Lessons learned:
EIGRP Scalability –
Eigrp can achieve sub-second reconvergence through the use
of backup routes
--backup routes are feasible successors if they pass the
feasibility condition
If no backup routes – QUERY message is sent.
-Asks other neighbors for an alternate path
-Query domain can be bounded by
--summarization
--stub router advertisement
==================================================
Other features that can be used if no backup route exists. Eigrp will use the QUERY message for an
alternate path to a particular destination.
This is also part of the scalability issues in EIGRP. If we
send a QUERY message out we have to wait for a reply from all our neighbors
before we can remove the old information from the topology table. The query domain can be too large so when I’m waiting
for the query response back within the required amount of time and this can
cause a SIA or stuck in active problem.
SIA:
We lost our primary path to a destination and there is no
backup available so we have to ask the neighbors for an alternate path. The neighbors
continue to pass the query request on and if the responses dot come back in a
certain amount of time. You have to reset the neighbors. The loss of one
destination could result in the loss of all destinations.
We can reduce the size of the query domain through
Summarization – combines multiple longer matches and
combines then into a single summary.
EIGRP summarization
Done on a per interface basis.
-ip summary-address eigrp (network) (mask) (AD) at the link
level.
Features:
-supports any bit boundary including 0.0.0.0/0 – does not
have the bit boundary limitation like RIP
-automatically suppressed subnet advertisements
Example: if we have 3 routes
10.0.0.0/24
10.0.1.0 /24
10.0.2.0 /24
10.0.3.0/24
----------------------
These can be summarized as 10.0.0.0/22.
Once we do this – these other subnets are no longer going to
be advertised out the interface.
Summarization reduces the size of the routing table and it
also binds the query domain.
---can advertise subnets through “leak-map” arguments
We can leak certain subnets through the summary with a leak
map. Routes will always chose the longest match when trying to reach a
destination.
-Administrative distance defaults to 5
---allows for floating summaries -
-automatically generates discard route for the exact match
of the summary.
What the discard route is meant to do is in case the local router
loses one of the subnets of the aggregate. Then it does not continue to forward
traffic to another router that does now know about the subnet.
---can be removed with AD of 255 – for the summary.
-------------------------------------------------------------------
Once the summary is created it should not change the path selection.
Were just now saving space in the routing table and also if the upstream router
loses the routes to one of the subnets, when it sends the query messages to the
neighbors they’re immediately going to send the reply back that it doesn't not
have alt routes to the destination
I created 3 loopbacks on R4 – R4 has two paths upstream to
R3 and R4.
The interfaces have the following subnets
10.2.1.0/24
10.2.2.0/24
10.2.3.0/24
Summary route would
then be 10.2.1.0 255.255.252.0
R4(config-if)#ip summary-address eigrp 500 10.2.1.0
255.255.252.0
Shows the two new
summary routes for the /22
R2#sh ip eigrp topology 10.2.0.0/22
IP-EIGRP (AS 500): Topology entry for 10.2.0.0/22
State is Passive,
Query origin flag is 1, 1 Successor(s), FD is 156160
Routing Descriptor
Blocks:
192.168.24.4
(FastEthernet1/0), from 192.168.24.4, Send flag is 0x0
Composite metric
is (156160/128256), Route is Internal
Vector metric:
Minimum
bandwidth is 100000 Kbit
Total delay is
5100 microseconds
Reliability is
255/255
Load is 1/255
Minimum MTU is
1500
Hop count is 1
192.168.23.3
(FastEthernet0/1), from 192.168.23.3, Send flag is 0x0
Composite metric
is (158720/156160), Route is Internal
Vector metric:
Minimum
bandwidth is 100000 Kbit
Total delay is
5200 microseconds
Reliability is
255/255
Load is 1/255
Minimum MTU is
1500
Hop count is 2
R2#
We can then shut down one of the loopbacks AND DEBUG THE
OUTPUT.
R2#debug eigrp packets query reply
R2#
*Mar 1 00:57:48.387:
EIGRP: Received QUERY on FastEthernet1/0 nbr 192.168.24.4 – We now received a query about the prefix from R4
*Mar 1
00:57:48.387: AS 500, Flags 0x0, Seq
33/0 idbQ 0/0 iidbQ un/rely 0/0 peerQ un/rely 0/0
*Mar 1 00:57:48.403:
EIGRP: Enqueueing REPLY on FastEthernet1/0 nbr 192.168.24.4 iidbQ un/rely 0/1
peerQ un/rely 0/0 serno 26-26
*Mar 1 00:57:48.411:
EIGRP: Sending REPLY on FastEthernet1/0 nbr 192.168.24.4 – We’re sending a reply saying we DO NOT have an alternate path
*Mar 1
00:57:48.411: AS 500, Flags 0x0, Seq 52/33
idbQ 0/0 iidbQ un/rely 0/0 peerQ un/rely 0/1 serno 26-26
Leak MAP:
We can use a leak map to advertise a certain prefix out or
upstream to influence routing for a certain path.
We would append this under the interface on the summary
address.
ip prefix-list LEAKMAP seq 5 permit 10.2.2.0/24
route-map LEAKMAP permit 10
match ip address
prefix-list LEAKMAP
Under the interface:
ip summary-address eigrp 500 10.2.0.0 255.255.252.0 leak-map
LEAKMAP
Resync log:
*Mar 1 01:10:15.079:
%DUAL-5-NBRCHANGE: IP-EIGRP(0) 500: Neighbor 192.168.24.2 (FastEthernet0/0) is
resync: summary configured
R4#
Now on the upstream neighbor I have the single prefix for
the route:
D 10.2.2.0/24
[90/156160] via 192.168.24.4, 00:00:20, FastEthernet1/0
Now all traffic to the prefix will go the path with the
longest match
We can also use the summary address to install a default
route but the issue here is if we do that it’s going to suppress all the
subnets. Everything is a subnet of /0.
If I removed the previous summary address from both
interfaces and replace it with a summary that goes to 0.0.0.0
Example:
R4(config-if)#ip summary-address eigrp 500 0.0.0.0 0.0.0.0
This will cause all the neighbors to reconverge.
We now see all the previous routes that were summarized are
now summarized behind the default /0.
D* 0.0.0.0/0
[90/30720] via 192.168.24.4, 00:01:42, FastEthernet1/0
R2#
-----------------------------------------------------------------------------------------
D* 0.0.0.0/0
[90/30720] via 192.168.34.4, 00:02:42, FastEthernet0/1
R3#
ALL the routes behind the summarized routes will be aggregated
route.
Summarization has multiple
purposes.
-used to reduce the size of the forwarding table by taking
multiple routes and combining them together to shorter matches.
-also be selectively advertising longer and shorter matches,
we end up in a design where the routers are always preferring the longest match
in the network.
Note: if you aggregate the routes to a default route. The
router generating the zero route will not be able to use a default route to any
other destinations in the network because now the discard route the route to
Null0 – is now the longest match.
Example:
D* 0.0.0.0/0 is a
summary, 00:09:40, Null0
R4#
The reason is because the AD of any new summary is not
greater the Null route.
We can change this on the summary route by adding the AD at
the end of the summary
Example:
R4(config-if)#ip summary-address eigrp 500 0.0.0.0 0.0.0.0 91
No comments:
Post a Comment