Solved: /32 static route problem in a vrf

Deniz AYDIN · ‎06-01-2015

Hi,

I have configured /32 route for a vrf which has /24 route for that prefix from another vrf. I am able to see /32 prefix in mpls forwarding table but AIB seems to be empty. When I have changed from /32 to /30, the problem is solved.

show route vrf test
Mon Jun 1 11:02:36.824 EEST

Codes: C - connected, S - static, R - RIP, B - BGP, (>) - Diversion path
D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
E1 - OSPF external type 1, E2 - OSPF external type 2, E - EGP
i - ISIS, L1 - IS-IS level-1, L2 - IS-IS level-2
ia - IS-IS inter area, su - IS-IS summary null, * - candidate default
U - per-user static route, o - ODR, L - local, G - DAGR
A - access/subscriber, a - Application route, (!) - FRR Backup path

Gateway of last resort is 10.0.0.2 to network 0.0.0.0

S* 0.0.0.0/0 [1/0] via 10.0.0.2, 2d19h
C 10.0.0.0/30 is directly connected, 2d20h, GigabitEthernet0/0/0/10.4045
L 10.0.0.1/32 is directly connected, 2d20h, GigabitEthernet0/0/0/10.4045
B 192.168.97.0/24 is directly connected, 2d17h, BVI97 (nexthop in vrf TN_NMC)
S 192.168.97.179/32 [1/0] via 10.0.0.2, 2d17h, GigabitEthernet0/0/0/10.4045

show mpls forwarding vrf test detail
Mon Jun 1 11:03:53.240 EEST
Local Outgoing Prefix Outgoing Next Hop Bytes
Label Label or ID Interface Switched
------ ----------- ------------------ ------------ --------------- ------------
16470 Unlabelled 0.0.0.0/0[V] Gi0/0/0/10.4045 10.0.0.2 828
Updated May 29 17:48:01.935
Version: 113, Priority: 3
MAC/Encaps: 18/18, MTU: 9000
Label Stack (Top -> Bottom): { Unlabelled }
NHID: 0
Packets Switched: 9

289891 Unlabelled 192.168.97.179/32[V] \
Gi0/0/0/10.4045 10.0.0.2 0
Updated May 29 17:56:09.341
Version: ?, Priority: ?
MAC/Encaps: ?/?, MTU: ?
Label Stack: { ? }
Packets Switched: ?

   show mpls forwarding vrf deltapetrol detail
   Mon Jun 1 11:05:17.918 EEST
   Local Outgoing Prefix Outgoing Next Hop Bytes
   Label Label or ID Interface Switched
   ------ ----------- ------------------ ------------ --------------- ------------
   16023 Unlabelled 192.168.97.176/30[V] \
   Gi0/0/0/10.4045 10.0.0.2 0
   Updated Jun 1 11:05:12.288
   Version: 690, Priority: 3
   MAC/Encaps: 18/18, MTU: 9000
   Label Stack (Top -> Bottom): { Unlabelled }
   NHID: 0x7a
   Packets Switched: 0

   16470 Unlabelled 0.0.0.0/0[V] Gi0/0/0/10.4045 10.0.0.2 828
   Updated May 29 17:48:01.936
   Version: 113, Priority: 3
   MAC/Encaps: 18/18, MTU: 9000
   Label Stack (Top -> Bottom): { Unlabelled }
   NHID: 0
   Packets Switched: 9

xthuijs · ‎06-26-2015

is a possibility indeed. jsut as additional FYI, the cisco live preso 2904 from sanfran 2014 has some detail about that rib behavior if you like more info and some show commands on how to track that down/verify.

xander

View solution in original post

DIEGO ZORRILLA · ‎06-01-2015

Can you provide additional info? configuration? the outputs when the /30 is configure insted of /32?

xthuijs · ‎06-03-2015

yeah a config would also help here I think. I'd like to see what vrf and config for Gi0/0/0/10.4045 was done. it seems a bit off that is unlabeled and then this interface is referenced in 2 vrf's, which can be done, if it was mpls enabled, but then it seems to target unlabeled.

show mpls ldp ne would also be helpful here.

