cancel
Showing results for 
Search instead for 
Did you mean: 
cancel
1917
Views
0
Helpful
2
Replies

Unicast Adjacency Routes Nexus (AM)

stephendrkw
Level 3
Level 3

I have an odd situation, there is static route on my Nexus 7700 that routes anything for specific Class A subnet /24 to a F5 LTM Load Balancer external VRRP interface 10.x.153.15 and as you can see further below command outputs the old host 10.x.135.70 is working fine which is using an AM Adjacency route using nexus FIB table with a 250 cost and the new server 10.x.135.71 that is not working is using a standard static route cost of 1, I guess using the nexus RIB table. At the moment we can route to .70 fine which is using the AM Route but not .71 which is using the Static Route. No new routes have been added, I'm using the /24 route to the F5 LB 10.x.135.15 already on the switch.

 

I've performed a debug on the Load balancer and I can see the F5 Interface 10.x.153.15 (external interface) quite rightly asking 10.x.135.13 (Nexus Switch Vlan interface) for an ARP,  no reply because the Nexus switch is not resolving the ARP of 10.x.135.71 I can ping the host fine from the switch. The Server .71 is configured correctly  DG & Subnet 

 

F5

 

 3661  23.056618 0.000228 137 F5Networ_15:76:07 → Broadcast       ARP OUT s1/tmm0 : Who has 10.x.135.71? Tell 10.x.135.13

 

How do I make the new server 10.x.135.71 part of the AM routing table and shouldn't the FIB automatically find and assign the new server ARP to it's table?  I even contacted F5 they went through issue and said this is a Cisco Switch issue!

 

Switch A# sh ip route 10.x.135.70
IP Route Table for VRF "default"
'*' denotes best ucast next-hop
'**' denotes best mcast next-hop
'[x/y]' denotes [preference/metric]
'%<string>' in via output denotes VRF <string>

10.x.135.70/32, ubest/mbest: 1/0, attached
*via 10.x.135.70, Vlan137, [250/0], 5d11h, am
Switch A #

Switch A# sh ip route 10.x.135.71
IP Route Table for VRF "default"
'*' denotes best ucast next-hop
'**' denotes best mcast next-hop
'[x/y]' denotes [preference/metric]
'%<string>' in via output denotes VRF <string>

10.x.135.0/25, ubest/mbest: 1/0
*via 10.x.153.15, [1/0], 51w3d, static
Switch A#

 

Switch A# sh ip arp 10.x.135.70

Flags: * - Adjacencies learnt on non-active FHRP router
+ - Adjacencies synced via CFSoE
# - Adjacencies Throttled for Glean
D - Static Adjacencies attached to down interface

IP ARP Table
Total number of entries: 1
Address Age MAC Address Interface
10.x.135.70 00:06:51 x.x.817c Vlanx
Switch A# sh ip arp 10.x.135.71

Flags: * - Adjacencies learnt on non-active FHRP router
+ - Adjacencies synced via CFSoE
# - Adjacencies Throttled for Glean
D - Static Adjacencies attached to down interface

IP ARP Table
Total number of entries: 0
Address Age MAC Address Interface
Switch A#

2 Replies 2

Hello

Have you tried a L3 consistency check on the FIB and also maybe refreshing the rib for that particular prefix.
sh ip route hashing <sip> <dst ip>
sh ip adj | in 10.x153.
test forwarding ipv4 
inconsistency
sh forwarding ipv4 inconsistency
clear forwarding ipv4 route x.x.x.x


Please rate and mark as an accepted solution if you have found any of the information provided useful.
This then could assist others on these forums to find a valuable answer and broadens the community’s global network.

Kind Regards
Paul

Spoiler
Hi Paul, thanks. I found a few inconsistencies but not my route or any related route from the subnet I'm investigating I have logged a TAC case.

One thing I don't understand is how the Switch decides whether the route will join the FIB or RIB table. I will investigate further.
Review Cisco Networking products for a $25 gift card