EIGRP – Poisoned floating summary-routes

Another cool little lab showcasing a new/redesigned functionality that apparently was introduced in IOS 15.x  that was new to me. The lab has the following requirements:

  • Configure EIGRP Classic on R4, R5 & R8 with AS100, only enable it on the .45.0 & .58.0 segments
  • Configure loopbacks on R4 & R5 with subnet 160.1.x.x/24 (x – router id) and advertise them
  • Advertise a summary default-route from R4
  • Advertise a summary route of the two 16.0.1.x.x/24 routes without overlap on R5 to R8
  • Modify the summary route of R5 so a null0 route isn’t installed

All basic steps until the last one so no explanation needed I feel. Keeping up with writing the full config in notepad only is starting to pay of as well, improving both speed and overall knowledge of syntax. Highly recommended! 🙂

! R4

interface Lo1
 ip add 160.1.4.4 255.255.255.0

router eigrp 100
 network 155.1.45.0 0.0.0.255 
 network 160.1.4.0 0.0.0.255

interface Gi1.45
 ip summary-address eigrp 100 0.0.0.0 0.0.0.0

! R5

interface Lo1
 ip add 160.1.5.5 255.255.255.0

router eigrp 100
 network 155.1.45.0 0.0.0.255
 network 155.1.58.0 0.0.0.255
 network 160.1.5.0 0.0.0.255

interface Gi1.58
 ip summary-address eigrp 100 160.1.4.0 255.255.254.0

! R8

router eigrp 100
 network 155.1.58.0 0.0.0.255

Checking the routes in R8 confirms that we “should” have connectivity to the two loopbacks.

R8#sh ip route eigrp | beg Gate
Gateway of last resort is 155.1.58.5 to network 0.0.0.0

D* 0.0.0.0/0 [90/3328] via 155.1.58.5, 00:18:58, GigabitEthernet1.58
155.1.0.0/16 is variably subnetted, 7 subnets, 2 masks
D 155.1.45.0/24 [90/3072] via 155.1.58.5, 00:18:58, GigabitEthernet1.58
160.1.0.0/23 is subnetted, 1 subnets
D 160.1.4.0 [90/130816] via 155.1.58.5, 00:00:43, GigabitEthernet1.58

But ping is still failing to R4:

R8#ping 160.1.4.4
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 160.1.4.4, timeout is 2 seconds:
U.U.U
Success rate is 0 percent (0/5)

R8#ping 160.1.5.5
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 160.1.5.5, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 4/5/6 ms

How come? All summary-routes created in ex. EIGRP & OSPF will also add a Null0-route to avoid routing-loops! As we only have a default-route to R4 longest match will hit our null0-route instead for 160.1.4.4.

R5#sh ip route eigrp | beg Gate
Gateway of last resort is 155.1.45.4 to network 0.0.0.0

D* 0.0.0.0/0 [90/3072] via 155.1.45.4, 00:22:19, GigabitEthernet1.45
160.1.0.0/16 is variably subnetted, 3 subnets, 3 masks
D 160.1.4.0/23 is a summary, 00:04:05, Null0

R5#sh ip cef 160.1.4.0 
160.1.4.0/23
attached to Null0

So how do we fix this? As mentioned a new feature was added to EIGRP in 15.x (I think in previous versions something similar was done on the interface instead). We can adjust the AD of our summary-route, and by setting it to 255 it will be deemed invalid (poisoned) and not installed in our routing table.

! R5
router eigrp 100
 summary 160.1.4.0/23 distance 255

The summary-route is still active in a way as our router will keep suppressing the advertisements of  160.1.5.0/24 and just forward the default-route instead.

R5#sh ip route eigrp | beg Gate
Gateway of last resort is 155.1.45.4 to network 0.0.0.0

D* 0.0.0.0/0 [90/3072] via 155.1.45.4, 00:26:02, GigabitEthernet1.45

We can now reach both R4 & R5’s loopback from R8.

R8#ping 160.1.4.4
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 160.1.4.4, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 5/6/7 ms

R8#ping 160.1.5.5
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 160.1.5.5, timeout is 2 seconds:
!!!!!
Success rate is 100 percent (5/5), round-trip min/avg/max = 5/5/6 ms

Lämna ett svar

E-postadressen publiceras inte. Obligatoriska fält är märkta *