Jump to content
Welcome to our new Citrix community!
  • 0

Bad network config during initial Pool Creation


Uwe Dittus

Question

Hi,

I believe I made a terrible mistake.

 

I had a XEN8.0 (Citrix Hypervisor) for tests running on a local machine here in the office. It has a local IP adress. Classic.

Then I wanted to move my VM guests to a dedicated server in the cloud. XEN installed, configured, cool.

 

Then, the transfer with in/export function in XenCenter of a guest was not working. Then I choose to create a pool, because I thought, that I simply can move the guest systems to the other server. You might know what happens.

 

Situation now:

XEN_1 in cloud, master, public accessible IP (new)

XEN_2 local, slave, local IP behind router (old)

 

 

XEN_1# xe pool-sync-database
You attempted an operation which involves a host which could not be contacted.
XEN_2# xe pif-list
The master reports that it cannot talk back to the slave on the supplied management IP address.
ip: 192.168.103.81

XEN_2# xe pif-list params=uuid,host-name-label,device,management
The host toolstack is still initialising. Please wait.

I assume, XEN_2 tells XEN_1 his IP which is crap and the so it's a deadlock situation.

 

Question: Is it possible to tell XEN_1 somehow the IP of my local router? Interrupt the process on the slave side, so that I could access the VMs again?

 

The very bad thing is, that the Local Storage of XEN_2 is not accessible and there is one guest system which has very very important content.

I saw this post here and now I'm actually frightend of loosing my data.

 

Any help is much appreciated! Thanks

Edited by citrix@disolva.de
Found additional information
Link to comment

4 answers to this question

Recommended Posts

  • 0

I'm so glad that I found a working solution. Many thanks to the guy here.

 

He motivated me to establish a VPN connection to XEN_2 and I can not tell you if I feel more surprized or amazed, that as soon as my two XENs are able to physically communicate, the pool is accessible and while I write these words, the VMs are getting exported.

 

Just in case one comes here and is desprate similarly... I had to use vpnc as VPN client, which is a bit of a journey on a XenServer as no repos are enabled and vpnc, in my case, had to be compiled manually. But as far as I recognized, it's a thing with the router I have (AVM FritzBox).

 

Btw. any admin is looking for a job?  ;)

  • Like 1
Link to comment
  • 0

This is because any transfers between hosts or pool have to take place on some common network. Non-routed networks won't work typically unless the network switch knows how to route them properly or you have a common private network within your organization that can be leveraged (like a 10 net), which is advisable, anyway, at least for XenServer hosts themselves from a security viewpoint.

 

-=Tobias

Link to comment
  • 0

You can always do an emergency network reset on a host that has a major network issue/mismatch and reset the primary management interface to something else. Afterwards, you would have to reset any of the networks.

You could also eject that host from the pool (though note that would result in all local SRs being destroyed!), do the network reset, and add it back into the pool in which case it will pick up the global pool network settings. Custom host-specific configurations, like iSCSI connections, will however not be maintained and will have to be redone by hand.

 

-=Tobias

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...