Hardisk Full 80GB less than a month?

[yuanyudistira]

Something is wrong with my ut box after re-instal.

I got full hardisk again today

and I find this bis files :

r-------- 1 root root 5.8G 2010-11-15 15:55 /proc/kcore
-rw-r--r-- 1 root root 29G 2010-11-15 01:36 /var/log/untangle-shield/debug.log.1
-rw-r--r-- 1 root root 39G 2010-11-15 11:55 /var/log/untangle-shield/debug.log


Whoa ... what that /var/log/untangle-shield/debug.log ? is all about?

it cost me 69 GB of files ...

Here is the debug.log files containt :

11-15 16:10:13.063650| Reputation[AC14C802,4, 1]: r      10003.7 a 00000000 r       1429.1 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.063717| Reputation[00000000,0,92]: r     0.878934 a 00000000 r      3.90336 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.063738| Reputation[AC14179F,1,81]: r      0.08718 a 00000000 r      4.40557 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.063754| Reputation[AC14179F,2,81]: r      1.18021 a 00000000 r      4.40557 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.063770| Reputation[AC14C801,3, 2]: r   0.00262089 a 00000000 r      21.7664 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.063800| Reputation[AC14C802,4, 1]: r      10004.9 a 00000000 r      1429.26 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.063937| Reputation[00000000,0,92]: r     0.878934 a 00000000 r      3.90336 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.063956| Reputation[AC14179F,1,81]: r      0.08718 a 00000000 r      4.40557 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.063990| Reputation[AC14179F,2,81]: r      1.18021 a 00000000 r      4.40557 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.064008| Reputation[AC14C801,3, 2]: r   0.00262089 a 00000000 r      21.7664 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.064026| Reputation[AC14C802,4, 1]: r      10005.8 a 00000000 r       1429.4 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.064089| Reputation[00000000,0,92]: r     0.878934 a 00000000 r      3.90336 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.064108| Reputation[AC14179F,1,81]: r      0.08718 a 00000000 r      4.40557 s            0 t            0 u 000000000000 e            0
11-15 16:10:13.064123| Reputation[AC14179F,2,81]: r      1.18021 a 00000000 r      4.40557 s            0 t            0 u 000000000000 e            0
11-15 16:10:13

What does it means?


YY

[jcoehoorn]

Note that this command is kind a hard to run on a production server, and can cause load spikes. I found the following command to be a little friendlier to run on loaded systems:

find / -size +10000k -print0 | xargs -0 ls -l

And even here, you still might want to restrict it to a sub folder, but this on my system at least was a lot easier on the RAM.

[jcoehoorn]

Wouldn't ya know it, I try the command again a few minutes later and it doesn't work the same any more. Weird.

[jcoehoorn]

What did end up working for me was the du command:

du -h --exclude=/proc --max-depth=2 /

[dr.gopher]

Good day!

Subscribe to the issue.

How to reclaim disk space and where it (free space) disappears?

How to reclaim disk space and where it (free space) propodaet?

HDD 80 GB


System information








Where are all the remaining space?

Thank you.

[hescominsoon]

NOt to reduce the problems but I ahve several UT and the only time they ever did this is when i was either logging too much..or had a misconfiguration that was causing this. My boxes typcially have 40 gig drives in them...i'm relaly curious to see the outcome of these issues.

[mrunkel]

grrr.

Sometimes I really wonder about the advice on this forum.

this command;

du -h --exclude=/proc --max-depth=2 /

is useless.

just do a

du -sh * 2>/dev/null

When it's done, look at the directories with the high disk usage.

cd into that directory, re-run

du -sh * 2>/dev/null

.

Continue until you find the large files.

If you're database is large, run this command:

psql -U postgres uvm -c "SELECT relname, reltuples, relpages FROM pg_class ORDER BY relpages DESC LIMIT 20 ;"

That will tell you your 20 largest tables.

psql -U postgres uvm -q -t -c "SELECT time_stamp from pl_endp order by time_stamp limit 1;"

Will tell you the oldest record in the database.

You can also run

curl http://www.untangle.com/download/patches/6.0/diskuse.sh | dash

for a complete report.

[dmorris]

curl http://www.untangle.com/download/patches/6.0/diskuse.sh | dash

use it!

[sky-knight]

mrunkel I think your command is off...

du -sh on my box gives me

du: cannot access `./proc/6846/task/19031/fd/1105': No such file or directory
du: cannot access `./proc/6846/task/19031/fd/1106': No such file or directory
du: cannot access `./proc/6846/task/19031/fd/1107': No such file or directory
du: cannot access `./proc/6846/task/19031/fd/1108': No such file or directory
du: cannot access `./proc/6846/task/19031/fd/1109': No such file or directory
du: cannot access `./proc/6846/task/19031/fd/1110': No such file or directory
du: cannot access `./proc/6846/task/19031/fd/1111': No such file or directory
du: cannot access `./proc/6846/task/19031/fd/1112': No such file or directory
du: cannot access `./proc/6846/task/19031/fd/1113': No such file or directory
du: cannot access `./proc/6846/task/19031/fd/1114': No such file or directory
du: cannot access `./proc/6846/task/11775': No such file or directory
du: cannot access `./proc/6846/task/11778': No such file or directory
du: cannot access `./proc/6846/task/11779': No such file or directory
du: cannot access `./proc/6846/task/11780': No such file or directory
du: cannot access `./proc/6846/task/11781': No such file or directory
du: cannot access `./proc/6846/task/11782': No such file or directory
du: cannot access `./proc/6846/task/11783': No such file or directory
du: cannot access `./proc/6846/task/11784': No such file or directory
du: cannot access `./proc/6846/task/11785': No such file or directory
du: cannot access `./proc/6846/task/11786': No such file or directory
du: cannot access `./proc/6846/task/11787': No such file or directory
du: cannot access `./proc/6846/task/11788': No such file or directory
du: cannot access `./proc/6846/task/11789': No such file or directory
du: cannot access `./proc/6846/task/11790': No such file or directory
du: cannot access `./proc/6846/task/11791': No such file or directory
du: cannot access `./proc/11923/task/11923/fd/4': No such file or directory
du: cannot access `./proc/11923/task/11923/fdinfo/4': No such file or directory
du: cannot access `./proc/11923/fd/4': No such file or directory
du: cannot access `./proc/11923/fdinfo/4': No such file or directory
9.0G    .
root@untangle#                                                                               / #                                                                        [root @ untangle]

Not a whole lot of useful information there either.

Now...
du -h --exclude=/proc --max-depth=1 /

That gives you a nice report of space by folder off root. And you can drill from there just by yanking the / off the end of the command.

root@untangle#                                                                               / # du -h --exclude=/proc --max-depth=1 /                                  [root @ untangle]
16K     /lost+found
6.1M    /bin
45M     /boot
4.0K    /cdrom
640K    /dev
20M     /etc
4.0K    /floppy
1.7G    /home
4.0K    /selinux
138M    /lib
36K     /mnt
4.0M    /root
3.5M    /sbin
0       /sys
19M     /tmp
1.7G    /usr
5.4G    /var
4.0K    /dvd
4.0K    /none
5.7M    /opt
9.0G    /
root@untangle#                                                                               / #                                                                        [root @ untangle]

[mrunkel]

mrunkel I think your command is off...

du -sh on my box gives me

Correct, I fixed it.

Now...
du -h --exclude=/proc --max-depth=1 /

That gives you a nice report of space by folder off root. And you can drill from there just by yanking the / off the end of the command.

That doesn't tell you what's in each subfolder, only what is in the folder directly under root. Not helpful if the files are in /var/lib/postgresql.