02-Oct-2008)
/dev/sda2: Memory allocation failed while setting up superblock
exit status 1
# ^[[A : not found
# in/busybox3.dir/sh:
# mke2fs -m 0 /dev/sda2
*** Device name for mke2fs: /dev/sda2
mke2fs 1.41.2 (02-Oct-2008)
/dev/sda2: Memory allocation failed while setting up superblock
exit status 1
#
-
maybe you ran out of memory, try
swapon /dev/sda1
before running mke2fs
TimeZlicer,
I was able to successfully unmount /dev/sda2 and ran fdisk. Everything worked until I got to the mke2fs command. I still tried hd_verify and rebooted but drive is coming up with pink again and will need to format. Any idea what's going on? I'm using ext2.
Command (m for help): p
Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sda1 64 1060289 530113 82 Linux swap / Solaris
/dev/sda2 3164808 3907029167 1951932180 83 Linux
/dev/sda4 1060290 2088449 514080 83 Linux
Partition table entries are not in disk order
Command (m for help): w
The partition table has been altered!
Calling ioctl() to re-read partition table.
Syncing disks.
# mkswap /dev/sda1
Setting up swapspace version 1, size = 542830592 bytes
# mke2fs -m 0 /dev/sda2
*** Device name for mke2fs: /dev/sda2
mke2fs 1.41.2 (02-Oct-2008)
/dev/sda2: Memory allocation failed while setting up superblock
exit status 1
# ^[[A : not found
# in/busybox3.dir/sh:
# mke2fs -m 0 /dev/sda2
*** Device name for mke2fs: /dev/sda2
mke2fs 1.41.2 (02-Oct-2008)
/dev/sda2: Memory allocation failed while setting up superblock
exit status 1
#
-
Hello,
Maybe I'm asking too much but could somebody who successfully managed to re-align and format (RAID 1) write a short summary or step by step of what needs to be done from the very beginning when you install fun_plug up to the end.
What kind of fun_plug do you install? I couldn't unumount with the regular fun_plug. I went through all the 5 pages at least 3 times but I'm still missing something.
Just by searching on google I found a lot of other guys out there struggling with the same issue. Even from Dlink tech team would be nice if somebody would write a step by step or a summary for a beginner linux user.
Regards,
Albert
-
maybe you ran out of memory, try
swapon /dev/sda1
before running mke2fs
That did it TimeZlicer. Thanks for all your help, couldn`t have done it without you!
-
;D cheers!
That did it TimeZlicer. Thanks for all your help, couldn`t have done it without you!
-
Hello,
Maybe I'm asking too much but could somebody who successfully managed to re-align and format (RAID 1) write a short summary or step by step of what needs to be done from the very beginning when you install fun_plug up to the end.
Just by searching on google I found a lot of other guys out there struggling with the same issue. Even from Dlink tech team would be nice if somebody would write a step by step or a summary for a beginner linux user.
I 2nd this..and nominate TimeZlicer... :) Hey, I'll offer you a batch of choc. chip cookies made by my daughter! :)
-
gee... thanks! i have tried before and was not successful...
will give it another go...
I 2nd this..and nominate TimeZlicer... :) Hey, I'll offer you a batch of choc. chip cookies made by my daughter! :)
-
I ran some iozone tests on 4K sector drive, and the key is not to write in 4K-8K chunks., other than this, it works well. Why there is a perception that it only affects Windows is because by default, NTFS uses 4K cluster. For Linux, depends on how you mke2fs, the block size stride all play a part.
-
gee... thanks! i have tried before and was not successful...
will give it another go...
you tried to format the drives to work the the NAS unsuccessfully or tried to write up a guide? Sorry, just a bit confused there...... :)
-
i was unsuccessful when i tried to create aligned raid partitions...
you tried to format the drives to work the the NAS unsuccessfully or tried to write up a guide? Sorry, just a
bit confused there...... :)
-
i was unsuccessful when i tried to create aligned raid partitions...
Hrrrrm... that's not good. I had hoped to use two 2TB EARS drives in RAID config.
You're saying it's not as "simple" as I thought to just format the two drives identically in Linux then setup the RAID array, eh?
-
Hrrrrm... that's not good. I had hoped to use two 2TB EARS drives in RAID config.
You're saying it's not as "simple" as I thought to just format the two drives identically in Linux then setup the RAID array, eh?
