Help ...../dev/mapper/pve-root got full.....

[Stryder]

I used I default installation nothing else, I have only a VM there working till today

if there is any old backup where is it??? 2 delete it

Why is /dev/mapper/pve-root fully ocupied

Filesystem Size Used Avail Use% Mounted on
/dev/mapper/pve-root 95G 95G 0 100% /
tmpfs 5.9G 0 5.9G 0% /lib/init/rw
udev 5.9G 244K 5.9G 1% /dev
tmpfs 5.9G 41M 5.9G 1% /dev/shm
/dev/mapper/pve-data 795G 197M 795G 1% /var/lib/vz
/dev/sda1 495M 34M 436M 8% /boot
overflow 1.0M 0 1.0M 0% /tmp
/dev/fuse 30M 32K 30M 1% /etc/pve


I make a search using lsof | awk '{print $7,$8,$9}' | sort -n and pops out a list of files I can´t understand what should I do???? guys this is the second time this happens 2 me

 

[tom]

see also http://forum.proxmox.com/threads/12512-pve-root-nearly-full

 

[Stryder]

I just read the post http://forum.proxmox.com/threads/12512-pve-root-nearly-full and I am so gratefull 2 someone tries 2 help me so fast but I don´t use Smb or CIFS or NFS Share so this files are factory default. And i can´t install ncdu first I have no space

I seeing this "overflow" that its mounted I dont´t what it is

Filesystem Size Used Avail Use% Mounted on
/dev/mapper/pve-root 95G 95G 0 100% /
tmpfs 5.9G 0 5.9G 0% /lib/init/rw
udev 5.9G 244K 5.9G 1% /dev
tmpfs 5.9G 41M 5.9G 1% /dev/shm
/dev/mapper/pve-data 795G 197M 795G 1% /var/lib/vz
/dev/sda1 495M 34M 436M 8% /boot
overflow 1.0M 0 1.0M 0% /tmp
/dev/fuse 30M 32K 30M 1% /etc/pve

 

[tom]

you just need to find the files filling up your root file system. also check the log files on /var/log/syslog*.

(you can safely remove old log files to free up some space)

 

[Stryder]

OK I´m back again...So this seems 2 happen because I have a failed backup operation I nano the file /var/log/vzdump/qemu-100.log were I Find this.
Feb 06 18:42:35 INFO: Starting Backup of VM 100 (qemu)
Feb 06 18:42:35 INFO: status = running
Feb 06 18:42:36 INFO: mode failure - unable to dump into snapshot (use option -$
Feb 06 18:42:36 INFO: trying 'suspend' mode instead
Feb 06 18:42:36 INFO: backup mode: suspend
Feb 06 18:42:36 INFO: ionice priority: 7
Feb 06 18:42:36 INFO: suspend vm
Feb 06 18:42:36 INFO: creating archive '/VM-Salvas/dump/vzdump-qemu-100-2013_02$
Feb 06 18:42:36 INFO: adding '/VM-Salvas/dump/vzdump-qemu-100-2013_02_06-18_42_$
Feb 06 18:42:36 INFO: adding '/Node1/images/100/vm-100-disk-1.raw' to archive ($
Feb 06 19:14:45 INFO: lzop: No space left on device: <stdout>
Feb 06 19:14:45 INFO: received signal - terminate process
Feb 06 19:14:46 INFO: unable to open file '/etc/pve/nodes/DUM-VS114/qemu-server$ --------> Here I find an old log file named 100.conf.tmp.24847
Feb 06 19:14:46 INFO: resume vm
Feb 06 19:14:48 ERROR: Backup of VM 100 failed - command '/usr/lib/qemu-server/$

But I can not delete it ....the system tells me "cannot remove permision denied"
I read this article http://forum.proxmox.com/archive/index.php/t-10284.html

Questions..

Erasing this file will help me releasing space on /dev/mapper/pve-root???
Does the new a fresh install of distro 2.3 solve the problem??

 

[Stryder]

OK I just solved my problem I´m really gratefull 2 all the ProXmoX forum friends (PFF) that so fast came in my help specially you tom has been a great Help....so once again proxmox rules

Oh I forgot something I did not understand fully FHS on linux Now I know better.....Thanks 2 all

 

[wingireesh]

OK I just solved my problem I´m really gratefull 2 all the ProXmoX forum friends (PFF) that so fast came in my help specially you tom has been a great Help....so once again proxmox rules

Oh I forgot something I did not understand fully FHS on linux Now I know better.....Thanks 2 all

could you explain the steps , how to fix it ?

 

[udo]

could you explain the steps , how to fix it ?

Hi,
find the used space with tools like ncdu (-x for one filesystem).

If it's not easy to find the big files, there are two possible reasons:
1. you used space inside an directory, where other stuff is now mounted on.
2. an "big" file is deleted, but still in use because one process use the file yet. In this case you don't find the file on the filesystem (lsof is you friend). This can happens if you delete an VM-disk while the VM is running, or the logfile during running syslog.

Udo