The volume "filesystem root" has only X bytes disk space remaining

[cmygeHm]

Hello!
Please help me.
Throws this error message: "The volume 'filesystem root' has only X bytes disk space remaining". Where X near 1 Gb. Capacity on the first device is 50Gb. Really used near 20Gb.

root@gafarov-pc:/usr/bin# du -hx --max-depth=1 / | sort -h

du: cannot access `/home/me/.gvfs': Permission denied
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
176K /root
792K /tmp
8.8M /bin
9.2M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.8G /home
3.3G /usr
12G /var
17G /

It seems, trouble in ~/.gvfs, I don't know what is it.
What another way to understand what reason of trouble?
OS Ubuntu 11.10

[spjackson]

Don't worry about .gvfs - that is normal. Please post the output of these commands.


df -h
df -i

[cmygeHm]

Don't worry about .gvfs - that is normal. Please post the output of these commands.


df -h
df -i




root@gafarov-pc:/home/me# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/sda1 49G 17G 29G 37% /
udev 1.5G 4.0K 1.5G 1% /dev
tmpfs 604M 836K 603M 1% /run
none 5.0M 0 5.0M 0% /run/lock
none 1.5G 9.7M 1.5G 1% /run/shm
/dev/sda3 406G 10G 375G 3% /home/me/Workspace

root@gafarov-pc:/home/me# df -i
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/sda1 3203072 231818 2971254 8% /
udev 384211 501 383710 1% /dev
tmpfs 386190 416 385774 1% /run
none 386190 1 386189 1% /run/lock
none 386190 19 386171 1% /run/shm
/dev/sda3 27009024 143 27008881 1% /home/me/Workspace

[spjackson]

Well, clearly the filesystem is far from full. Do you get this message now or does it come up intermittently? It's possible that something is creating very large temporary files that are eventually removed.

12GB for /var seems pretty high. The most common things to use a lot of disk space in /var are a) installing and updating packages in /var/cache b) excessive logging in /var/log. Less commonly, /var/spool, /var/mail, /var/tmp. Additionally, if the machine is a server, there may be websites and/or databases in /var.



du -hx --max-depth=1 /var | sort -h
However, the current content of /var does not explain the message you are getting.

[cmygeHm]

This message does come up intermittently, 2-3 two times in day.
It is very large postgresql db files. I think it is normal. I can't understand what big files (30Gb) can be temporary created?


4.0K /var/crash
4.0K /var/games
4.0K /var/local
4.0K /var/mail
4.0K /var/opt
96K /var/spool
9.3M /var/backups
11M /var/tmp
19M /var/log
575M /var/cache
754M /var/www
16G /var/lib
17G /var

It is more then 12Gb here bcz I restored dump. When I wrote first message it was 12GB. New db restored right now is 16Gb.

[cmygeHm]

Right now:


root@gafarov-pc:/var/www/h/Service/Holder/Scripts/Vkontakte# du -hx --max-depth=1 / |sort -h
du: cannot access `/home/me/.gvfs': Permission denied
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
144K /tmp
180K /root
8.8M /bin
9.1M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.7G /home
3.3G /usr
17G /var
23G /

[cmygeHm]

Anybody can help me? Please...

[spjackson]

You need to see whether the disk is actually filling up, or whether the message is incorrect. I think it unlikely that the message is incorrect, but it's always possible.

Whenever you have posted, the disk has been nowhere near full. You need to do your df and du commands when the message appears. If these don't catch it, then perhaps have a cron job to log df and du to a file once a minute. Something like this:

sudo crontab -e


* * * * * (date ; df /) >> /var/log/df.log
* * * * * (date ; du -hx --max-depth=1 / | sort -h) >> /var/log/du1.log
* * * * * (date ; du -hx --max-depth=1 /var | sort -h) >> /var/log/du2.log
Since you are running a database installation of several gigabytes, my money would be on an errant query - probably an unintended cartesian product. Is there any relevant message in the postrgesql log? (probably in /var/log/postgresql/ )

[cmygeHm]

ok, I will set jobs in cron.
So big database bcz I develop one program and use outside database. It db really very big.

[cmygeHm]

It is strange:


13G /var
18G /
Mon Aug 20 13:50:01 MSK 2012
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
136K /tmp
184K /root
8.8M /bin
9.1M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.7G /home
3.3G /usr
13G /var
18G /
Mon Aug 20 13:55:01 MSK 2012
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
136K /tmp
184K /root
8.8M /bin
9.1M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.7G /home
3.3G /usr
13G /var
18G /
Mon Aug 20 14:00:02 MSK 2012
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
136K /tmp
184K /root
8.8M /bin
9.1M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.7G /home
3.3G /usr
13G /var
18G /
Mon Aug 20 14:05:02 MSK 2012
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
136K /tmp
184K /root
8.8M /bin
9.1M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.7G /home
3.3G /usr
13G /var
18G /
Mon Aug 20 14:10:01 MSK 2012
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
136K /tmp
184K /root
8.8M /bin
9.1M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.7G /home
3.3G /usr
13G /var
18G /
Mon Aug 20 14:15:02 MSK 2012
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
136K /tmp
184K /root
8.8M /bin
9.1M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.7G /home
3.3G /usr
13G /var
18G /
Mon Aug 20 14:20:01 MSK 2012
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
136K /tmp
184K /root
8.8M /bin
9.1M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.7G /home
3.3G /usr
13G /var
18G /