xander

Deniz AYDIN · ‎06-26-2015

Hi all,

Sorry for delay. Its related with cef. I found a simalar case ;

CSCse46790 CEF prefers ARP adjacency over RIB next hop - See more at: https://supportforums.cisco.com/document/12098096/cscse46790-cef-prefers-arp-adjacency-over-rib-next-hop#sthash.7w4Dy3rb.dpuf

cef adjacency route override rib disable command solved my problem and we are planning it to apply on live env. But I am sure its gonna solve the problem.

xthuijs · ‎06-26-2015

is a possibility indeed. jsut as additional FYI, the cisco live preso 2904 from sanfran 2014 has some detail about that rib behavior if you like more info and some show commands on how to track that down/verify.

xander

Deniz AYDIN · ‎06-29-2015

definitely the same scenario (I remember that I saw this in the cisco live docs, but unable to find it than I posted:)

More detail about the configuration.

Two VRF'es, TEST and CE_TEST are configured. One VRF, TEST, has 10.20.97.0/24 prefix connected and the customer vrf, CE_TEST, is importing that but also has more specific route inside that prefix, 10.20.97.2/32, configured statically. Without that command you see the host route in the CE_TEST rib but fib shows the remote adjency in the TEST vrf for the 10.20.97.2/32.

after the command applied rib install in fib of the vrf.

vrf TEST
address-family ipv4 unicast
import route-target
12735:910
!
export route-target
12735:900
!
vrf CE_TEST
address-family ipv4 unicast
import route-target
12735:900
!
export route-target
12735:910
!

interface BVI97
description ### TEST ###
mtu 9014
vrf TEST
ipv4 address 10.20.97.3 255.255.255.0
!

interface GigabitEthernet0/0/0/0.17
vrf CE_TEST
ipv4 address 10.0.0.1 255.255.255.252
encapsulation dot1q 4045
!

router static
vrf CE_TEST
address-family ipv4 unicast
0.0.0.0/0 10.0.0.2
10.20.97.2/32 10.0.0.2
!

show route vrf CE_TEST
Mon Jun 29 06:59:15.194 UTC

Codes: C - connected, S - static, R - RIP, B - BGP, (>) - Diversion path
D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
E1 - OSPF external type 1, E2 - OSPF external type 2, E - EGP
i - ISIS, L1 - IS-IS level-1, L2 - IS-IS level-2
ia - IS-IS inter area, su - IS-IS summary null, * - candidate default
U - per-user static route, o - ODR, L - local, G - DAGR
A - access/subscriber, a - Application route, (!) - FRR Backup path

Gateway of last resort is 10.0.0.2 to network 0.0.0.0

S* 0.0.0.0/0 [1/0] via 10.0.0.2, 2d19h
C 10.0.0.0/30 is directly connected, 2d19h, GigabitEthernet0/0/0/0.17
L 10.0.0.1/32 is directly connected, 2d19h, GigabitEthernet0/0/0/0.17
B 172.16.1.0/29 is directly connected, 2d17h, BVI16 (nexthop in vrf TEST)
B 10.20.97.0/24 is directly connected, 2d17h, BVI97 (nexthop in vrf TEST)
S 10.20.97.2/32 [1/0] via 10.0.0.2, 2d17h

sh cef vrf CE_TEST 10.20.97.2/32 hardware egress location 0/0/CPU0
10.20.97.2/32, version 0, internal 0x4480001 0x0 (ptr 0x88448264) [1], 0x0 (0x8838cf70), 0x0 (0x0)
Updated Jun 29 06:54:41.910
local adjacency 10.20.97.2
Prefix Len 32, traffic index 0, Adjacency-prefix, precedence n/a, priority 0
via 10.20.97.2, BVI97, 6 dependencies, weight 0, class 0 [flags 0x0]
path-idx 0 NHID 0x0 [0x8a491aa8 0x0]
next hop VRF - 'TEST', table - 0xe0000014
next hop 10.20.97.2
local adjacency

   Mon Jun 29 07:00:18.249 UTC
   10.20.97.2/32, version 26, internal 0x4000001 0x0 (ptr 0x88448264) [1], 0x0 (0x0), 0x0 (0x0)
   Updated Jun 29 06:59:52.650
   Prefix Len 32, traffic index 0, precedence n/a, priority 3
   via 10.0.0.2, 5 dependencies, recursive [flags 0x0]
   path-idx 0 NHID 0x0 [0x88447c64 0x0]
   next hop 10.0.0.2 via 10.0.0.2/32

Sumanta Ghosh · ‎09-15-2017

Hi Experts

I am facing similar issue in IOS-XR. We have two different VRFs (A and B) and we are trying to leak routes selectively via inter-VRF static. It is not working everytime as expected. Appears to be intermittent issue. Does the below look ok in IOS-XR? It broke down 3 days back. Although it is working fine for few other /32 routes.