Eh, I'm too old for this hassle, going to return to Newegg and get the Sammy 2TB F3EG's. Ah well, glad I hadn't opened them yet...
-
Anyone with progress to report on getting EARS partitions realigned for 4K?
I have my 323 up and running with the hardware and RAID0, but I haven't yet tried to realign, based on time constraints and reports of no-success.
I might take a crack at it this weekend; since I haven't yet used it to store data I'm not concerned about trashing the partitions.
-
Finally, I manage to get the RAID1 with the alignment things.
It wasn't that hard but it do take times. (I'm not too sure why but there are 2 re-sync happeded)
Below are the guide.
http://docs.google.com/document/pub?id=1FIcG6ORtV9yqIK-4MLNgNLlQeHOb3Fxpu_OpLbd4Npg (http://docs.google.com/document/pub?id=1FIcG6ORtV9yqIK-4MLNgNLlQeHOb3Fxpu_OpLbd4Npg)
Fell free to let me know if you encounter any issue.
Thank You
-
i salute you efforts, great post! :o
Finally, I manage to get the RAID1 with the alignment things.
It wasn't that hard but it do take times. (I'm not too sure why but there are 2 re-sync happeded)
Below are the guide.
http://docs.google.com/document/pub?id=1FIcG6ORtV9yqIK-4MLNgNLlQeHOb3Fxpu_OpLbd4Npg (http://docs.google.com/document/pub?id=1FIcG6ORtV9yqIK-4MLNgNLlQeHOb3Fxpu_OpLbd4Npg)
Fell free to let me know if you encounter any issue.
Thank You
-
I wonder anyone would test it on a RAID5 and see if the WDC Engineer was correct on saying that it don't support that.
However I do suggest that you turn off the Disk hibernation as it seem that the spin time is higher then normal after the HDD go hibernation.
And I wonder if it can harm the array by declaring it removed...
-
Finally, I manage to get the RAID1 with the alignment things.
It wasn't that hard but it do take times. (I'm not too sure why but there are 2 re-sync happeded)
Below are the guide.
http://docs.google.com/document/pub?id=1FIcG6ORtV9yqIK-4MLNgNLlQeHOb3Fxpu_OpLbd4Npg (http://docs.google.com/document/pub?id=1FIcG6ORtV9yqIK-4MLNgNLlQeHOb3Fxpu_OpLbd4Npg)
Fell free to let me know if you encounter any issue.
Thank You
Awesome, thanks! I'm using this guide to get myself set up now. ;D
I noticed that you didn't realign the last partition on each drive. Why did you choose not to do that?
-
This is the guide for single HDD.
Well actually if you are following the thread you would be able to do that already.
But still having a guide would be better.
http://docs.google.com/document/pub?id=1ZnfQ8kvkBAHSa4EsT9kRPzPVxLz1zU2MLQKLmMROF7A (http://docs.google.com/document/pub?id=1ZnfQ8kvkBAHSa4EsT9kRPzPVxLz1zU2MLQKLmMROF7A)
Thank You
-
If anyone is slow on the raid rebuild...
Try
echo 150000 > /proc/sys/dev/raid/speed_limit_max
This should kick the process to 150MB/s
But will slow down all other process...
But I donno if that will caused the system to halt :)
Please do let me know...
-
Please check on the doc again
I've make some change on the telnetd link.
Thank You
-
Hum...
You can't touch that partation..
once you touch it ..
The RAID will die.
I think that the reason is that it is unable to mount it..
Any auggestion about the partation format type and etc?
this would be the last to try....
For though of you who have the issue, temporaty would be disable the upnp av, itune server and other which need to store on this partition...
-
I just kill my Raid again...
Anyone have anyway to telnet to the system without HDD?
The volume is not detected... :(
They think that this is a new hdd...
OMG...
I need some help here...
-
Finally... I got a firmware which have telnet enable by default and it help me to fix my raid and also the 3rd partition.
Also I've confirm that if you are using the workaround (the 3rd partation is still not align)
you will still suffer some performance issue when you are using the itune & upnpav server on the DNS-343.
Try to disable it to see if you have a better result. (For me I seem to found that disable it is way better)
Fixing the 3rd partition wasn't easy. The Firmware with telnet enable is a must also I found that this firmware will help me to fix my broken raid also :)
However I don't know if I can release it...
Let me check with them...
-
Finally... I got a firmware which have telnet enable by default and it help me to fix my raid and also the 3rd partition.
