SUGGESTION: Store VM NAME in vzdump backup filename

[gkovacs]

I have an idea for an enhancement of vzdump: when creating a backup job, it would be great to have an option to store the guest's NAME in the backup filename (in addition to the VM ID).

So with the option disabled the filenames would look unchanged:
vzdump-qemu-240-2016_09_02-01_27_32.log
vzdump-qemu-240-2016_09_02-01_27_32.vma.lzo

But with the option enabled they would include the name of the VM as well:
vzdump-qemu-240-debian7mysql-2016_09_02-01_27_32.log
vzdump-qemu-240-debian7mysql-2016_09_02-01_27_32.vma.lzo

This option would help administrators a lot when looking at dump folders containing many vzdump files!

I have also filed a bug for this on the PVE bugzilla:
https://bugzilla.proxmox.com/show_bug.cgi?id=1098

 

[claudiomiguelmuller]

https://bugzilla.proxmox.com/show_bug.cgi?id=438

 

[moxmox]

is this likely to be implimented any time soon?

Also I think it should go further than just backup names it should be the same for the disk image names for containers and vm's.

so instead of just vm-120-disk-0 it would be vm-120-debian7mysql-disk-0 or similar.

looking at the snapshots in "zfs list" or similar or trying to do other admin from the command line is prone to error due to having to lookup the vm id number.

I understand that it would mean you could never change the name of the vm but that is preferable to the current naming convention.

 

[LnxBil]

is this likely to be implimented any time soon?

The bug report is from 2013 and still NEW, so I don't think so.

Also I think it should go further than just backup names it should be the same for the disk image names for containers and vm's.

so instead of just vm-120-disk-0 it would be vm-120-debian7mysql-disk-0 or similar.

looking at the snapshots in "zfs list" or similar or trying to do other admin from the command line is prone to error due to having to lookup the vm id number.

I understand that it would mean you could never change the name of the vm but that is preferable to the current naming convention.

Everything in PVE is VMID centric, so this won't change. It's impossible to change filenames that are in access on the fly, so it will technically not work.

 

[guletz]

Everything in PVE is VMID centric, so this won't change. It's impossible to change filenames that are in access on the fly, so it will technically not work.

Hi to all,

Agree! But like I wrote in a old post, it could be use a TAG(user define) for almost anything in PMX(VM/CT, networks, interfaces, and so on). A VM TAG it could be used in such a case!

Good luck /Bafta

 

[moxmox]

The bug report is from 2013 and still NEW, so I don't think so.



Everything in PVE is VMID centric, so this won't change. It's impossible to change filenames that are in access on the fly, so it will technically not work.

I don't really see why that would be an issue. The ID would stay as is, it could easily parse the filename to get the vm id or disk id and ignore the label - would just be a simple regex...

I think a lot of people would agree with me that current naming convention is pretty dangerous when doing manualy file operations or zfs snapshots etc.

 

[lixaotec]

I agree.

VM name show be on files too. IDs are not helpful within time past.

It should even not have id number on it vm-debian7mysql-disk-0