Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

mailx says no space left, but there's plenty of it

1,314 views
Skip to first unread message

LostInTheLoop

unread,
Aug 18, 2010, 6:17:38 AM8/18/10
to
Hi all,

Slackware 13.0.

root@box:~# echo test | mailx -s test -r ema...@domain1.it
email2@domain2

It says

temporary mail file: No space left on device
No space left on device
. . . message not sent.


This morning I found a big bunch of old messages generated by
fail2ban in /var/spool/clientmqueue/. I deleted them all and
adjusted fail2ban settings. Now the directory is empty and I get
these output:

root@box:~# df -h
Filesystem Size Used Avail Use% Mounted on
/dev/root 749M 627M 85M 89% /
/dev/sda3 183M 85M 89M 49% /var
/dev/sda4 2.8G 705M 2.0G 27% /home
tmpfs 27M 0 27M 0% /dev/shm


root@box:~# mount
/dev/root on / type ext4 (rw,barrier=1,data=ordered)
proc on /proc type proc (rw)
sysfs on /sys type sysfs (rw)
usbfs on /proc/bus/usb type usbfs (rw)
/dev/sda3 on /var type ext4 (rw)
/dev/sda4 on /home type ext4 (rw)
tmpfs on /dev/shm type tmpfs (rw)

Moreover I can create an delete files in all of the subdirectories
of /var/spool/, whith touch /var/spool/test and rm...

What's wrong? Thanks.

Henrik Carlqvist

unread,
Aug 18, 2010, 3:31:50 PM8/18/10
to
LostInTheLoop <bum...@ro.und> wrote:
> What's wrong?

If the error message from an application is not enough to tell what is
wrong strace might be able to give a clue about what is going on:

root@box:~# echo test | strace -f mailx -s test

regards Henrik
--
The address in the header is only to prevent spam. My real address is:
hc123(at)poolhem.se Examples of addresses which go to spammers:
root@localhost postmaster@localhost

LostInTheLoop

unread,
Aug 19, 2010, 10:18:52 AM8/19/10
to
Henrik Carlqvist, on 08/18/2010 09:31 PM, wrote:
> LostInTheLoop <bum...@ro.und> wrote:
>> What's wrong?
>
> If the error message from an application is not enough to tell
> what is wrong strace might be able to give a clue about what is
> going on:
>
> root@box:~# echo test | strace -f mailx -s test
>
> regards Henrik

I' think there's some deep problem with the disk, a compactflash
card actually. Using vi it keeps saying that it can't write on
.viminfo. But I can delete .viminfo.
I've installed strace with slapt-get, an yet there is no
/usr/bin/strace, as it says in the package's file list.
I think I'll just reformat it, checking for bad blocks.

Lew Pitcher

unread,
Aug 19, 2010, 10:58:24 AM8/19/10
to
On August 18, 2010 06:17, in alt.os.linux.slackware, bum...@ro.und wrote:

> Hi all,
>
> Slackware 13.0.
>
> root@box:~# echo test | mailx -s test -r ema...@domain1.it
> email2@domain2
>
> It says
>
> temporary mail file: No space left on device
> No space left on device
> . . . message not sent.
>
> This morning I found a big bunch of old messages generated by
> fail2ban in /var/spool/clientmqueue/. I deleted them all and
> adjusted fail2ban settings. Now the directory is empty and I get
> these output:

[snip]
> What's wrong? Thanks.

Given that you seem to have lots of space available on your filesystems, and
your subsequent posts hint that you manipulated directories at some point,
it /could be/ that the permissions on your spool directories have got
screwed up a bit.

Look to see if the /var/spool/ mail directories have the proper permissions.
If the target directory is read-only, then apps won't be able to
write/append files to it. Which might explain the "No space" error.

HTH
--
Lew Pitcher
Master Codewright & JOAT-in-training | Registered Linux User #112576
Me: http://pitcher.digitalfreehold.ca/ | Just Linux: http://justlinux.ca/
---------- Slackware - Because I know what I'm doing. ------


Sylvain Robitaille

unread,
Aug 19, 2010, 11:27:47 AM8/19/10
to
Lew Pitcher wrote:

> If the target directory is read-only, then apps won't be able to
> write/append files to it. Which might explain the "No space" error.

I'm quite certain that would return EACCESS, resulting in a "permission
denied" error, not "no space left on device" (ENOSPC).

