[one-users] VMs IP Clash after failure

[Valentin Bud]

Hello Cyrille,

I have noticed something this morning about IP addresses management
inside OpenNebula and I thought I'd add my opinion.

As far as I have noticed whenever you instantiate a machine OpenNebula
uses the next IP address from the defined pool in the vnet. That's
almost DHCP. 

One can check the status of the network using `onevnet show <vnet_id>`. 
Under `USED LEASES` one can see what leases are in use. 

If, for whatever reason, one wants a certain ip address to use for a
particular VM, one can use `IP=x.x.x.x` in the template.

I think here arises an issue. Human beings aren't good with numbers,
that's one of the reasons we have DNS. I for one, forget which IP is
allocated to which VM, I am talking about the ones I specify in the
template using `IP`. I don't care which IP is allocated to which VM, I
just care which IPs are allocated or, better said, reserved from the pool.

I know there tools that perform the specific function of IP address
management.

OpenNebula could check if the IP address is in use whenever one
creates/clones a template. The reserved IP address could be listed
with `onevnet show <vnet_id>` under `RESERVED LEASES` or something.

What do you people think about this? 

I also am aware that you guys welcome patches. I wish my Ruby skills
were better so I could code this feature by myself. Speaking of Ruby,
does anyone have some good docs for a noob? :)

Cyrille, I hope I haven't hijacked your thread. I think the above
thoughts are connected with your last question from your email. If
that's not the case I am sorry.

Cheers and Goodwill,
v