This section will focus on how to troubleshoot Cisco ACI. It is important to understand that you can troubleshoot ACI using different methods such as:
Either of the three methods listed above will provide the information needed to diagnose a problem in ACI. In this particular section, we will be providing some important tips on how to troubleshoot ACI and which commands to leverage when someone needs to identify an issue. We will be focusing on the APIC Command Line Interface (CLI).
Click on the console ICON on the left to start the connection to the APIC Controller Username: acilabpod01 Password: cisco.123
At this point you should be in the apic prompt.
show switch: Provides the information of the entire "switch". Some times we refer to ACI as a "switch" because of the stateless and distributed architecture ACI has, with this command users can identify every single leaf and spine, and their respective information such as:
show switch
apic1# show switch ID Pod Address In-Band IPv4 In-Band IPv6 OOB IPv4 OOB IPv6 Version Flags Serial Number Name ---- ---- --------------- --------------- ------------------------- --------------- ------------------------- ------------------ ----- ---------------- ------------------ 101 1 10.9.200.129 10.0.236.177 :: 10.0.226.37 :: n9000-16.0(1j) asiv FDO221425XC S1 102 1 10.9.216.160 10.0.236.178 :: 10.0.226.38 :: n9000-16.0(1j) asiv FDO221425TL S2 203 1 10.9.200.130 10.0.236.173 :: 10.0.226.33 :: n9000-16.0(1j) aliv FDO21280JKY L3 204 1 10.9.88.66 10.0.236.174 :: 10.0.226.34 :: n9000-16.0(1j) aliv FDO21400SS1 L4 205 1 10.9.200.128 10.0.236.175 :: 10.0.226.35 :: n9000-16.0(1j) aliv FDO21360AU2 L5 206 1 10.9.216.161 10.0.236.176 :: 10.0.226.36 :: n9000-16.0(1j) aliv FDO21293NG8 L6 207 1 10.9.88.64 0.0.0.0 :: 10.0.226.110 :: n9000-16.0(1j) aliv FDO211218CD L7 208 1 10.9.240.32 0.0.0.0 :: 10.0.226.111 :: n9000-16.0(1j) aliv FDO20270CT3 L8 Flags - a:Active | l/s:Leaf/Spine | v:Valid Certificate | i:In-Service
ACI provides a very easy way to find out where a particular EndPoint is located anywhere in the fabric. The command is called show endpoints which provides valuable information from the EndPoints.
In this particular scenario, we will be troubleshooting POD01-WEB-SRV-02. If you recall from the previous section we left a continuous ping. This was done in order to make sure the EndPoint was still active in the ACI.
show endpoints ip 10.0.144.3
apic1# show endpoints ip 10.0.144.3 Legends: (P):Primary VLAN (S):Secondary VLAN Dynamic Endpoints: Tenant : aci_p01_tenant Application : aci_p01_ap AEPg : aci_p01_epg_web End Point MAC IP Address Source Node Interface Encap Multicast Address Create TS ----------------- ---------------------------------------- ------------ ---------- ------------------------------ --------------- --------------- -------------------- 00:50:56:02:01:18 10.0.144.3 learned,vmm 207 208 vpc aci_p01_intpolg_vpc vlan-1884 not-applicable 2023-01-26T15:29:44. 519+00:00 Total Dynamic Endpoints: 1 Total Static Endpoints: 0
As you can see POD01-WEB-SRV-02 is behind vPC in leafs 207 208 with the Interface Policy Group aci_p01_intpolg_vpc
The next step is to identify which port(s) are part of this vPC, since the show endpoints ip have already provided the vPC Policy Group aci_p01_intpolg_vpc. Let's identify the ports by doing the following command.
show vpc map aci_p01_intpolg_vpc
apic1# show vpc map aci_p01_intpolg_vpc Legends: N/D : Not Deployed Virtual Port-Channel Name Domain Virtual IP Peer IP VPC Leaf Id, Name Fex Id PC Id Ports -------------------------------- ---------- ---------------- ---------------- ---------- -------------------------------- ---------- ---------- -------------------- aci_p01_intpolg_vpc 207 10.9.104.67/32 10.9.240.32/32 347 208,L8 po4 eth1/1 aci_p01_intpolg_vpc 207 10.9.104.67/32 10.9.88.64/32 347 207,L7 po4 eth1/1
As you can see the port members are eth1/1 in Leaf 207 and eth1/1 on Leaf 208.
Then, you can follow the commands that you have been using on a daily basis such as - show interface, show port-channel, etc. You can execute these Leaf commands directly from the APIC. Here is an sample on how to check the interface eth1/1 on Leaf 207.
