Mail get queued up due to disk quota problems??

23 posts / 0 new
Last post
#1 Wed, 05/25/2011 - 06:21
-eclipse-

Mail get queued up due to disk quota problems??

Hi All

I am totally Newbie to VirtualMin and Webmin but I am doing the best to search forums and Google before posting. But I have a problem here I can't figure out.

I have installed a CentOS 5.6 server 64 bit and installed VirtualMin with WebMin on a 100GB disk. I have setup 5 sites and they have been working great until now (only one month), when one of our customers told us that mail suddenly doesn't work any more. I looked into the maillog and I can see the following errors

May 25 13:14:16 glolinweb001 postfix/cleanup[15971]: warning: mail_queue_enter: create file incoming/46671.15971: Disk quota exceeded May 25 13:14:26 glolinweb001 postfix/cleanup[15971]: warning: mail_queue_enter: create file incoming/47729.15971: Disk quota exceeded May 25 13:14:36 glolinweb001 postfix/cleanup[15971]: warning: mail_queue_enter: create file incoming/49779.15971: Disk quota exceeded May 25 13:14:46 glolinweb001 postfix/cleanup[15971]: warning: mail_queue_enter: create file incoming/51771.15971: Disk quota exceeded

I can't see any quota errors in the system information page, but somehow there is not enough space to send mails?? We are using a mail relay server which have been setup in the "General Options" under "Postfix Mail Server" in the entry "Send outgoing mail via host".

Looking forward for any help regarding the above error.

Best Regards Tim Nielsen Denmark

Wed, 05/25/2011 - 08:29
andreychek

It sounds like the disk quota for that particular user may have been reached.

You may want to go into Edit Mail and FTP Users for the domain in question to see the quota usage for all the email accounts. From there, you can modify the quota for any users who have reached their limit.

-Eric

Wed, 05/25/2011 - 10:17
-eclipse-

Hi Eric

We don't have mail's setup for the specific sites on the VirtualMin server. Only web and FTP have been activated and when I look in the quota area no one have filled out their quotes at all.

There is no visible quote issue when using the interface.

And no one can send mails anymore, even sites which have 1Tb but only uses 10Mb. So it sounds more to me that it is a temporary area for the postfix program been filled up??

I think it is really strange!?

Looking forward to hear from you.

Tim

Wed, 05/25/2011 - 10:41
andreychek

What output do you receive if you run this command:

repquota /

It looks like something thinks it's over quota, the above command should help determine what exactly that is :-)

-Eric

Wed, 05/25/2011 - 18:23
-eclipse-

Hi Eric

I think so to :/

Here is the output

*** Report for user quotas on device /dev/mapper/VolGroup00-LogVol00
Block grace time: 7days; Inode grace time: 7days
                        Block limits                File limits
User            used    soft    hard  grace    used  soft  hard  grace
----------------------------------------------------------------------
root      +- 3594156 1048576 1048576   none  103909     0     0
daemon    --      20       0       0              3     0     0
lp        --      16       0       0              2     0     0
news      --    7336       0       0            187     0     0
uucp      --     168       0       0              1     0     0
ftp       --       8       0       0              1     0     0
nobody    --      12       0       0              3     0     0
apache    --      32       0       0             19     0     0
rpc       --       4       0       0              1     0     0
smmsp     --       8       0       0              1     0     0
webalizer --      32       0       0              4     0     0
dovecot   --       8       0       0              1     0     0
squid     --      16       0       0              2     0     0
rpcuser   --      36       0       0              5     0     0
named     --      68       0       0              9     0     0
avahi     --      12       0       0              3     0     0
mailman   --     108       0       0             22     0     0
postfix   --    4408       0       0            334     0     0
mysql     --   28320       0       0            616     0     0
postgres  --   29444       0       0            372     0     0
clamav    --   33628       0       0              5     0     0
glolinweb001 --   15096       0       0            912     0     0
vilever   --   46584  512000  512000           3908     0     0
bsl       --  169824  512000  512000          12079     0     0
seatainersagency --   75256  512000  512000          10503     0     0
seatainers --  199496  512000  512000          18995     0     0
seatainersatlantic --   74736  512000  512000          10417     0     0
slangerup-ck --   25272  512000  512000           2016     0     0
sailxtremeshop --  387508  768000  768000          18743     0     0
lydemann  --   57564  512000  512000           4837     0     0
karlslundedagskole --   44968  768000  768000           6320     0     0
infonet   --    7188  512000  512000            125     0     0

I don't know if you can read it though??

