Issue with Gen2 Compute profiles

[ksagle77]

Hey,

I'm running in to an issue when creating a Gen2 compute profile. I am able to create the a gen 2 profile:

but then once I try to open the profile, I get the following error:

Oops, we're sorry but something went wrong Hyper-V was unable to find a virtual machine with name "foreman_1511888691".

The name 'foreman_1511888691' is static, it does not change based on the name of the profile/host/etc.

I looked for that string in the code, but was unable to find it. Stack trace is attached.

stack_trace_compute.txt

Kevin

[ksagle77]

Note, I am able to create a gen 2 vm successfully using fog hyper-v directly.

vm = compute.new_vm(
name: 'fogtest',
generation: 2,
path: 'd:\newvms\fogtest',
bootdevice: 'VHD',
vhdpath: 'd:\newvms\fogtest\fogtest.vhdx'
)

[ksagle77]

Interestingly, if I try to create a gen 2 VM, the same error is thrown during the first step of the process.

The VM is created anyways, but the rest of the deployment steps do not complete.

[ananace]

Worth noting is that the 'foreman_1511888691' is from Foreman itself, when creating a VM object Foreman gives it an unique identifier until the Compute Resource can provide the actual one.

I think I have an idea what the issue might be, which version of fog_hyperv are you running it against?

[ksagle77]

Ah, cool.

-bash-4.2$ gem list | grep fog-hyperv
fog-hyperv (0.0.5)

[ananace]

Then that's not the issue I'm thinking of.

Can you give me the Foreman version as well, and I'll throw up an environment for that when I'm back at my workstation.

[ksagle77]

rpm -qa 'foreman' || dpkg -l 'foreman' | sort

foreman-1.15.6-1.el7.noarch
tfm-rubygem-hammer_cli_foreman-0.10.2-1.el7.noarch
foreman-debug-1.15.6-1.el7.noarch
foreman-proxy-1.15.6-1.el7.noarch
foreman-cli-1.15.6-1.el7.noarch
foreman-installer-1.15.6-2.el7.noarch
foreman-selinux-1.15.6-1.el7.noarch
foreman-release-scl-3-1.el7.noarch
foreman-postgresql-1.15.6-1.el7.noarch
tfm-rubygem-foreman_setup-5.0.0-1.fm1_13.el7.noarch
foreman-release-1.15.6-1.el7.noarch

[ananace]

👍 Thanks, I'll try and get back to you tomorrow once I've had time to test it out myself, see if I can replicate the bug.

[ksagle77]

Awesome, thanks! I'm trying to debug as well, but am fairly new to ruby. Do you have an idea on whether the issue would be in fog-hyperv, or foreman-hyperv?

[ananace]

I think it's most likely a foreman-hyperv issue, and I have a rather good idea what might be causing it as well.

Sadly don't have any of the things I need to try the changes out myself at hand.

[ksagle77]

Cool. If there's anything I can do to help, let me know.

[ananace]

As a quick hack just to see if my hypothesis holds any water, you could try applying the change from the commit I just pushed.

[ksagle77]

Awesome, the error no longer appears after applying those changes!

[ananace]

Great, then I'll have a starting point for making a proper fix once I'm in place and can test my code properly.

[ananace]

Fix is merged into master as of bfbb7af, closing this issue.