Hyper-v 2012 r2 woes.....

[Jaan]

Just a thought -

 

On the screenshots the Intel NICs all appear to be working, whereas the Broadcom's are not. When you bought a new NIC (I'm guessing a Quad adapter), was the original also Broadcom or were they an Intel?

 

the server (dell r610) has 4 broadcom nics on the board, the new card is the INTEL 4 port i350

 

cheers for looking

[sted]

im guessing that when you transfer a to the host new machine it dosent have any saved configuration for a network adapter so defaults to has no network so as to "fail safe" in as much as it dosent interrupt any of its normal vms. I think that's one of the advantages of replication

[Jaan]

im guessing that when you transfer a to the host new machine it dosent have any saved configuration for a network adapter so defaults to has no network so as to "fail safe" in as much as it dosent interrupt any of its normal vms. I think that's one of the advantages of replication

 

when using the "move wizhard" all the settings are retained....static ips, and network configuration..... it just stops working..... if i transfer it back... it starts to work again.

[sted]

when using the "move wizhard" all the settings are retained....static ips, and network configuration..... it just stops working..... if i transfer it back... it starts to work again.

 

that's that virtual pcs network config not the hosts so it may still have the same ip address but isn't bound to any virtual network adapter

[Jaan]

i thought the vswitch created in hyper-v manager did all that via layer 2?

 

only my vms have static ips and the management nic on the host.

 

but i've looked at this for far to long, non of t makes sense now lol

[edutech4schools]

What would happen if you used windows backup (which now can backup an entire vm) to backup and then restore a vm to the other host.

[IanT]

Can you check the logs again, also how is your switch ports setup?

[Jaan]

Can you check the logs again, also how is your switch ports setup?

 

Sorry for the late reply, end of term rush and all......

 

anyway..... the logs look fine, we have 4 core switches that are stacked..... with each physical nic (of the team) connected to 1 of each of the four switches.

 

The ports aren't aggregated/trunked. As the team has been configured as "switch independent"/hyper-v port.

 

The other hyper-v host we have is configured the same way.

 

cheers

[IanT]

What switching you using, HP or Cisco?

[rich_tech]

I am not a lover of stacked switches myself, we had Netgear GS724TS switches here and I had them stacked initially as it was relatively easy as they use a HDMI cable, but what I found was that pretty much any maintenance or upgrading of their firmwares requires you to break the stack and reboot the switches.

 

Also if one of the switches in the stack crashes as we had happen, it takes out the entire set for you with no redundancy.

 

Do you have both virtual switches on each server named the same ? Also do you have two sets of network adapters with the same IP ranges and gateways in the machine thats dropping the connection, windows caused issues with that with me before when that was the case.

[Jaan]

Yes we have HP Switches a5500-ei-sfp-24 from memory.

 

Both switch are named the same, both have different MAC ranges, the vms are all on the same ip range and use the same gateway.

 

thanks all.

[IanT]

I'll come back to the switching later.

 

Try this, create a new VM in Hyper-V Manager and manually attach the VHD's to the VM - try this.

[Jaan]

I'll come back to the switching later.

 

Try this, create a new VM in Hyper-V Manager and manually attach the VHD's to the VM - try this.

 

 

create new, or import?

[IanT]

New VM

[Jaan]

Thought i'd report back to say i've sorted this issue.

 

Setting the Teaming mode to "LACP", load balancing mode to "Dynamic" and configuring the switch port to allow LACP sorted the migrate issues.