Tim

Wed, 05/25/2011 - 18:30
andreychek

Hmm, it looks like a quota is setup for your root user.

That's probably not something you'd want :-)

My suggestion would be to go into Webmin -> System -> Disk Quotas, search for the user "root", and then set root's quota to unlimited.

-Eric

Thu, 05/26/2011 - 07:22
-eclipse-

Strange, I haven't set any quota for the root account, that's kind of strange. And I agree that it is not something I want :)

I found the user and I have changed the quota to unlimited.

I better check our other test environment if it has the same "strange" limitation. If it has, I may have forced the quota to the whole system on a mistake.

I just checked it, and yes same 1Gb quota, I better watch my changes in the future.. :/ Do I have to restart the system to get it working "correctly" again or may I have wrecked it now when there has been a quota enforced on the root user??

Looking forward to hear from you.

Best Regards Tim Nielsen

Thu, 05/26/2011 - 07:26
andreychek

You shouldn't need to restart... try typing:

repquota / | grep root

The 2nd and 3rd column should now show "0" if it's in there correctly.

-Eric

Thu, 05/26/2011 - 08:11
-eclipse-

Hi Eric

Here is the output

root -- 3594252 0 0 103911 0 0

I just checked the /var/log/maillog and it still says that the quote have been reached :/

May 26 15:10:38 glolinweb001 postfix/cleanup[4027]: warning: mail_queue_enter: create file incoming/807995.4027: Disk quota exceeded May 26 15:10:48 glolinweb001 postfix/cleanup[4027]: warning: mail_queue_enter: create file incoming/809823.4027: Disk quota exceeded May 26 15:10:50 glolinweb001 postfix/qmgr[4576]: fatal: qmgr_move: rename 061338981DA from active to incoming: Disk quota exceeded May 26 15:10:51 glolinweb001 postfix/master[3996]: warning: process /usr/libexec/postfix/qmgr pid 4576 exit status 1 May 26 15:10:51 glolinweb001 postfix/master[3996]: warning: /usr/libexec/postfix/qmgr: bad command startup -- throttling May 26 15:10:58 glolinweb001 postfix/cleanup[4027]: warning: mail_queue_enter: create file incoming/811401.4027: Disk quota exceeded

Thx for your help so far :)

Looking forward to hear from you.

Best Regards Tim Nielsen

Thu, 05/26/2011 - 08:48
-eclipse-

Hi Eric

I tried to reboot the server, still the same issue :( Do you think it has crashed the root because of the quote??

Tim

Thu, 05/26/2011 - 10:05
andreychek

What output do you get if you run the command "mount"?

-Eric

Thu, 05/26/2011 - 10:27
-eclipse-

Hi Eric

Here is the output of the command

/dev/mapper/VolGroup00-LogVol00 on / type ext3 (rw,grpquota,usrquota) proc on /proc type proc (rw) sysfs on /sys type sysfs (rw) devpts on /dev/pts type devpts (rw,gid=5,mode=620) /dev/sda1 on /boot type ext3 (rw) tmpfs on /dev/shm type tmpfs (rw) none on /proc/sys/fs/binfmt_misc type binfmt_misc (rw) sunrpc on /var/lib/nfs/rpc_pipefs type rpc_pipefs (rw)

Maybe I need to install a new server and move everything to that one??

Looking forward to hear from you.

Best Regards Tim Nielsen

Thu, 05/26/2011 - 10:44
andreychek

I don't see anything unusual in your mount output.

One thing you may want to try is to go into Limits and Validation -> Check Disk Quotas, and have it run a check.

After that's done, what is the output of "repquota /" again?

