07-15-2011 09:59 AM - edited 03-01-2019 09:59 AM
We are implementing our first UCS platform. We have everything physically connected, and can get to the HTTP GUI via IE. while SSH'ed into fabric A, I can ping any vlan in the network except for the one I created a management IP pool on the UCS system in. (we are using a class B address pool for all of the servers in the network, and using 1/2 of a class C portion of that network in the UCS). When I try to ping the default gateway of the management pool, I get a destination host unreachable message from one of the IPs in the pool (local to the UCS system). I can however traceroute to the default IP. I'm sure I am missing something, anyone with UCS knowledge able to see it? If you need any more detail let me know - if I didn't keep my hair cut short, I'd be pulling it out
07-15-2011 10:25 AM
Just want to confirm that the management pool IP addresses are from the same subnet as the addresses assigned to the Fabric Interconnects? These need to be on the same network. And is the address you are pinging one that has actually been asigned to a blade?
07-15-2011 11:28 AM
Hmmm...so far I am not trying to ping anything inside the UCS environment, so I think that invalidates the premise of your second question. when you say "same subnet as the addresses assigned to the Fabric Interconnects" do you mean the management IP/HTTP address of the UCS Heads? If you mean that, then no - but we should be able to change that easily.
If we do that, would the default gateway for the servers be the HTTP address of the heads, or the layer 3 address of the vlans on the cores (like it is for a normal server)?
07-15-2011 11:58 AM
Alright!!! I renumbered the management ports of A/B to be in the same subnet as the local block. However, that leads me to wonder: Does every server on the UCS chassis have to be in that same subnet? E.G. if we have a financial vlan of 7 and a voice vlan of 207, can voice servers and financial servers both exist on the same chassis? or even on two different chassis plugged into the same set of heads?
07-15-2011 01:17 PM
UCS is managed in an out-of-band fashion through the mgmt0 interface. Data (server-side) VLANs are not in contact with that interface. There is no internal bridging between mgmt0 and the DCE interfaces.
07-15-2011 06:56 PM
do mean by managment IPs the IP range used for KVM access ?
do you have a vlan created for that range and a L3 interface of the same vlan ( default GW) added in an upstream switch switch for example !
07-16-2011 01:27 AM
also check out this link it might be usefull for you !
https://supportforums.cisco.com/docs/DOC-6157
hope this help
if helpful Rate
Find answers to your questions by entering keywords or phrases in the Search bar above. New here? Use these resources to familiarize yourself with the community: