This may be a stupid question for most but with my limited knowledge of nexus 7000 gear I'll ask anyway. If you say ssh into a created vdc how do you get back to the default vdc if you didn't use a mgmt 0 interface . Used mgmt 0 as a vpc peer keepalive link . I know if you get into the box via console it takes you into the default vdc then you use the switchto command and you can just switchback to the default . Is this comand still available if you ssh into a given vdc ??? Hope this made sense.
Thanks Reza , understand that. The question for me is if I am using the mgtm 0 interface for vpc peer keep alive link how do you get to the default vdc other than thru the console port then if mgt 0 is being used ? I think i am missing something easy here .
Do you have an L3 routed link in the VDC? You could use that IP address to reach the VDC. Reza is correct when he says if you SSH to a VDC, you don't have the privilege to switchback to the default VDC.
Just as a pointer, don't use your mgmt0 interface for vPC peer keepalive links if you have dual SUPs. Use dedicated L3 links in a line card for this purpose.
Thanks , right I can ssh to the vdc that was created but how do you get to the default vdc . Is the only way to get to the default thru the console port then , that certainly doesn't seem correct or even practical. We don't have dual sup's so the mgt 0 should be ok for the keep alive link.
When you use mgmt0 for vpc peer link, you usually use a 3rd switch to connect the switches together and not back-to-back.
Now, you connect the 3560 to the rest of the network and you can use telnet or SSH to get to the default vdc or any other vdc you have.
default mgmt0 172.16.1.1
vdc test mgmt0 172.16.1.2
vdc prod mgmt0 172.16.1.3
and so on .....
Did I answer your question?
Reza, isn't it true that when you allocate mgmt0 interface on the 7K to multiple VDCs, even though the mgmt0 is shared, you cannot have mgmt0 of VDC 2 talk to mgmt0 of VDC3?
And, wouldn't the three VDC's mgmt0 IP be in different subnets?
However, I looked this up on Cisco,
"Basic device management is local to the virtual device context (VDC). The mgmt0 interface exists in the management virtual routing and forwarding instance (VRF) and is accessible from any VDC. You can configure a unique IP address for the mgmt0 interface in each VDC."
Say I have VDC 1 (default) and 3 other VDCs - (2,3,4), could you certify the following?
VDC 1 - mgmt0 - 172.16.0.1/30
VDC 2 - mgmt0 - 172.16.0.5/30
VDC 3 - mgmt0 - 172.16.0.9/30
VDC 4 - mgmt0 - 172.16.0.13/30
VDC 1 - mgmt0 - 172.16.0.1/24
VDC 2 - mgmt0 - 172.16.0.2/24
VDC 3 - mgmt0 - 172.16.0.3/24
VDC 4 - mgmt0 - 172.16.0.4/24
I thought the vrf management of each VDC is isolated. You'd need a physical L3 interface across VDCs for them to route to each other.
I have never tried putting the VDC management IPs in different subnets. I put all the VDCs in the same subnet.
And you are correct when you give vdc1 and IP and the same for vdc2, the IPs can talk to each other and that the correct behavior. After all the VDCs are no supposed to talk to each other unless you physically cable them together.