Also I've confirm that if you are using the workaround (the 3rd partation is still not align)
you will still suffer some performance issue when you are using the itune & upnpav server on the DNS-343.
Try to disable it to see if you have a better result. (For me I seem to found that disable it is way better)
Fixing the 3rd partition wasn't easy. The Firmware with telnet enable is a must also I found that this firmware will help me to fix my broken raid also :)
However I don't know if I can release it...
Let me check with them...
That really doesn't resolve the problem. In fact I believe after you reboot the unit it will go back to what it was doing before and you will have to telnet in again and fix the mount. Also since that is not a final release version of firmware the simplest answer is no, you can't/shouldn't release it. That version you are using is based on an old version of 1.03 which has some critical bugs in it.
-
That really doesn't resolve the problem. In fact I believe after you reboot the unit it will go back to what it was doing before and you will have to telnet in again and fix the mount. Also since that is not a final release version of firmware the simplest answer is no, you can't/shouldn't release it. That version you are using is based on an old version of 1.03 which has some critical bugs in it.
Maybe for you it didn't but for me.
It did help to solve my problem.
My Raid 1 come back.
(I've a backup of the HD_a4 and HD_b4) before I change the 3rd partation.
I just restore the files after the fdisk and Good my raid is back online.
I know what you would not like to see "unauthorized" firmware flying around.
But Frankly speaking, This firmware with telnet do help me a lot. (In this case)
If I've this firmware 3 months ago might no need to do linux read and etc things...
-
hi,
I'm new to linux and was able to telnet to my DLINK dns-323. I followed this link's instructions (http://docs.google.com/document/pub?id=1ZnfQ8kvkBAHSa4EsT9kRPzPVxLz1zU2MLQKLmMROF7A)
and i get stuck at this part of the telnet:
jeff@xbmc:~$ telnet 192.168.0.192
Trying 192.168.0.192...
Connected to 192.168.0.192.
Escape character is '^]'.
stdin, stdout, etderr: 0 1 2
BusyBox v1.2.1 (2006.10.30-10:27+0000) Built-in shell (ash)
Enter 'help' for a list of built-in commands.
# /tmp/fdisk -lu
/bin/busybox3.dir/sh: /tmp/fdisk: not found
# ./fdisk -lu
/bin/busybox3.dir/sh: ./fdisk: not found
Here is my starttelnet.sh:
# improved starttelnet.sh, enabling the filesystems to be unmounted and checked, if need be.
# copy the provided components to a directory on the ramdisk
# the ramdisk is regenerated with every boot, so the copy has no lasting effects at all
# simply copy the two files over
cp /mnt/HD_a2/lnx_bin/utelnetd /sbin/utelnetd
cp /mnt/HD_a2/lnx_bin/busybox3 /bin/busybox3
cp /mnt/HD_a2/ffp /ffp
cp /mnt/HD_a2/fdisk /tmp/fdisk
# create the terminal device as usual
/bin/busybox3 mknod /dev/ptyp0 c 2 0
/bin/busybox3 chmod 0666 /dev/ptyp0
/bin/busybox3 mknod /dev/ttyp0 c 3 0
/bin/busybox3 chmod 0666 /dev/ttyp0
# make a shell link on the ramdisk
mkdir /bin/busybox3.dir/
PATH="$PATH:/bin/busybox3.dir"
ln -s /bin/busybox3 /bin/busybox3.dir/sh
# and start the Telnet service from the ramdisk as well
/sbin/utelnetd -l /bin/busybox3.dir/sh -d
what am i doing wrong? Any help would be appreciated thanks.
-
My raid is up. But the speed is horrible. I have a direct 1Gb link and before the alignment the transfer rate was 2.5 M/sec, now its 4.5 M/s. Its not even 100Mb link. This is terrible. When the raid was re-syncing, the speed was about 100 000 K/s from # cat /proc/mdstat command. Twice higher than in the guide. I am so disappointed. If anybody has any suggestions I’d really appreciate it.
For others, who is trying (hoping) to get it working, here’s my experience, I hope other people can get something from it.
1 x DNS-321
2 x 2TB WD EARS
Linux experience – eeh, I’ve heard about it. (Beginner)
Looks like I’ve got all the problems mentioned in this discussion and some more. I will not go through all of them, but want to mention some things.
1. fdisk: not found.