--
----------------------------------------------------------------------
Sylvain Robitaille s...@encs.concordia.ca

Systems analyst / AITS Concordia University
Faculty of Engineering and Computer Science Montreal, Quebec, Canada
----------------------------------------------------------------------

LostInTheLoop

unread,
Aug 19, 2010, 11:45:26 AM8/19/10
to
Sylvain Robitaille, on 08/19/2010 05:27 PM, wrote:
> Lew Pitcher wrote:
>
>> If the target directory is read-only, then apps won't be able
>> to write/append files to it. Which might explain the "No space"
>> error.
>
> I'm quite certain that would return EACCESS, resulting in a
> "permission denied" error, not "no space left on device"
> (ENOSPC).
>

Permissions seem to be ok, but I'm starting to get errors on varius
applications. Vim for example says that can't write on .viminfo, but
I can rm it. I think that the compactflash is dying on me. Is there
anything I can do to check the disk without unmounting the root
partition? The disk is a compactflash card on /dev/sda
Thanks everybody.

Helmut Hullen

unread,
Aug 19, 2010, 11:44:00 AM8/19/10
to
Hallo, LostInTheLoop,

Du meintest am 18.08.10:

> root@box:~# echo test | mailx -s test -r ema...@domain1.it
> email2@domain2

> It says

> temporary mail file: No space left on device
> No space left on device
> . . . message not sent.

What tells

which -a mailx
ls -l $(which mailx)
file $(which mailx)

Viele Gruesse
Helmut

"Ubuntu" - an African word, meaning "Slackware is too hard for me".

LostInTheLoop

unread,
Aug 21, 2010, 2:36:20 PM8/21/10
to
Helmut Hullen, on 08/19/2010 05:44 PM, wrote:

> What tells
>
> which -a mailx
> ls -l $(which mailx)
> file $(which mailx)

root@box:~# which -a mailx
/usr/bin/mailx
root@box:~# ls -l $(which mailx)
-rwxr-xr-x 1 root root 306856 Aug 11 2009 /usr/bin/mailx*
root@box:~# file $(which mailx)
/usr/bin/mailx: ELF 32-bit LSB executable, Intel 80386, version 1
(SYSV), dynamically linked (uses shared libs), stripped

Grant

unread,
Aug 21, 2010, 5:12:46 PM8/21/10
to

You didn't run out of inodes? That's way to 'fill' a non-full disk.

Though it seems a rare complaint in recent years. If you run out
of inodes there's nowhere to catalog (directory area) where new
information gets stored in the filesystem.

I didn't follow this thread from start.

Grant.

LostInTheLoop

unread,
Aug 22, 2010, 7:41:48 AM8/22/10
to
Grant, on 08/21/2010 11:12 PM, wrote:

> You didn't run out of inodes? That's way to 'fill' a non-full
> disk.

It would be quite a weird occurrence, the whole root is a 700MB
partition. I don't think I have so many files in it. How can I check
for the inodes shortage?

Thanks

Helmut Hullen

unread,
Aug 22, 2010, 8:43:00 AM8/22/10
to
Hallo, LostInTheLoop,

Du meintest am 22.08.10:

>> You didn't run out of inodes? That's way to 'fill' a non-full
>> disk.

> It would be quite a weird occurrence, the whole root is a 700MB
> partition. I don't think I have so many files in it. How can I check
> for the inodes shortage?

df -i

LostInTheLoop

unread,
Aug 23, 2010, 6:37:07 AM8/23/10
to
Helmut Hullen, on 08/22/2010 02:43 PM, wrote:

> df -i


Ouch! I did run out of inodes!!
Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/root 48768 48767 1 100% /

Removed a few packages, now

Filesystem Inodes IUsed IFree IUse% Mounted on
/dev/root 48768 48187 581 99% /

Now mailx and everything else work fine

Thanks all!

Eef Hartman

unread,
Sep 6, 2010, 8:10:05 AM9/6/10
to
LostInTheLoop <bum...@ro.und> wrote:
> It would be quite a weird occurrence, the whole root is a 700MB
> partition. I don't think I have so many files in it. How can I check
> for the inodes shortage?

df -i
gives the same kind of overview for I-node usage.

I recently had that problem when I had initialized a data partition
with "-T largefile", but the user then ran out of I-nodes on it
(largefile allocates 1 I-node per MB, the default is 1 per 8 KB)
There even is a type "largefile4" which only allocates one I-node
for each 4 MB of diskspace.

PS: this is all for ext2/3, other fs'es do it anothe way.
--
******************************************************************
** Eef Hartman, Delft University of Technology, dept. SSC/ICT **
** e-mail: E.J.M....@tudelft.nl - phone: +31-15-27 82525 **
******************************************************************

0 new messages