v-backup-user: Not enough space on device Topic is solved

[steens]

Hi.

I'm having problems on multiple Debian 8 and 9 servers when running v-backup-user (Or the automatic backups for that matter).
Output

Code: Select all

-- SYSTEM --
2018-01-18 10:21:51 cmo.conf
2018-01-18 10:21:51 stats.log
2018-01-18 10:21:51 history.log
2018-01-18 10:21:51 pam

-- WEB --
2018-01-18 10:21:51 domain.dk

gzip: stdout: No space left on device
2018-01-18 10:22:52 *** 1 domain ***

-- DB --
2018-01-18 10:22:52 cmo_wp (mysql)
grep: write error: No space left on device
Error: dump cmo_wp failed

I don't understand why it says no space left on device, as there is clearly enough.
Disk space / inodes

Code: Select all

root@33454:/usr/local/vesta/func# df -h
Filesystem      Size  Used Avail Use% Mounted on
/dev/simfs       40G  3.4G   37G   9% /
devtmpfs        512M     0  512M   0% /dev
tmpfs           512M     0  512M   0% /dev/shm
tmpfs           512M   14M  499M   3% /run
tmpfs           5.0M     0  5.0M   0% /run/lock
tmpfs           512M     0  512M   0% /sys/fs/cgroup
tmpfs           512M   15M  498M   3% /tmp
none            512M     0  512M   0% /run/shm

root@33454:/usr/local/vesta/func# df -hi
Filesystem     Inodes IUsed IFree IUse% Mounted on
/dev/simfs       977K   62K  916K    7% /
devtmpfs         128K    57  128K    1% /dev
tmpfs            128K     1  128K    1% /dev/shm
tmpfs            128K   222  128K    1% /run
tmpfs            128K     4  128K    1% /run/lock
tmpfs            128K    10  128K    1% /sys/fs/cgroup
tmpfs            128K    34  128K    1% /tmp
none             128K     1  128K    1% /run/shm

Any ideas?

[asdcxz]

Can any one help me, i am also facing the same issue.

i am having 3 domains on server 2 domains are working fine with backup but the thrid is getting error

[mehargags]

check if inodes full, post results of

Code: Select all

df -i

[asdcxz]

Code: Select all

Filesystem            Inodes  IUsed    IFree IUse% Mounted on
/dev/mapper/vg-root 60448768 109306 60339462    1% /
devtmpfs              987368    374   986994    1% /dev
tmpfs                 990121      1   990120    1% /dev/shm
tmpfs                 990121    450   989671    1% /run
tmpfs                 990121     16   990105    1% /sys/fs/cgroup
/dev/sda2             128016    343   127673    1% /boot
/dev/mapper/vg-tmp     65536     77    65459    1% /tmp
tmpfs                 990121      1   990120    1% /run/user/0

[steens]

check if inodes full, post results of

Code: Select all

df -i

I already posted this, they are not used.

[dsystem]

I have the same problem, 5 accounts can not back up two.

I use:
CentOS 7.4.1708

VestaCP 0.9.8-18

[mehargags]

Can you paste Vesta logs here ?

[alexcy]

Same issue here. Ubuntu 16.04

[steens]

Can you paste Vesta logs here ?

error.log

Code: Select all

2018-01-21 05:11:28 v-backup-user  'cmo' [Error 17]
2018-01-22 05:15:09 v-backup-user  'cmo' [Error 17]
2018-01-23 05:13:22 v-backup-user  'cmo' [Error 17]

backup.log

Code: Select all

================================
admin
--------------------------------

-- SYSTEM --
2018-01-23 05:10:02 admin.conf
2018-01-23 05:10:02 stats.log
2018-01-23 05:10:02 history.log
2018-01-23 05:10:02 pam


-- WEB --
2018-01-23 05:10:02
2018-01-23 05:10:02 *** 1 domain ***

-- DB --
2018-01-23 05:10:02 admin_default (mysql)
2018-01-23 05:10:02 *** 1 database ***

-- CRON --
2018-01-23 05:10:02 *** 8 jobs ***

-- User Dir --
2018-01-23 05:10:03 adding .bash_logout
2018-01-23 05:10:03 adding .bashrc
2018-01-23 05:10:03 adding .profile
2018-01-23 05:10:03 adding tmp
2018-01-23 05:10:03 *** 4 directories ***

-- SUMMARY --
2018-01-23 05:10:03 Rotated: 2018-01-19_05-10-02
2018-01-23 05:10:03 Rotated: 2018-01-20_05-10-03
2018-01-23 05:10:03 Local: /backup/admin.2018-01-23_05-10-03.tar
2018-01-23 05:10:03 Size: 1 MB
2018-01-23 05:10:03 Runtime: 1 minute

--------------------------------

================================
cmo
--------------------------------

-- SYSTEM --
2018-01-23 05:10:04 cmo.conf
2018-01-23 05:10:04 stats.log
2018-01-23 05:10:04 history.log
2018-01-23 05:10:04 pam


-- WEB --
2018-01-23 05:10:04 domain.dk

gzip: stdout: No space left on device
2018-01-23 05:13:22 *** 1 domain ***

-- DB --
2018-01-23 05:13:22 cmo_wp (mysql)
grep: write error: No space left on device
Error: dump cmo_wp failed

--------------------------------

system.log is a bunch of additions/deletions to the firewall on port 22 (from fail2ban, I assume). It does say v-backup-user admin during the backup cron, but does not mention backing up the cmo user at all.
I can't really find any clues as to what is wrong.

Also, as others mention, it's interesting that this is only a problem with some users. Is it maybe packing up the same things too many times requiring a lot of disk space?
Since the backup structure is basically tarballs in tarballs in tarballs, if these archives are not deleted once the higher level tar is created, that could require a lot of space I suppose.

As far as I can tell from the code in v-backup-user, the /tmp folder is used, which in my case cannot be bigger than 512 MB. That seems to be the issue, as changing it to use the home folder makes it work.
Do you know of any solution that does not require hard patching this?

[alexcy]

It used to be /backup, why change that? Many people mount their /tmp as tmpfs as well.