The line in starttelnet.sh:
cp /mnt/HD_a2/ffp /ffp
does not work for me. There is no /ffp after reboot. So, I had to modify my starttelnet.sh as:
mkdir /ffp
mkdir /ffp/lib
cp /mnt/HD_a2/ffp/lib/*.* /ffp/lib
Now fdisk can find all the libs and starts just fine.
2. Double echo with telnet (can’t modify first sector in fdisk). Just hold Ctrl when you press Enter. That’s it. You only need to do that in fdisk. Other than that place everything works ok with duplicating the Enter.
3. Some files are missing (again I am using DNS-321), or in a different places. The most important
/mnt/HD_a4/mdadm.conf and
/mnt/HD_b4/mdadm.conf
In my setup they are:
/mnt/HD_a4/.systemfiles/mdadm.conf and
/mnt/HD_b4/.systemfiles/mdadm.conf
I also didn’t have
write_hdverify
I did
hd_verify –w
just for the heck of it. Don’t know if it is the same. It gave me bunch of crap, but who cares. Everything worked.
That is probably it. I followed the guide and was able to recreate raid with partitions aligned the way it should be. Thank you very much for the guide and all your help.
-
Finally, I manage to get the RAID1 with the alignment things.
It wasn't that hard but it do take times. (I'm not too sure why but there are 2 re-sync happeded)
Below are the guide.
http://docs.google.com/document/pub?id=1FIcG6ORtV9yqIK-4MLNgNLlQeHOb3Fxpu_OpLbd4Npg (http://docs.google.com/document/pub?id=1FIcG6ORtV9yqIK-4MLNgNLlQeHOb3Fxpu_OpLbd4Npg)
Fell free to let me know if you encounter any issue.
Thank You
Hi.
Thanks for the guide. I'm part way through it but having a few problems. Was wondering if you could help me out:
1) After I Telnet into the 323, I see the "#". Am I supposed to type in 5784468? I tried it the first time, and I got an error message, but then the prompts started appearing as "# #".
2) When I tried the "cat /mnt/HD_b4/raidtab2web" command, I got an error. I copied and pasted this command into telnet, but it keeps giving me an error. What am I doing wrong?
Thanks.
-
Hi. I have realigned my 1.5TB WDEARS GREEN but upload speed to my DNS343 is capping at about 140Mbps. The download speed is about 300Mbps.
is that normal? i thought that was a little slow compared to what i googled.
just want to sort out the peformance issues before i start putting data back into the drive.
any help is appreciated
-
Omg...
That was important but if seem that I've forgot...
My bad. Will modify the guide accordingly..
Thank for letting me know.
cp /mnt/HD_a2/ffp /ffp
does not work for me. There is no /ffp after reboot. So, I had to modify my starttelnet.sh as:
mkdir /ffp
mkdir /ffp/lib
cp /mnt/HD_a2/ffp/lib/*.* /ffp/lib
Now fdisk can find all the libs and starts just fine.
2. Double echo with telnet (can’t modify first sector in fdisk). Just hold Ctrl when you press Enter. That’s it. You only need to do that in fdisk. Other than that place everything works ok with duplicating the Enter.
3. Some files are missing (again I am using DNS-321), or in a different places. The most important
/mnt/HD_a4/mdadm.conf and
/mnt/HD_b4/mdadm.conf
In my setup they are:
/mnt/HD_a4/.systemfiles/mdadm.conf and
/mnt/HD_b4/.systemfiles/mdadm.conf
I also didn’t have
write_hdverify
I did
hd_verify –w
-
my success story, non-RAID (SLEDs, JBOD; RAID makes no sense for me): http://www.consumedconsumer.org/2010/06/upgrading-my-dns-323-to-2-x-2-tb-ii_08.html
-
Hello Everyone!
Dcmwai,
Thanks for the Single HDD 4K Alignment issue Guide acticle. I am trying to use it for my D-Link DNS-323 with no success. I've made the following according to the instruction:
1) Format drive
2) Reboot
3) Extract files from utelnetd-ram.tar.gz into \Volume_1\ folder
4) Reboot
Due to some reason I am unable to connect using Telnet. File dmesg.out was created. Any ideas how to debug this?
This is the guide for single HDD.
Well actually if you are following the thread you would be able to do that already.