You should notice that you will be executing the commands on the leaf from the APIC prompt. There is no need to ssh to Leaf 207 since APIC is able to query Leaf 207.
fabric 207 show interface eth1/1
apic1# fabric 207 show interface eth1/1 ---------------------------------------------------------------- Node 207 (L7) ---------------------------------------------------------------- Ethernet1/1 is up admin state is up, Dedicated Interface Belongs to po4 Hardware: 100/1000/10000/25000/auto Ethernet, address: 00a3.8ebf.ff00 (bia 00a3.8ebf.ff00) MTU 9000 bytes, BW 10000000 Kbit, DLY 1 usec reliability 255/255, txload 1/255, rxload 1/255 Encapsulation ARPA, medium is broadcast
In the previous step, we configured ACI - VMWare integration. Because of this valuable integration, ACI has full visibility to the Virtualization environment and can provide information around the Virtual Machines. For example, we can do the following command to understand the characteristics of a VM from where the VM is located to the Guest OS.
show vmware domain name aci_p01_dc3_vds vm name POD01-WEB-SRV-02
apic1# show vmware domain name aci_p01_dc3_vds vm name POD01-WEB-SRV-02 VM Name : POD01-WEB-SRV-02 DVS : aci_p01_dc3_vds vCenter : 10.0.226.193 Host : pod01-compute1.ecatsrtpdmz.cisco.com Guest OS : CentOS 7 (64-bit) Configured OS : CentOS 7 (64-bit) VM OID : vm-7651 VM GUID : 500c648e-ae86-9753-4147-45345076d111 Power State : poweredOn Virtual Nics: Name : Network adapter 2 Type : Vmxnet3 MAC : 00:50:56:02:01:05 IP : 10.0.144.3 State : up Switch : aci_p01_dc3_vds Port Group : aci_p01_tenant|aci_p01_ap|aci_p01_epg_web Encap : vlan-551 PrimaryEncap : -- Adjacency : leafNone aci_p01_intpolg_vpc Name : Network adapter 1 Type : Vmxnet3 MAC : 00:50:56:02:00:05 IP : 10.0.145.2 State : up Switch : aci_p01_dc3_vds Port Group : aci_p01_tenant|aci_p01_ap_mgmt|aci_p01_epg_mgmt Encap : vlan-550 PrimaryEncap : -- Adjacency : leafNone aci_p01_intpolg_vpc
When troubleshooting routing issues, it is important to understant what are you troubleshooting. If you recall, ACI has an Underlay and an Overlay routing table. The Underlay, called Overlay-1, carries the routes of the VTEP IP Addresses. The Overlay carries the Tenant information, in your case it would be aci_p01_tenant.
Check the Underlay (overlay-1) routing table.
fabric 207 show ip route vrf overlay-1
apic1# fabric 207 show ip route vrf overlay-1 ---------------------------------------------------------------- Node 207 (L7) ---------------------------------------------------------------- IP Route Table for VRF "overlay-1" '*' denotes best ucast next-hop '**' denotes best mcast next-hop '[x/y]' denotes [preference/metric] in via output denotes VRF 3.0.0.102/32, ubest/mbest: 1/0 *via 10.9.216.160, eth1/49.15, [115/2], 01w06d, isis-isis_infra, L1 10.9.0.0/27, ubest/mbest: 1/0, attached, direct *via 10.9.0.30, vlan13, [1/0], 02w06d, direct 10.9.0.1/32, ubest/mbest: 2/0 *via 10.9.216.160, eth1/49.15, [115/12], 02w06d, isis-isis_infra, L1 *via 10.9.200.129, eth1/50.16, [115/12], 02w05d, isis-isis_infra, L1
Check the routing table for aci_p01_tenant
fabric 207 show ip route vrf aci_p01_tenant:aci_p01_vrf
apic1# fabric 207 show ip route vrf aci_p01_tenant:aci_p01_vrf ---------------------------------------------------------------- Node 207 (L7) ---------------------------------------------------------------- IP Route Table for VRF "aci_p1_tenant:aci_p1_vrf" '*' denotes best ucast next-hop '**' denotes best mcast next-hop '[x/y]' denotes [preference/metric] '%' in via output denotes VRF 0.0.0.0/0, ubest/mbest: 1/0 *via 10.9.200.130%overlay-1, [200/1], 08:39:07, bgp-65503, internal, tag 65503 1.1.1.1/32, ubest/mbest: 1/0 *via 10.9.200.130%overlay-1, [200/13], 08:39:07, bgp-65503, internal, tag 65503 10.0.0.5/32, ubest/mbest: 1/0 *via 10.9.200.130%overlay-1, [1/0], 08:39:07, bgp-65503, internal, tag 65503 10.0.4.0/31, ubest/mbest: 1/0 *via 10.9.200.130%overlay-1, [200/12], 08:39:07, bgp-65503, internal, tag 65503 10.0.5.0/31, ubest/mbest: 1/0 *via 10.9.200.130%overlay-1, [200/0], 08:39:07, bgp-65503, internal, tag 65503
Then you can check OSPF as well
fabric 203 show ip ospf neighbors vrf aci_p01_tenant:aci_p01_vrf
apic1# fabric 203 show ip ospf neighbors vrf aci_p01_tenant:aci_p01_vrf Total number of neighbors: 1 Neighbor ID Pri State Up Time Address Interface 10.0.1.1 1 FULL/DR 23:54:20 10.0.1.1 Eth1/1
As you can see, with ACI, the user can still leverage the same CLI commands that they have been leveraging in the past. More importantly, users can apply the same concepts when it comes to troubleshooting an issue. ACI leverages the same methodology as routing and forwarding as before.
Hopefully this troubleshooting session has provided some clarification on how to troubleshot an issue within the Cisco ACI Fabric. If you still have some specific troubleshooting questions/concerns please contact one of the proctors.