First leaf on ACI simulator is inactive

[mazenbassiouny60876]

Dears,

I installed the ACI simulator software in a VM following all the instructions of the simulator documentation.

https://www.cisco.com/c/en/us/td/docs/switches/datacenter/aci/apic/sw/3-x/simulator/install_vm/b_Cisco_ACI_Simulator_VM_Install_Guide.html

The APIC is up and running and i find the first leaf is there to be registered. After i register the leaf to the pod, it appears as registered but in "Inactive" state. For sure the spine and the other leaf did not appear too. Please support and advise if it is a bug or i can troubleshoot the problem somehow.

Thank you 

Mazen Bassiouny

[Claudia de Luna]

Hi @mazenbassiouny60876, 

Did you change the node ID by any chance.  A screen shot of Fabric > Inventory >Fabric Membership (either Registerd or Pending depending on where it is) would be helpful.   I'm assuming you took all the defaults during the APIC set ups as well?

[mazenbassiouny60876]

Hello Claudia, 

Thanks for your response. I did all default configuration except only VTEP pool because the default one is used in my VMs environment( Attached all initial config). I changed also the NodeID but no effect. Attached screenshots for the node before and after registration which still in active node.

Best Regards,

Mazen Bassiouny 

[Claudia de Luna · Accepted solution]

Hi @mazenbassiouny60876,

That may be the issue. You really need to take all the default values on the Simulator.   From the Docs it says:
" 10.0.0.0/16 This value is for the infrastructure virtual routing and forwarding (VRF) only.  Default value. Do not change.  Do not set up your host on the network 10.0.0.0/8. "

Node id should also not be changed.

A quick test is to bring up the simulator with all defaults to see if that resolves the problem.  Don't change anything.  The TEP pool is internal and in this state there should be no conflict with your VM address pool.   Obviously you don't want the OOB Management on 10/8.  

That should be a quick and easy test to confirm if that is the issue. 

See Table 1
https://www.cisco.com/c/en/us/td/docs/switches/datacenter/aci/apic/sw/3-x/simulator/getting_started/b_ACI_Simulator_Guide_Rel_3_x/b_ACI_Simulator_Guide_Rel_3_x_chapter_010.html

[robertwieters]

So, I am working though this same issue. I have some follow up questions to help me understand the simulator and its effect on other networks:

1. Is the TEP pool completely internal? no traffic from this VRF leaves the simulator?

2. Is the Multicast Address Pool completely internal also?? Can I trust this traffic never leaves the simulator?

3. Why does this solution above mention to NOT put the OOB address on the 10.0.0.0/8 network? I only have this network available for OOB: 10.8.20.11/24 which is within that network. I can't put that interface in a classB or classC space due to how our VM infrastructure is layed out. 

[Robert Burns]

Both Infra TEP & MCast address are completely internal to the fabric operations.   As for the oob Management Address, as long as it doesn't overlap with the infra TEP subnet you're fine.  So if you need a 10.0.0.0/16 address for your out of band management network for example, configure your Infra TEP pool something like 11.0.0.0/16.  The reason is that APICs are not VRF-aware, so having an overlapping IP assigned on both as it's overlay TEP address and OOB Management IP presents a routing problem.

Robert

[RedNectar]

Hi @robertwieters ,

Possibly should be a new thread - this one has already been answered. But what the heck!

Firstly, since the original question specified "ACI simulator software in a VM" I'm going to assume your situation is the same.

1. Is the TEP pool completely internal? no traffic from this VRF leaves the simulator?

Correct. No traffic from this VRF leaves the simulator.

2. Is the Multicast Address Pool completely internal also?? Can I trust this traffic never leaves the simulator?

Correct. No multicast traffic  leaves the simulator.

3. Why does this solution above mention to NOT put the OOB address on the 10.0.0.0/8 network? I only have this network available for OOB: 10.8.20.11/24 which is within that network. I can't put that interface in a classB or classC space due to how our VM infrastructure is layed out.

The document is WRONG - it SHOULD read 
Do not set up your host on the network 10.0.0.0/16.

The reason for this can be seen if you issue a netstat -rn command on an APIC - simulator or real life APIC.

apic1# netstat -rn
Kernel IP routing table
Destination     Gateway         Genmask         Flags   MSS Window  irtt Iface
0.0.0.0         172.16.20.1     0.0.0.0         UG        0 0          0 apic1-eth1
0.0.0.0         192.168.11.254  0.0.0.0         UG        0 0          0 bond0.5
10.0.0.0        10.0.0.30       255.255.0.0     UG        0 0          0 bond0.4
10.0.0.30       0.0.0.0         255.255.255.255 UH        0 0          0 bond0.4
10.253.0.0      10.0.0.30       255.255.0.0     UG        0 0          0 bond0.4
172.16.20.0     0.0.0.0         255.255.255.0   U         0 0          0 apic1-eth1
192.168.11.0    0.0.0.0         255.255.255.0   U         0 0          0 bond0.5
192.168.11.254  0.0.0.0         255.255.255.255 UH        0 0          0 bond0.5

In your case, 10.8.20.11/24 should be fine for an OOB IP address unless there is something in the setup process that prevents it.  The reason that it can't be on 10.0.0.0/16 is because the APIC will attempt to send any reply packets it gets from a 10.0.x.x address to the infra subnet.

[robertwieters]

Thank you for that clarity!!

[samdilloway]

Currently attempting the Lab myself and having this same very problem and unable to find the fix online.

Default values used, mgmt network well away from the TEP 10.x.x.x range. VM resources up and over requirements.

Can discover the spine and two leafs but sitting deactivated.

[RedNectar]

Hi @samdilloway ,

Do you have a question? This post is already answered. If the answer didn't answer your question, ask a new one (with reference to this one if necessary)