Failed to get whole disk by device number

[romain1985]

Hello,
I installed Veeam agent 1.0.0.499 (32 bits, i386) to my personal OpenVZ (vps), Debian 7 .

I press [C] to create my first backup job and I get this error message :
No such file or directory
Failed to get whole disk by device number [243:136]

Do you have an idea to resolve this issue ?

Regards,
Romain

[nielsengelen]

This is related to OpenVZ way of creating the containers. Certain folders are linked to the host OS such as /bin, /usr/bin,... and we depend on certain things (I guess in this case /dev) which can't be read.

[romain1985]

Hello,
That mean it is impossible to use Veeam to backup OpenVZ container ?

Thanks.

[nielsengelen]

Additional testing will be needed to confirm if it will work or not. Thanks for the feedback already!

[hawkeye80]

Got a similar error on initial configuration.

Code: Select all

Failed to get whole disk by device number [2:12].

Test system is a Oracle Linux 6 VM with UEK kernel and LVM configured.

[PTide]

Hi,

Please post a download link to an archive with logs. Also please post the outputs of dmesg -T and lsblk.

Thak you

[hawkeye80]

Hi PTide,

please find the logs and outputs here https://file.io/01WACh
dmesg -T is not available on OEL6, so I've cleared the dmesg, rebooted and tried veeam again. So the entries are all fresh.

Kind regards,
Daniel

[pguima]

Also receiving similar error CentOS 7 running on openVZ.

[PTide]

Hi,

Also receiving similar error CentOS 7 running on openVZ.

Does it fail when you try to backup the openVZ host, or it fails if you start veeam inside the container? If the former then please provide a detailed description of the job settings, host configuration. Please include the exact error message, dmesg -T output and lsblk -af output, also please share an archive that contains /var/log/veeam directory.

Thank you.

[pguima]

Unfortunately, my dmesg log are blank and the output for the command are null. Please see the veeam logs, here https://file.io/JpHnFY

[PTide]

<...>and the output for the command are null

What does lsblk -af shows? Also I see /dev/simfs in the log, are you trying to run VAL inside an openVZ container?

Thanks

[pguima]

Mostly likely, I have running on a hosted VPS, I installed the release version but still have issues. This is what I get now
boost::filesystem::directory_iterator::construct: No such file or directory: "/sys/block"

[PTide]

Hi,

VAL is not designed to operate inside a container. Anyway please provide logs (use 'M' key on the main screen). Also could tell who is your hoster, please?

Thanks

[vchrizz]

hi,

im getting the same error when using the linux agent within an openvz container.
the mountpoint /sys is mounted and available, but /sys/block is not available due to the nature of (openvz-) containers.
so i guess the agent tries to access physical disks which are not available.

are there any plans for the linux agent to also back up containers in another way than accessing the physical disk(s)?

thanks

[PTide]

Hi,

No, currently we don't have that functionality on the list due to a low demand. Anyway your feedback is noted, thank you!

[Daniel N.]

Hi,

No, currently we don't have that functionality on the list due to a low demand. Anyway your feedback is noted, thank you!

Please increase the demand counter by one.
I run a VPS by www.hosteurope.de and also see the error message

Code: Select all

boost::filesystem::directory_iterator::construct: No such file or directory: "/sys/block/ploop23837p1/slaves"

[willinit]

Would love to see this feature as well. Currently have about 30+ OpenVZ servers and being able to set up jobs for specific container folders would be a huge help.

To chime in on the error front, what I get when trying to add a backup job is the following error (CTID is in place of the actual container ID). But the thing is, this isn't an NFS or SMB mounted partition.
"Network filesystems (NFS, SMB) cannot be backed up. Please exclude: /vz/root/CTID/sys."

Another container gave the following error:
"Network filesystems (NFS, SMB) cannot be backed up. Please exclude: /vz/root/CTID/proc."

[efranklin]

I also love to have this feature as well. Encountered this trying to backup a VPS at Media Temple.

[0riole]

Hi All,

Has there been any progress on this? I have two VMs in an OpenVZ hosted environment, with all care and no responsibility, it would be greatly beneficial if I could get VAL working on these two VMs.

Thanks in advance,
Chris

[PTide]

Hi,

Unfortunately, the upcoming version of VAL will not support containers

Thanks