But still having a guide would be better.
http://docs.google.com/document/pub?id=1ZnfQ8kvkBAHSa4EsT9kRPzPVxLz1zU2MLQKLmMROF7A (http://docs.google.com/document/pub?id=1ZnfQ8kvkBAHSa4EsT9kRPzPVxLz1zU2MLQKLmMROF7A)
Thank You
-
Hey Pavlo,
I'm experiencing the same issue. I suspect we're missing something, perhaps even a option setting within our telnet command? I'm more accustomed to SSH so I have to familiarize myself with telnet. Any luck on this? I receive the following:
$ telnet 192.168.1.26 -l root
Trying 192.168.1.26...
Connected to 192.168.1.26.
Escape character is '^]'.
Connection closed by foreign host.
From what I could find online this would suggest that our client is not on the telnet trusted list, but I suspect there isn't a trusted list in this case so it should accept connections from anywhere.
-
Pavlo,
It seems the files included within utelnetd-ram.tar.gz have some lines commented out that shouldn't be. Have a look here for the correct details on getting telnet working:
http://wiki.dns323.info/howto:telnet (http://wiki.dns323.info/howto:telnet)
-
Hello everyone and thanks for help,
Actually I've resolved the trouble by just connecting second (temporary) 80GB HDD together with those 2TB WD.
This scenario is more simple to be done as does not require using of RAM DRIVE.
I would lovely try other cases to give something back for community but I am afraid to make my HDD inaccessible while having important data there.
-
Hello everyone,
I don't know if any of you is still watching this thread, but I had a question for any of you that was able to repartition/reformat the drives for the 4K alignement issue.
First of all, thanks a lot to dcmwai for his step by step procedure and also to maxbis who was kind enough to post corrections to dcmwai's guide. I followed these steps, had a bit of problems at the begining, but in the end, I managed to partition and format my disks properly (well I think). After synching completed, I ran some last checks (as suggested by dcmwai), I was able to start/stop the raid array properly and I rebooted the device.
The smb share is up, I can read/write stuff on it. However, when I start dlink's web page, it asks me to configure my drivers (raid type, format, etc...). If I click skip, most of the settings are available, under Tools -> Raid, it says:
Current RAID Type :
Volume_1 : RAID 1 (Degraded)
And I have a button "Set RAID Type and Re-Format"
Under Tools -> Disk Tools, I can't use the Scan Disk feature.
Anyone experienced the same thing? Any suggestion? Could it be related to the fact that the guid of my raid changed (i.e., could linksys's software refer to that guid?)
Thanks,
Sebastien
-
Have you seen this thread? http://forums.dlink.com/index.php?topic=14484.0
I believe I hit the same problem and it was due to
WARNING: Re-reading the partition table failed with error 16: Device or resource busy.
The kernel still uses the old table.
The new table will be used at the next reboot.
Try:
swapoff /dev/sdb1
in step (6).
-
please help !
I get stuck at the end...
when I type
#mdadm -Es > /mnt/HD_b4/.systemfile/mdadm.conf
response is
cannot create /mnt/HD_b4/.systemfile/mdadm.conf: No such file or directory
what I have to do ?
-
I tried one more time to align in RAID1 but unsuccessful
just finished to sync and #cat /proc/mdstat give this:
Personalities : [linear] [raid0] [raid1]
md0 : active raid1 sdb2[1] sda2[2](F)
1951171776 blocks [2/1] [_U]
resync=DELAYED
unused devices: <none>
# mdadm --detail /dev/md0
/dev/md0:
Version : 00.90.01
Creation Time : Thu Nov 25 02:28:21 2010
Raid Level : raid1
Array Size : 1951171776 (1860.78 GiB 1998.00 GB)
Device Size : 1951171776 (1860.78 GiB 1998.00 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 0
Persistence : Superblock is persistent
Update Time : Thu Nov 25 06:15:23 2010
State : active, degraded
Active Devices : 1
Working Devices : 1
Failed Devices : 1
Spare Devices : 0
UUID : 93671b49:337d2c92:cdb3fb02:17f99cc2
Events : 0.4
Number Major Minor RaidDevice State
0 0 0 0 removed
1 8 18 1 active sync /dev/sdb2
2 8 2 - faulty spare /dev/sda2
# /tmp/fdisk -ul
Disk /dev/sda: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sda1 64 1060289 530113 82 Linux swap / Solaris
/dev/sda2 2088456 3904432206 1951171875+ 83 Linux
/dev/sda4 1060290 2088449 514080 83 Linux
Partition table entries are not in disk order
Disk /dev/sdb: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders, total 3907029168 sectors
Units = sectors of 1 * 512 = 512 bytes
Disk identifier: 0x00000000
Device Boot Start End Blocks Id System
/dev/sdb1 64 1060289 530113 82 Linux swap / Solaris
/dev/sdb2 2088456 3904432206 1951171875+ 83 Linux
/dev/sdb4 1060290 2088449 514080 83 Linux
Partition table entries are not in disk order
Please can someone give me help ?