*******************************************

router static
vrf A
address-family ipv4 unicast
10.0.91.5/32 vrf B description ***Route-1***
10.20.3.6/32 vrf B 10.20.3.18 description ***Route-2***

10.33.211.152/29 vrf B description ***Route-3***

router static
vrf B
address-family ipv4 unicast
10.20.3.6/32 10.20.3.18 description ***Route-4***
10.32.0.0/22 vrf A description ***Route-5***

*********************************************************

sh cef vrf A 10.33.211.156 detail
Fri Sep 15 18:19:50.238 BST
10.33.211.152/29, version 6, internal 0x1000401 0x0 (ptr 0x72164914) [1], 0x0 (0x71fc0090), 0xa20 (0x723fa230)
Updated Sep 15 10:38:17.013
Prefix Len 29, traffic index 0, precedence n/a, priority 3
gateway array (0x71dcedd4) reference count 4, flags 0x20068, source rib (7), 0 backups
                [5 type 4 flags 0x208401 (0x7242c5d0) ext 0x0 (0x0)]
LW-LDI[type=1, refc=1, ptr=0x71fc0090, sh-ldi=0x7242c5d0]
gateway array update type-time 1 Sep 15 10:38:17.012
LDI Update time Sep 15 10:38:17.013
LW-LDI-TS Sep 15 10:38:17.013
   via 0.0.0.0/32, 0 dependencies, weight 0, class 0 [flags 0x0]
    path-idx 0 NHID 0x0 [0x7156b110 0x0]
    next hop VRF - 'B', table - 0xe0000011
    next hop 0.0.0.0/32
     local label 24007      labels imposed {None}

    Load distribution: 0 (refcount 5)

    Hash OK Interface                 Address
    0     Y   Unknown                   Lookup in table
RP/0/RSP0/CPU0:LD6-DCGW-SR01#

sh cef vrf B 10.32.0.132 detail
Fri Sep 15 18:20:43.630 BST
10.32.0.128/25, version 72, internal 0x1000001 0x0 (ptr 0x72170fb4) [1], 0x0 (0x71fc11e8), 0x0 (0x0)
Updated Sep 15 17:56:05.472
Prefix Len 25, traffic index 0, precedence n/a, priority 3
gateway array (0x71dcddd8) reference count 2, flags 0x0, source rib (7), 0 backups
                [3 type 3 flags 0x8401 (0x71e94710) ext 0x0 (0x0)]
LW-LDI[type=3, refc=1, ptr=0x71fc11e8, sh-ldi=0x71e94710]
gateway array update type-time 1 Sep 15 10:35:56.509
LDI Update time Sep 15 10:35:56.509
LW-LDI-TS Sep 15 17:56:05.472
   via 0.0.0.0/32, 0 dependencies, weight 0, class 0 [flags 0x0]
    path-idx 0 NHID 0x0 [0x7156b090 0x0]
    next hop VRF - 'A', table - 0xe0000014
    next hop 0.0.0.0/32

    Load distribution: 0 (refcount 3)

    Hash OK Interface                 Address
    0     Y   Unknown                   Lookup in table
RP/0/RSP0/CPU0:LD6-DCGW-SR01#

smailmilak · ‎09-15-2017

Hi,

Please provide the configuration for route leaking.

Sumanta Ghosh · ‎09-29-2017

Hi

Initially I had only given inter-VRF routes, for summary subnets between two VRFs. However TAC suggested me to use next hop or directly attached interface in the static route syntax. The IOS-XR was allowing the statement without any next hops and it was working well, when suddenly it broke down.