AND in this time:


/dev/sda1 50395844 32932236 14903608 69% /
Mon Aug 20 13:46:01 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 33215992 14619852 70% /
Mon Aug 20 13:47:01 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 33551632 14284212 71% /
Mon Aug 20 13:48:01 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 33853420 13982424 71% /
Mon Aug 20 13:50:01 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 34395428 13440416 72% /
Mon Aug 20 13:55:01 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 36052812 11783032 76% /
Mon Aug 20 14:00:02 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 37764636 10071208 79% /
Mon Aug 20 14:05:02 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 39437880 8397964 83% /
Mon Aug 20 14:10:01 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 41111448 6724396 86% /
Mon Aug 20 14:15:02 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 42781360 5054484 90% /
Mon Aug 20 14:20:01 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 44438256 3397588 93% /
Mon Aug 20 14:25:02 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 46112620 1723224 97% /

I restared PC and it is again


Mon Aug 20 14:35:01 MSK 2012
Filesystem 1K-blocks Used Available Use% Mounted on
/dev/sda1 50395844 18579648 29256196 39% /

[spjackson]

There are several reasons why df can show much higher values than du, but we can rule out some of these possibilities because the problem goes away when you reboot, and then the problem repeats.

This usually means that a file is being written to that has been deleted. Try monitoring as follows.

sudo crontab -e


* * * * (date ; lsof | egrep 'COMMAND|deleted') >> /var/log/lsof.log
If you find a process that is doing this, especially if SIZE/OFF column is large, then note the PID from the first column and do


ps -F pid_number
substituting the PID for pid_number.

Did you look at the postrgesql log?

A query such as this which does an unintended cartesian product will result in a huge temporary file if there are many rows in the table.


select a.* from mytable a, mytable b

[GeorgeWhite5]

Hello!
Please help me.
Throws this error message: "The volume 'filesystem root' has only X bytes disk space remaining". Where X near 1 Gb. Capacity on the first device is 50Gb. Really used near 20Gb.

root@gafarov-pc:/usr/bin# du -hx --max-depth=1 / | sort -h

du: cannot access `/home/me/.gvfs': Permission denied
0 /dev
0 /proc
0 /run
0 /sys
4.0K /cdrom
4.0K /lib64
4.0K /mnt
4.0K /selinux
4.0K /srv
8.0K /media
16K /lost+found
176K /root
792K /tmp
8.8M /bin
9.2M /sbin
15M /etc
15M /lib32
24M /boot
125M /opt
224M /lib
1.8G /home
3.3G /usr
12G /var
17G /

It seems, trouble in ~/.gvfs, I don't know what is it.
What another way to understand what reason of trouble?
OS Ubuntu 11.10

How strange... how old is the harddrive in question?

EDIT: You aren't connecting to remote servers (using Nautilus or another gnome program), are you? If you don't, then it is nothing to do with .gvfs

[GeorgeWhite5]

This may help:

You can change where PostregreSQL stores its temporary files (through a variable). I found it here:

http://archives.postgresql.org/pgsql-general/2008-10/msg01406.php

You could possibly change this (or get the variable)?

[cmygeHm]

spjackson, thank you very much!
Excuse me, please =)
One day I play with mplayer and cron =)
I past task for mplayer to start play in on time. And I found


cron 13577 root 5u REG 8,1 6284358798 786599 /tmp/tmpfUQsuAA (deleted)
sh 13578 me 1u REG 8,1 6284360256 786599 /tmp/tmpfUQsuAA (deleted)
sh 13578 me 2u REG 8,1 6284360418 786599 /tmp/tmpfUQsuAA (deleted)
mplayer 13579 me 1u REG 8,1 6284373783 786599 /tmp/tmpfUQsuAA (deleted)
mplayer 13579 me 2u REG 8,1 6284373864 786599 /tmp/tmpfUQsuAA (deleted)

when tried
lsof | egrep 'COMMAND|deleted'
I killed this process and space on hard disk released =)
Thank you very much =)
Of course I removed task for mplayer from crontab.

[spjackson]

I killed this process and space on hard disk released =)
Thank you very much =)
Of course I removed task for mplayer from crontab.
Great. I'm glad you've managed to solve it. Could you please mark the thread as solved?

Also, make sure to remove all the logging you added to crontab, or that will fill your disk eventually!

[cmygeHm]

I marked it.
Thank you =) Ok, I have only one task in crontab, which creates so big logs.