Quota problem


My server loses quota settings randomly. I think it's related to some schedule used by virtualmin or by server reboot.

Closed (fixed)


Howdy -- well, that's not normally an issue that comes up. Is there anything unusual about your setup? Are you using Amazon's EC2 service, for example?

Also, what does your /etc/fstab file look like?

Whenever your quota stops working, how do you typically go about re-enabling it?

Hi Andrey

It a Dell Poweredge 2900 server running CentoOS 5.x nothing virtual.

I did not make it clear, but the quota does not disapear it just changes back to default values. 200 MB. I can then manually set it using the shell or via virtualmin edit virtual server. It then works a while again. My best guess is a couple of days.

/dev/VolGroup00/LogVol00 / ext3 grpquota,usrquota,rw 0 1 LABEL=/boot /boot ext3 defaults 1 2 devpts /dev/pts devpts gid=5,mode=620 0 0 tmpfs /dev/shm tmpfs defaults 0 0 proc /proc proc defaults 0 0 sysfs /sys sysfs defaults 0 0 /dev/VolGroup00/LogVol01 swap swap defaults 0 0

Best Regards Sebastian

I've seen this sometimes with Virtualmin running under an OpenVZ instance - are you using that there perhaps?

No, it's a normal server with no virtualization. No Xen, VMware, Hyper-V, OpenVZ or anything. :o)

Is the aquota.user or aquota.group file in the filesystem's root directory perhaps being restored from a backup or copied from some other location? That is the file that contains user and group quota information..

Both files are there. Modified a few minutes ago by virtualmin I guess. I'm the only one with shell access and I did not apply any quota.

I have a site/user with quota problems. The user is currently ok. even though the quota files were just changed.

[root@sekhmet /]# ls -lt total 660 -rw------- 1 root root 90112 Dec 16 10:40 aquota.group -rw------- 1 root root 91136 Dec 16 10:40 aquota.user

Are there perhaps aquota.user.new and aquota.group.new files in the same directory? I've seen issues where old files with those names confuse the quota system ..

No, theres only those 2 quota files in the root directory.

If you make a quota change with Virtualmin and then display it with a command like :

quota -v -u username

is it shown correctly?

Yes, that works.

[root@sekhmet ~]# quota -v -u mmd3semautumn2009.sub.net Disk quotas for user mmd3semautumn2009.sub.net (uid 979): Filesystem blocks quota limit grace files quota limit grace /dev/mapper/VolGroup00-LogVol00 1048576* 1048576 1048576 1272 0 0

[root@sekhmet ~]# quota -v -u mmd3semautumn2009.sub.net Disk quotas for user mmd3semautumn2009.sub.net (uid 979): Filesystem blocks quota limit grace files quota limit grace /dev/mapper/VolGroup00-LogVol00 1048576 2097152 2097152 1272 0 0

Ok, so clearly the quotas are being updated correctly by Virtualmin ..

Is it a reboot that causes them to get reset?

After a reset, does that quota command show the old values?

I do not have anything new about this issue. I do not think it has been fixed.

I'm still interested to know what triggers the quota reset .. does a reboot cause it?


I've the same issue and discovered the quotas daemon is stopped by a script (don't figured out why) /etc/webmin/quota/email.pl which is launched each 5 Minutes.

Hope it can help.

Quotas daemon? I don't think there is any such thing .. what process are you referring to exactly?

No sorry by quota daemon i mean the answer of the quotaon command:

quotaon -ap

Which send "disabled" after each cron execution.

root@XXXXXXXX:/# quotaon -ap

group quota on / (/dev/disk/by-uuid/a4d06b3d-a269-409a-9790-421010b401e4) is disabled user quota on / (/dev/disk/by-uuid/a4d06b3d-a269-409a-9790-421010b401e4) is disabled

(translated from french)

Interesting, it looks like quotaon -p outputs a different format on your system to what Virtualmin expects, which is causing it to get confused :-( This will be fixed in the next Webmin release (1.540).

Which linux distribution and version are you running there?

I'm running on Linux Ubuntu 10.10 (I Know it's not supported with virtualmin but Webmin support it, doesn't it ?)

As it is a Webmin issue...

Yes, it is a Webmin bug .. Ubuntu 10.10 isn't nearly as well tested with Webmin as 10.04.

Sorry for hijacking this issue, but I had a similar issue on Centos 5.5 (OpenVZ kernel). See also


I suspected it something to with reboots, but I could not reproduce it consistently. It hasn't happened again in the last two weeks fortunately.

No response required. This is just FYI

Automatically closed -- issue fixed for 2 weeks with no activity.


I just installed VirtualMin 1,534 on Debian 6,0,1 x64 and I also have a "quota display issue". Indeed, when I click on "Edit Virtual Server", under "Quotas and limits", I get the following:

Home disk space used (including sub-servers) -8 kB (7.61 GB by server administrator, 0 bytes by mail / FTP users)

It seems that I have the same error as this user: http://www.virtualmin.com/node/16805 "Looks like it doesn't make the sum between the mail /FTP users quote and the server administrator's one and juste remove the quote of the database."

I printed some debug information from the function "show_domain_quota_usage" of file "./virtual-server/virtual-server-lib-funcs.pl". So here are they (this virtual server just has files in its home directory):

$homequota: 0
$mailquota: 0
$duser: HASH(0x438d870)
$dbquota: 8192
$dbquota_home: 8192
$subhomequota: 0
$submailquota: 0
$subdbquota: 0
$totalhomequota: 0
$bsize: 1024

After line "$totalhomequota -= $dbquota_home/$bsize;", the concerned variable becomes:

$totalhomequota: -8

If it can help you, here is the output of "quotaon -ap" on my Debian 6,0,1 x64 system:

group quota on / (/dev/disk/by-uuid/4462c922-0b8d-45a2-b682-0f38ac3eef86) is on
user quota on / (/dev/disk/by-uuid/4462c922-0b8d-45a2-b682-0f38ac3eef86) is on


Tristan - could you open a new ticket for this? Your issue sounds quite different to the original bug report..

Same issue here running Debian Linux 6.0, Virtualmin 3.86.gpl GPL, Webmin 1.550 when executing /etc/webmin/quota/email.pl. Is this cronjob virtualmin or webmin related? Also I don't get any quota exceed email warnings. As it seems the job doesn't work properly at all.

I disabled it for now but I really need to have quota warnings about exceeding file limits. I hope this issue will be fixed soon...

I just came across this and I believe it was fixed back in 2012/2013 => DONE