virtual labs and 2008 R2 servers

[mpoling]

Hello All,
I've been dealing with support for a while now, and they have been unable to tell me exactly why Surebackup restores are not working with my 2008 R2 servers. We even did a webex session for an hour. Everything works perfectly with my 2003 servers. I've gone through all the other posts I can find on this issue (making sure to use E1000 nic, making sure to use production gateway as IP for isolated network, naming isolated VM network same as production VM currently is connected to, etc). That being said, I apologize if this information is in the forum somewhere.

My setup is as follows:
The total environment is composed of 5 physical servers and 2 SANs.
Servers 1and2:
ESXi 4.1 (Essentials Plus) cluster of two machines sharing storage on a Dell MD3000i iscsi SAN. All is working wonderfully well.
Server 3:
Veeam and Vcenter (Essentials Plus) are sharing a physical 2008 64bit server that is connected to a 2nd, identical Dell SAN. While I understand some people don't like housing both Veeam and VCenter on the same server, I have had no issues with it (assuming my current issue is not related). Obviously
Server 4:
A stand alone VCenter (Enterprise ...don't ask. It was a purchase from a former IT Mgr.) server used to manage the ESXi servers that are dedicated to restores.
Server 5:
A single ESXi 4.1 (Enterprise) server to use as a sandbox and for ...hopefully U-Air restores...ah, someday.
Future server 6 (after I get the restores working):
Another ESXi server dedicated to Veeam Replications.

As I said, all is working well in this environment. Backups fly. We are currently backing up over the network (due to an early recommendation by Veeam tech), but due to info I've found on this forum, we'll soon be moving to backup via direct SAN.

In any case, the problem is creating a virtual lab that will correctly function (without error) on 2008 R2 64bit servers, like our Exchange 2007 server. I've changed the vlab config a dozen times, as has a Veeam tech, and every time we either get "No Networks" in which case there is no nic even present in the restored vm or "1 unmapped network" in which case the nic is present but fails to ping. Both of these errors cause the restore process to fail its tests and close.

Ideas? I'm happy to supply as much info as needed, and I guarantee that I'll supply the resolution when found. There are a couple other threads that have similar problems, but no solution. Awesome.

Thanks All.

Matt

[Frosty]

A shot in the dark, as I have no experience with SureBackup and haven't seen this problem, but maybe this might help ... login as administrator on one of the VMs where you have this problem and then run a command line:

SET devmgr_show_nonpresent_devices=1
START DEVMGMT.MSC

When Device Manager starts, go to View, Show Hidden Devices. See if there are any hidden network cards which might not be active but might still have IP addresses assigned to them? Just be careful; don't delete anything unless you are 100% certain it is not needed.

I have had to use this to clean up the setup of some of my VMs and templates.

[Frosty]

Maybe also double-check that the vSwitch config on the host where you are running the vLab is exactly the same as the vSwitch setup on the host where the VM was running? Case-sensitive and all that stuff. Also, do your servers have static IP addresses, or IPs allocated by a DHCP server?

EDIT: I see that you've already been through that sort of checking, sorry.

[Bunce]

How many restore points does the vm have? I had a few with many incrementals which wouldn't restore until I ran a full backup..

[foggy]

Matt, which version of Veeam B&R do you use and what is your support case ID number?

[mpoling]

Hey All,
Thanks for the replies.
Frosty,
I'll definitely check out the hidden nics. And while I have checked out my vSwitch config, I'm not 100% confident that I don't just have some dumb vlab config issue. Although you would think that a veeam tech could have solved that when they were logged into my system for an hour. As for IPs, yes, all my servers are static. And when the restored server boots into Windows, it does appear that the correct static IP does get set.

Bunce,
I have 2 different 2008 R2 servers. 1 (Exchange) with probably 14 or so restore points, mostly incremental, with the synthetic full every Sat. But I've also tried a fresh 2008 R2 server, and right from the very first backup, same issue.

Foggy,
My B&R version is 5.0.2.230 (64bit). Case # is 5143782. I'm just about to send them some screenshots of my production server and network configs.

Thanks again, All. Very much looking forward to a fully functional recovery solution.

[foggy]

Matt, you might try to use another installation of w2008R2 - maybe this is a template issue (if you are using a template to create those VMs), so trying to install it anew from another source will give some information. Also I would recommend to continue investigate the issue with support team. Thanks.

[mpoling]

Just wanted to post a solution to this issue. Neither I nor the Veeam tech knew exactly why, but I ended up having to put both the production Exchange server and the production DC on the same production virtual network. Typically (and ideally) I like to keep them on separate networks. In any case after moving the production DC to the Exchange network, then the restore process works. Both recovery VMs are able to use a single connection to the single production network. Not sure why creating 2 connections to the 2 original networks did not work. The Veeam tech did not know either.

[Gostev]

Hi Matt, I will check on your support case. Virtual labs natively support multiple networks and communication between those within the virtual lab, so this is somewhat unexpected issue to have. Thanks.

[Gostev]

Hi Matt, it looks like your case was handled by Tier 1 support engineer (there are 3 more tiers). Naturally, Tier 1 engineers may not have deep knowledge of specific components. I understand that he was ready to escalate this issue just before you have asked to close the case. We can re-open to investigate if necessary, since in theory this should work just fine with 2 network configuration. Most likely configuration related issue. Please ask to re-open the support case if you would like us to do so. Thanks!

[markfellows]

I’m experiencing a very similar problem, Veeam Support have been on and can’t seem to find a solution either.

Once the VMs have booted in the virtual lab I can manually add the network card and continue to use the lab as usual.

I presumed it could be to do with the VMs being on different host/switches but didn’t really want to put the DC and Exchange on the same host.

I’ll look to re-open the case.

[mi.systems]

Hi there,

I'm having a similar issue to that described in this post. I've tried the suggestions here - but still no luck...

Of the 4 VM's being restored to the virtual Lab only two of them both Win 2k8 R2 VMs deployed from the same image have the problem. The other two VMs are Win 7 and Win 2008.

All vm's have static assignments though do not keep there IP addresses - 3 of the 4 VMs are multi-homed though this is not a problem with the win 2008 vm.

Is there an official solution for this issue somewhere other than re-installing these VMs?

Thanks,

Daniel

[foggy]

Daniel, I would suggest to open a support case as OP did not reopen his case to find the solution, and Mark did not post his support case ID either, so I cannot check what was the solution.

[markfellows]

Can't entirely remember the solution to the problem but I think it had something to do with Virtual Network name. If I’m remembering this incident correctly the servers in my Virtual Lab originated from different ESXi hosts and had different Virtual Network Names. When I was attempting to use the Virtual Lab on server 1 the VM from server 2 didn’t restore with a NIC as the Virtual Network couldn’t be found.

As I said I can’t completely remember this issue but the Case ID was ID#5146380.

[foggy]

Mark, thanks for the update. As I see from your case, you had a webex with our tech who mapped the missing port group and after that you got the missing vNIC in SureBackup job.

[paulkroon]

I hate to reply to an older post, but there doesn't seem to be a clear resolution, and I believe I ran into the same issue. I haven't tested it yet, but this appears to be the resolution: http://kb.vmware.com/selfservice/micros ... Id=1020078. I will post again if this does solve the issue, as this is the first result when searching Google about the problem.

[Gostev]

Yes, this is what you need.

[Cokovic]

It's discussed in other threads here too. From my own experience i can tell you this will work.