-
:'( :'(
-
Does anyone know if fdisk should show the same (properly aligned information) after formatting and rebooting? I just happened to check and it shows this now:
Command (m for help): p
Disk /dev/sdb: 2000.3 GB, 2000398934016 bytes
255 heads, 63 sectors/track, 243201 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0xb6b1183f
Device Boot Start End Blocks Id System
/dev/sdb1 1 66 530112 82 Linux swap / Solaris
Partition 1 does not end on cylinder boundary.
/dev/sdb2 197 243202 1951932184 83 Linux
/dev/sdb4 66 197 1052256 83 Linux
Partition 4 does not end on cylinder boundary.
Partition table entries are not in disk order
Command (m for help):
-
Hi,
If one of the hhd failed and requires rebuild of the raid after align of the sector, do we need to re-align again?? or just let it rebuild the raid will do? ??? ???
-
Ok... following this thread and the guide... and ripping my hair out (don't worry I was already balding...) I have finally gotten this to mostly work.
EDIT: I am using a DNS-323 with two 2TB seagate drives.
However the first time the partition wasn't unmounted (sdb4 was sticking for some reason) and upon reboot I had individual volumes instead of raid... I just tried again making sure no partitions were mounted and now I got through everything without error except for one thing...
# mdadm -Es > /mnt/HD_b4/mdadm.conf
cannot create /mnt/HD_b4/mdadm.conf: No such file or directory
# mdadm -Es > /mnt/HD_a4/mdadm.conf
cannot create /mnt/HD_b4/mdadm.conf: No such file or directory
# mdadm -Es > /etc/mdadm.conf
# cat /mnt/HD_b4/mdadm.conf
cat: /mnt/HD_b4/mdadm.conf: No such file or directory
# cat /mnt/HD_a4/mdadm.conf
cat: /mnt/HD_b4/mdadm.conf: No such file or directory
# write_hdverify
write_hdverify: No such file or directory
# hd_verify -w
hd verify v1.23.10072009
Force write verify to hd
WriteMagicNumToMTD 486036
Can not mount md0
There is nothing in HD_[ab]4 (except an empty folder called backup) and the hd_verify -2 can not mount md0?
I just finished formatting and am waiting for the sync to complete so I can reboot and see if this worked...
EDIT: It finished and from the web interface it showed the raid 1 with sync completed but upon restarting the unit I had two separate volumes...
Just telneted back in and fdisk shows all the partitons are still correctly there and both volumes work fine... it just didn't stick with the raid...
EDIT2: Since the partitioning stuck I just tried recreating the raid 1 array. This time when I did hd_verify -w it was able to mount md0 and write something but then spat out
hd verify v1.23.10072009
Force write verify to hd
WriteMagicNumToMTD 328986
WriteMagicNumToHD 328986
cp: /mnt/HD_a4/.systemfile/hd_magic_num: No such file or directory
rm: cannot remove `/tmp/hd_verify_warning': No such file or directory
Also this time I had to umount md0 first before it would let me format... not sure why this time it was mounted and last time it wasn't... I followed all the same steps.
Waiting for it to finish syncing again...
I should also note the first time I did this when I created the raid array I had the right side giving an yellow degraded light. This time both lights are normal and blue.
-
sync completed successfully. raidstop and start did not trigger another sync. Restart DNS-323 and back to to separate volumes...
Help?
EDIT:
Ok went back in... after after going through each step I found that sda4 and sdb4 were unmounting when attempting to start md0 without a FS on it or when doing hd_verify. So I went back in, recreated the raid, formatted it, then did raidstart md0, checked mount and I had md0 as HD_a2 and sda4/sdb4 as HD_a4/HD_b4 and when I did mdadm -Es > /mnt/HD_[ab]4/mdadm.conf it worked as well as hd_verfiy -w didn't give the one error...
Waiting for sync again...
Sorry for the crampness of this post... posting quickly from my phone at work...