Schmill

What is /dev/md0 ?

Recommended Posts

What is /dev/md0 ?

 

If I try to login via the DSM web interface I get the message;

 

You cannot log into the system because the disk space is full currently. Please restart the system and try again

 

However, there is no way that my disks should be full, and if I log in via SSH and run a 'df' I get this:

 

df -h
Filesystem                Size      Used Available Use% Mounted on
/dev/md0                  2.3G      2.3G         0 100% /
/tmp                    940.2M    800.0K    939.4M   0% /tmp
/dev/vg1000/lv          224.8G      1.0G    223.7G   0% /volume1
/dev/vg1001/lv            1.8T    672.2G      1.1T  37% /volume2
/dev/sdu1                30.8M     12.5M     18.4M  40% /volumeUSB1/usbshare
/dev/sdv1                 1.8T      1.3T    495.2G  73% /volumeUSB2/usbshare

 

The only thing that is 100% used is /dev/md0, but I don't know what this is....

 

Any ideas?

 

Thanks :smile:

Share this post


Link to post
Share on other sites

Seems I may have solved it :smile:

 

So I mounted /dev/md0 as /tmp/work/ and ran "du -hs *" to find the largest directory...

 

/tmp/work # du -hs *

796.0K bin

0 core_perl

112.0K dev

6.0M etc

5.6M etc.defaults

4.0K initrd

146.7M lib

5.3M lib64

4.0K lost+found

4.0K mnt

4.0K no

188.8M opt

4.0K proc

1.5G root

12.0K run

5.2M sbin

4.0K sys

8.0K tmp

422.9M usr

4.4M var

156.0K var.defaults

4.0K volume

4.0K volume1

8.0K volume2

8.0K volumeUSB1

4.0K volumeUSB2

 

So then drilling into the 'root' subdirectory:

 

/tmp/work # cd root

/tmp/work/root # ls -la

drwxr-xr-x 5 root root 4096 Nov 23 15:45 .

drwxr-xr-x 30 root root 4096 Nov 25 14:38 ..

drwx------ 3 root root 4096 Nov 23 11:59 .config

drwx------ 2 root root 4096 Nov 12 22:15 .gnupg

-rw-r--r-- 1 root root 364 Oct 4 07:30 .profile

drwx------ 3 root root 4096 Nov 23 15:45 VirtualBox VMs

-rw-r--r-- 1 root root 1007 Nov 20 00:03 comp.pl

/tmp/work/root # du -hs *

1.5G VirtualBox VMs

4.0K comp.pl

/tmp/work/root # mv VirtualBox\ VMs/ /volume2/

 

So it seems as if by default the VirtualBox install (viewtopic.php?f=15&t=3497) locates the VM's directly on this /dev/md0, rather than on one of the volumes.

By doing a "mv VirtualBox\ VMs/ /volume2/" I have now freed space on the /dev/md0 and can log back into the web gui :smile:

Share this post


Link to post
Share on other sites

/md0 is the DSM install partition. md1 is swap file partition, and md2 is your array volume. if you have multiple volumes, you'll have md3,4, ect.

 

edit: my experience with getting data off a broken array: http://forum.cgsecurity.org/phpBB3/foun ... t2600.html

 

try googling "You cannot log into the system because the disk space is full currently. Please restart the system and try again" you'll find lots with similar problems. You probably have a misbehaving app or service filling it with log files, or mem dumps.

Share this post


Link to post
Share on other sites
Seems I may have solved it :smile:

 

So I mounted /dev/md0 as /tmp/work/ and ran "du -hs *" to find the largest directory...

 

/tmp/work # du -hs *

796.0K bin

0 core_perl

112.0K dev

6.0M etc

5.6M etc.defaults

4.0K initrd

146.7M lib

5.3M lib64

4.0K lost+found

4.0K mnt

4.0K no

188.8M opt

4.0K proc

1.5G root

12.0K run

5.2M sbin

4.0K sys

8.0K tmp

422.9M usr

4.4M var

156.0K var.defaults

4.0K volume

4.0K volume1

8.0K volume2

8.0K volumeUSB1

4.0K volumeUSB2

 

So then drilling into the 'root' subdirectory:

 

/tmp/work # cd root

/tmp/work/root # ls -la

drwxr-xr-x 5 root root 4096 Nov 23 15:45 .

drwxr-xr-x 30 root root 4096 Nov 25 14:38 ..

drwx------ 3 root root 4096 Nov 23 11:59 .config

drwx------ 2 root root 4096 Nov 12 22:15 .gnupg

-rw-r--r-- 1 root root 364 Oct 4 07:30 .profile

drwx------ 3 root root 4096 Nov 23 15:45 VirtualBox VMs

-rw-r--r-- 1 root root 1007 Nov 20 00:03 comp.pl

/tmp/work/root # du -hs *

1.5G VirtualBox VMs

4.0K comp.pl

/tmp/work/root # mv VirtualBox\ VMs/ /volume2/

 

So it seems as if by default the VirtualBox install (viewtopic.php?f=15&t=3497) locates the VM's directly on this /dev/md0, rather than on one of the volumes.

By doing a "mv VirtualBox\ VMs/ /volume2/" I have now freed space on the /dev/md0 and can log back into the web gui :smile:

 

how did you mount? i literally spent 2 hours and i cannot for the fcking life of me get my shit to work.. turns out this is my problem

 

df -h

Filesystem Size Used Available Use% Mounted on

/dev/root 2.3G 2.3G 0 100% /

 

 

 

 

I tried mnt, mount, could not do it. I cant even browse the damn root directory to see what the hell is taking up the space. I cant even enable ssh im lucky to access the server via vga cable and a keyboard but i cant start the damn ssh service :sad: if i had winscp access to the server im sure everything would be resolved. Does anyone know how to find the file that is preventing the server from being logged into.

 

the problem with ^^^^ i dont have a work dir under temp so i just did

 

Mount /dev/md0 /tmp/ (mounted and i was able to take care of business)

 

Another command that didn't work for dick was " du -hs * "

 

in a nutshell when shit hits the fan and you cant login its because there's something fucked up in the log folder so browse

 

/var/log/synolog

 

and remove any bullshit that's over a couple of megs.

 

I did hours and hours of searching all that turned up was unanswered problems and format your shit... talk about worthless answers out there. So if you get stuck its all about the syntax some of the OPS syntax did not work and that's why I blame myself for not understanding more commands and ignoring some of the ones that are posted relying on those and figuring well if they dont work im screwed.. not the case its just some commands dont work or display what you're looking for.

 

** the root cause of my system shitting the bed was plex i tried to watch a video and it was freezing halfway and being unusable. via the web gui cpu was stuck at 50% because syslog-ng would not quit and was stuck i thought it would be ok if i did a hardboot. Bad idea

Share this post


Link to post
Share on other sites

hi schmill sorry for the extremely late reply. my account was locked, and it literally took 5 months to get it unlocked. to answer your question my rant of a post / reply was yes I fixed it, but wanted to know why.

Share this post


Link to post
Share on other sites
md0 is a mdadm drive. 

 

Is it a regular type of MD or something special to Synology as md0 and md1 obviously use all available disks but in Xpenology, as it is mimicking a 12 disk system and most people do not have this fully populated, these two md devices always appear degraded e.g. for a 4 disk system 12/4 [uUUU________] ?

Share this post


Link to post
Share on other sites