IPv6 Allocation Error When Trying to Clone a Newly Created CentOS 7.3 1611 KVM System

Good morning,

To start, we have recently migrated two older physical servers from Ubuntu 12.04LTS to CentOS as follows:

Server A:

HP Proliant DL360 GEN5 | Running: CentOS 6.9 (CentOS 7 did not support the RAID controller)

Server B:

HP Proliant DL360 GEN6 | Running: CentOS 7.3 1611

Both servers are configured as KVM hosts within Cloudmin.

While testing these two newly migrated servers, we created a CentOS7 Minimal KVM system from the following ISO:

CentOS-7-x86_64-Minimal-1611.iso

Once setup, we configured it to have a status of "Webmin". We then shut the KVM system down, with a plan to clone it. When trying to clone it however, it will not clone to any of the hosts in our KVM hosts infrastructure, it just sends back the following error:

".. failed : Host system hostname.domain does not have any automatic IPv6 allocation ranges defined, but no v6 addresses were supplied for the clone"

I'm not sure what to troubleshoot here. We have other KVM VMs shutdown setup as "templates", and they seem to clone between hosts without issues. Any support/feedback on why this may be happening would be appreciated. Please just let me know what other information would be pertinent to resolving this.

Thanks!

Status: 
Active

Comments

Do you care about IPv6? If not, one quick fix would be to bring up the VM, login to Webmin on it, go to Networking -> Network Configuration -> Network Interfaces, and remove all IPv6 addresses.

chadwick89's picture
Submitted by chadwick89 on Thu, 04/20/2017 - 10:57

Hello Jamie - thanks for the feedback. I have done as you've suggested and removed all IPv6 addresses from the KVM guest system, and both host systems. No systems have IPv6 enabled now. I then shutdown the KVM host and re-attempted to clone it from Server A to Server B - I ended up getting the same error.

I have deleted the KVM guest system, and I am going to try and create a new one from scratch and see if it clones properly. I will update this post once I've tried it.

Ok, please let us know. Also, if you removed the IPv6 addresses directly on the VM, make sure you click the "Refresh Fully" button at least once first before cloning (to update Cloudmin's view of the VM state).

chadwick89's picture
Submitted by chadwick89 on Fri, 04/28/2017 - 12:53

Ok, near as I can tell, there is no IPv6 enabled anywhere on these hosts. However I made a new KVM guest with the CentOS 7.3 1611 minimal ISO, and we are still getting the IPv6 error noted above. Any additional support/feedback on this would be hugely appreciated, I'm not sure what else to troubleshoot.

Thanks!

Any chance we could login to your Cloudmin master system to see what's going wrong here?

chadwick89's picture
Submitted by chadwick89 on Wed, 05/03/2017 - 12:20

Hi Jamie - I would be happy to provide you with the credentials. What would be the best way to coordinate?

Was this ever fixed? I am having the same problem.

@HarryZink - this is still open as we were unable to re-produce the problem on our test systems

chadwick89's picture
Submitted by chadwick89 on Tue, 08/15/2017 - 16:47

@JamieCameron

This issue is still happening to us. I'm not sure how to further troubleshoot, can we arrange remote access for you folks to see what may be going on? We were recently trying to clone a CentOS7 1611 VM which gave the same error again. We have not been able to figure out what's causing it.

Thanks again!

Yes, if you could give us remote access to a system that's exhibiting this problem it would be really useful to debug,

chadwick89's picture
Submitted by chadwick89 on Tue, 08/15/2017 - 17:37

Hi Jamie,

I'd be happy to. Can I email you with the access credentials and additional details?

chadwick89's picture
Submitted by chadwick89 on Wed, 10/25/2017 - 10:40

JamieCameron,

Good morning - so we tried to connect about this (back in early September), but apparently you didn't get the email containing access information. I tried re-sending it from my personal Gmail, but did not receive a response...so not sure if you got that one.

As this issue is still happening, can we try this again? I would be happy to provide the access information in whatever way works best.

Thanks again folks!

Can you re-send me the login details , and reference this ticket in the email subject?

chadwick89's picture
Submitted by chadwick89 on Wed, 11/01/2017 - 15:55

JamieCameron

Hi Jamie, I will resend the information as requested in the form of email. Thanks again!

Got the email ... I am taking a look now.

Ok, it does look like Cloudmin thinks that some of your VMs have IPv6 addresses, and so when trying to clone it wants to allocate a new IP.

Which specific VM are you trying to clone?

chadwick89's picture
Submitted by chadwick89 on Thu, 11/02/2017 - 10:59

JamieCameron,

Interesting - IPv6 isn't necessary in our infrastructure, perhaps a misconfig on our side somewhere? Anyway, we are trying to clone our CentOS7 1611 VMs, specifically:

centos1611.virtual (.73) - This is a template VM

We also previously tried cloning:

nodejs.virtual (.72) - This VM was cloned from a previous working template, but does not successfully clone now.

What would need to be done to ensure we don't run into this cloning snag?

Ok, so it looks like Cloudmin thought those VMs previously had IPv6 addresses, but actually they no longer did. Unfortunately this information was cached, and used when you tried cloning.

I forced a refresh of this cache with a command like cloudmin refresh-systems --host centos1611.virtual --regenerate , so cloning should work now. I suspect the problem arose because at some point the VM was switched from a static IP address to using DHCP.

chadwick89's picture
Submitted by chadwick89 on Fri, 11/03/2017 - 10:32

JamieCameron,

Morning, thank you very much for looking into this. I haven't had a chance to test yet...but I'll be doing that this AM and I'll get back to you. Assuming all is working, I'll document this for our future reference.

-Chad

Ok, let us know if this fixes the issue for you.