Transcript Distance Vector Routing - Bismarck State College

Distance Vector Routing Protocols
Distance Vector Routing
Distance Vector Routing
Updates
• Routing updates occur periodically or
when the topology of the network
changes.
• Updates proceed from router to router.
• Each router sends its entire routing table
to each of its adjacent neighbors.
• Routing tables include information about:
 Total path cost as defined by the metrics
 Logical address of the first router on the path
to each network contained in the table
Distance Vector Update
Process
Routing Loops Introduction
• Routing loops can occur when routing tables
are inconsistent due to slow convergence in a
changing network.
Converged Network
• Just before the failure of the 192.168.0.0/24 network, all
routers have consistent knowledge and correct routing
tables.
• The network is said to have converged.
• For RTC, the preferred path to the 192.168.0.0/24
network is through RTB, and the distance from RTC to
the 192.168.0.0/24 network is 3.
Converged Network Example
RTE#show ip route
<output omitted>
C
192.168.0.0/24
is directly connected, FastEthernet1/0
RTA#show ip route
<output omitted>
R
192.168.0.0/24
[120/1] via 192.168.1.2, FastEthernet0/0
RTB#show ip route
<output omitted>
R
192.168.0.0/24
[120/2] via 192.168.2.2, Serial2/0
RTD#show ip route
<output omitted>
R
192.168.0.0/24
[120/2] via 192.168.3.2, Serial2/0
RTC#show ip route
<output omitted>
R
192.168.0.0/24
[120/3] via 192.168.4.2, Serial2/0
Network Failure
•
•
•
•
•
After the 192.168.0.0/24 network fails and an update is
scheduled to be sent, RTE sends an update to RTA.
RTA stops routing packets to the 192.168.0.0/24 network , but
RTB, RTC, and RTD continue to do so, because they have not
yet been informed of the failure.
When RTA sends out its normal update, RTB and RTD stop
routing to the 192.168.0.0/24 network .
However, RTC has not received an update.
For RTC, the 192.168.0.0/24 network can still be reached
through RTB.
Network Failure Example
RTE#show ip route
<output omitted>
No route to 192.168.0.0/24
RTA#show ip route
<output omitted>
No route to 192.168.0.0/24
RTB#show ip route
<output omitted>
No route to 192.168.0.0/24
RTD#show ip route
<output omitted>
No route to 192.168.0.0/24
RTC#show ip route
<output omitted>
R
192.168.0.0/24
[120/3] via 192.168.4.2, Serial2/0
The Routing Loop
•
•
•
•
Now RTC sends a periodic update to RTD, which indicates a
path to the 192.168.0.0/24 network through of RTB.
RTD changes its routing table to reflect this incorrect
information, and sends the information to RTA.
RTA sends the information to RTB and RTE, and the process
continues.
Any packet destined for the 192.168.0.0/24 network will now
loop from RTC to RTB to RTA to RTD and back again to RTC.
The Routing Loop Example
RTE#show ip route
<output omitted>
C
192.168.0.0/24
is directly connected, FastEthernet1/0
RTA#show ip route
<output omitted>
R
192.168.0.0/24
[120/5] via 192.168.3.1, Serial3/0
RTB#show ip route
<output omitted>
R
192.168.0.0/24
[120/6] via 192.168.2.2, Serial2/0
RTD#show ip route
<output omitted>
R
192.168.0.0/24
[120/4] via 192.168.5.1, Serial3/0
RTC#show ip route
<output omitted>
R
192.168.0.0/24
[120/3] via 192.168.4.2, Serial2/0
Looping Forever
• Packets will loop continuously around the
network, despite the 192.168.0.0/24 network
being down.
• Each time a router handles a packet it
increases the packet’s hop count value.
 Counting to infinity
Combating Routing Loops
• A number of dynamic routing
protocol measures can prevent
routing loops from occurring or limit
the problems they cause.





Defining a maximum hop count
Split horizon
Route poisoning
Triggered updates
Holddown timers
Defining Infinity
• Distance-vector protocols define infinity as a specific
maximum number of hops.
 This number usually refers to a maximum hop count.
 15 for RIP, 255 for IGRP
• The routing protocol permits the routing loop to continue
until the hop count reaches ‘infinity’.
• When the maximum hop count is reached, the
192.168.0.0/24 network is considered unreachable.
Split Horizon
• Using split horizon, a router will not
advertise a route update to the
router that originally informed it of
the route.
Split Horizon Example (1)
• RTA, RTB, and RTD exchange updates
indicating the 192.168.0.0 network is
down.
 RTC has not received the message, yet.
Split Horizon Example (2)
• Now, RTC sends a periodic update to RTD, which
indicates a path to the 192.168.0.0/24 network through
of RTB.
• RTD changes its routing table to reflect this incorrect
information.
 RTD will not send the updated information to RTA, as RTD
originally learned the 192.168.0.0/24 network was down
through RTA.
Split Horizon Results
RTE#show ip route
<output omitted>
No route to 192.168.0.0/24
RTA#show ip route
<output omitted>
No route to 192.168.0.0/24
RTB#show ip route
<output omitted>
No route to 192.168.0.0/24
RTD#show ip route
<output omitted>
R
192.168.0.0/24
[120/4] via 192.168.5.1, Serial3/0
RTC#show ip route
<output omitted>
R
192.168.0.0/24
[120/3] via 192.168.4.2, Serial2/0
Split Horizon Completion
• Eventually, RTB will send a periodic update to
RTC, indicating the 192.168.0.0/24 network is
unavailable.
• Then RTC will update its routing table and
send an update to RTD.
 Then, RTD will clear its route, again.
Route Poisoning
• Poisoned routes are advertised with a
distance higher than the maximum hop
count.
 For RIP, poisoned routes are advertised as
being 16 hops away.
 Indicates an unreachable network
• By itself, route poisoning does little.
 When used with triggered updates, route
poisoning speeds-up convergence time, by
immediately notifying other routers when a
network is down.
Triggered Updates
• By default a router running RIP advertises its
routing table very 30 seconds.
 Every 90 seconds for routers running IGRP
• Using triggered updates, when a route fails, an
update is sent immediately.
• The router that detects a topology change
immediately sends an update message to
neighbor routers that, in turn, send triggered
updates to their neighbors.
Holddown Timers
• When a router receives an update
indicating a route is down, it marks the
route as inaccessible and starts a holdtime timer.
• If the holddown timer expires, the route
is completely removed from the routing
table.
Holddown Timer Changes
• If at any time before the holddown timer
expires an update is received from:
 The same neighbor indicating that the network is
again accessible, the router marks the network as
accessible and removes the holddown timer.
 A different neighbor with a better metric than
originally recorded for the network, the router marks
the network as accessible and removes the
holddown timer.
 A different neighbor with a poorer metric:
• The update is ignored.
• This allows more time for the knowledge of a disruptive
change (routes frequently flapping from up to down) to
propagate through the entire network.
Holddown Timers Illustrated