-Eric

Thu, 05/26/2011 - 10:56
-eclipse-

Hi Eric

Here is the output

*** Report for user quotas on device /dev/mapper/VolGroup00-LogVol00
Block grace time: 7days; Inode grace time: 7days
                        Block limits                File limits
User            used    soft    hard  grace    used  soft  hard  grace
----------------------------------------------------------------------
root      -- 3595596       0       0         103960     0     0
daemon    --      20       0       0              3     0     0
lp        --      16       0       0              2     0     0
news      --    7336       0       0            187     0     0
uucp      --     168       0       0              1     0     0
ftp       --       8       0       0              1     0     0
nobody    --      12       0       0              3     0     0
apache    --      32       0       0             14     0     0
rpc       --       4       0       0              1     0     0
smmsp     --       8       0       0              1     0     0
webalizer --      32       0       0              4     0     0
dovecot   --       8       0       0              1     0     0
squid     --      16       0       0              2     0     0
rpcuser   --      36       0       0              5     0     0
named     --      68       0       0              9     0     0
avahi     --      12       0       0              3     0     0
mailman   --     112       0       0             22     0     0
postfix   --    4620       0       0            391     0     0
mysql     --   28320       0       0            616     0     0
postgres  --   29444       0       0            372     0     0
clamav    --   33632       0       0              5     0     0
glolinweb001 --   15324       0       0            913     0     0
vilever   --   46548  512000  512000           3899     0     0
bsl       --  170440  512000  512000          12078     0     0
seatainersagency --   75804  512000  512000          10503     0     0
seatainers --  201992  512000  512000          18993     0     0
seatainersatlantic --   76292  512000  512000          10422     0     0
slangerup-ck --   25312  512000  512000           2014     0     0
sailxtremeshop --  386936  768000  768000          18249     0     0
lydemann  --   57620  512000  512000           4854     0     0
karlslundedagskole --   45076  768000  768000           6320     0     0
infonet   --    8616  512000  512000            128     0     0

Best Regards

Tim

Thu, 05/26/2011 - 11:08
-eclipse-

Hi Eric

Hmmm, it just sent 5 mails out here at 5:54 PM DK time somehow, I don't know how... But there is still 406 mails in the queue.

Tim

Thu, 05/26/2011 - 11:25
-eclipse-

Hi Eric

Look at this

May 26 18:21:07 glolinweb001 postfix/qmgr[19952]: fatal: qmgr_move: rename 5769B8981AF from active to incoming: Disk quota exceeded May 26 18:21:08 glolinweb001 postfix/master[3996]: warning: process /usr/libexec/postfix/qmgr pid 19952 exit status 1 May 26 18:21:08 glolinweb001 postfix/master[3996]: warning: /usr/libexec/postfix/qmgr: bad command startup -- throttling May 26 18:22:08 glolinweb001 postfix/qmgr[20002]: fatal: qmgr_move: rename 5769B8981AF from active to incoming: Disk quota exceeded May 26 18:22:09 glolinweb001 postfix/master[3996]: warning: process /usr/libexec/postfix/qmgr pid 20002 exit status 1 May 26 18:22:09 glolinweb001 postfix/master[3996]: warning: /usr/libexec/postfix/qmgr: bad command startup -- throttling May 26 18:22:10 glolinweb001 postfix/postsuper[20024]: warning: bogus file name: incoming/530410.19193 May 26 18:22:29 glolinweb001 postfix/smtpd[20092]: warning: database /etc/postfix/virtual.db is older than source file /etc/postfix/virtual

Tim

Thu, 05/26/2011 - 11:30
andreychek

Yeah, if Postfix thinks it's out of space (or out of quota), those errors may be expected.

I asked Jamie if he could take a peek at the thread here -- let's see if he has any input :-)

-Eric

Thu, 05/26/2011 - 12:31
-eclipse-

Hi Eric

That will be great. :)

We added the Pro license two days ago on both systems, which now have a corrupted root quote. If we at some point is not able to get it working again and we have to reinstall the servers again, what do we then do with the license???

Tim

Thu, 05/26/2011 - 13:17
andreychek

Well, I'd like to think that whatever is wrong is solvable... however, you're certainly welcome to transfer this same Pro license to another server. You'll see a notice about it being installed on two servers until your old server is decommissioned. But, that's no problem, we want you to be able to migrate to a new server when necessary :-)

-Eric

Fri, 05/27/2011 - 04:21
-eclipse-

Hi Eric

Do you know if Jamie had time to look at this thread?

Tim

Fri, 05/27/2011 - 04:23
-eclipse-

Hi Eric

I am planning to do a fresh install as well. Can you give me some tips on how to move the configuration correctly to the new server?

I know where to backup the virtual sites and the configurations for webmin and virtualmin, but any guide lines are welcome :)

Tim

Fri, 05/27/2011 - 10:33
andreychek

We've been discussing it in the Support request you had opened up here:

https://www.virtualmin.com/node/18308

Jamie's last question is when this issue comes up, is it when sending email, receiving email, or both.

Maybe we could pick this up in that support request, then Jamie can follow along a bit easier :-)

-Eric

Fri, 05/27/2011 - 10:34
andreychek

If you're looking to migrate from one server to another -- it's generally as simple as generating a set of Virtualmin backups on one server, then restoring them on the other server.

-Eric

Topic locked