Classless Inter-Domain
Routing
CIDR
CIDR (Classless Inter-Domain Routing)
• By 1992, members of the IETF were having serious concerns about the
exponential growth of the Internet and the scalability of Internet routing tables.
• The IETF was also concerned with the eventual exhaustion of 32-bit IPv4
address space.
• Projections were that this problem would reach its critical state by 1994 or
1995.
• IETF’s response was the concept of Supernetting or CIDR, “cider”.
• To CIDR-compliant routers, address class is meaningless.
• The network portion of the address is determined by the network subnet
mask or prefix-length (/8, /19, etc.)
• The first octet (first two bits) of the network address (or network-prefix) is
NOT used to determine the network and host portion of the network
address.
• CIDR helped reduced the Internet routing table explosion with supernetting and
reallocation of IPv4 address space.
CIDR• (Classless Inter-Domain Routing)
First deployed in 1994, CIDR dramatically improves IPv4’s scalability and
efficiency by providing the following:
• Eliminates traditional Class A, B, C addresses allowing for more efficient
allocation of IPv4 address space.
• Supporting route aggregation (summarization), also known as
supernetting, where thousands of routes could be represented by a single
route in the routing table.
• Route aggregation also helps prevent route flapping on Internet
routers using BGP. Flapping routes can be a serious concern with
Internet core routers.
• CIDR allows routers to aggregate, or summarize, routing information and thus
shrink the size of their routing tables.
• Just one address and mask combination can represent the routes to
multiple networks.
• Used by IGP routers within an AS and EGP routers between AS.
Without CIDR, a
router must
maintain individual
routing table
entries for these
class B networks.
With CIDR, a
router can
summarize
these routes
using a single
network
address by
using a 13-bit
prefix:
172.24.0.0 /13
Steps:
1. Count the number of left-most matching bits, /13 (255.248.0.0)
2. Add all zeros after the last matching bit:
172.24.0.0 = 10101100 00011000 00000000 00000000
CIDR (Classless Inter-Domain Routing)
• By using a prefix address to summarizes routes, administrators can keep
routing table entries manageable, which means the following
• More efficient routing
• A reduced number of CPU cycles when recalculating a routing table, or
when sorting through the routing table entries to find a match
• Reduced router memory requirements
• Route summarization is also known as:
• Route aggregation
• Supernetting
• Supernetting is essentially the inverse of subnetting.
• CIDR moves the responsibility of allocation addresses away from a
centralized authority (InterNIC).
• Instead, ISPs can be assigned blocks of address space, which they can then
parcel out to customers.
ISP/NAP Hierarchy - “The Internet: Still hierarchical after all
these years.” Jeff Doyle (Tries to be anyways!)
Supernetting Example
• Company XYZ needs to address 400 hosts.
• Its ISP gives them two contiguous Class C addresses:
• 207.21.54.0/24
• 207.21.55.0/24
• Company XYZ can use a prefix of 207.21.54.0 /23 to supernet these two
contiguous networks. (Yielding 510 hosts)
• 207.21.54.0 /23
• 207.21.54.0/24
• 207.21.55.0/24
23 bits in common
Supernetting Example
• With the ISP acting as the addressing authority for a CIDR block of addresses, the
ISP’s customer networks, which include XYZ, can be advertised among Internet
routers as a single supernet.
CIDR Restrictions
• Dynamic routing protocols must send network address and mask
(prefix-length) information in their routing updates.
• In other words, CIDR requires classless routing protocols for dynamic routing.
Example from online curriculum
Number of Networks Aggregated = 2^(network bits borrowed)
Are we over summarizing here?
Summarized and Specific Routes: Longest-bit Match
(more later) ISP
Summarized Update Specific Route Update
172.16.0.0/16 172.16.5.0/24
172.16.5.0/24
172.16.1.0/24
Sub1 Sub2
172.16.2.0/24 172.16.10.0/24
• ISP receives a summarized /16 update from Sub1 and a more specific /24
update from Sub2.
• ISP will include both routes in the routing table.
• ISP will forward all packets matching at least the first 24 bits of 172.16.5.0 to
Sub2 (172/16/5/0/24), longest-bit match.
• ISP will forward all other packets matching at least the first 16 bits to Sub1
(172.16.0.0/16).
Another example from online curriculum
Route flapping
• Route flapping occurs when a router interface alternates rapidly between the up and down
states.
• Route flapping can cripple a router with excessive updates and recalculations.
• However, the summarization configuration prevents the RTC route flapping from affecting any
other routers.
• The loss of one network does not invalidate the route to the supernet.
• While RTC may be kept busy dealing with its own route flap, RTZ, and all upstream routers, are
unaware of any downstream problem.
• Summarization effectively insulates the other routers from the problem of route flapping.
Short Term Solutions: IPv4 Enhancements
• CIDR (Classless Inter-Domain Routing) – RFCs 1517,
1518, 1519, 1520
• VLSM (Variable Length Subnet Mask) – RFC 1009
• Private Addressing - RFC 1918
• NAT/PAT (Network Address Translation / Port Address
Translation) – RFC
VLSM (Variable Length Subnet Mask)
• Limitation of using only a single subnet mask across a
given network-prefix (network address, the number of
bits in the mask) was that an organization is locked into
a fixed-number of of fixed-sized subnets.
• 1987, RFC 1009 specified how a subnetted network
could use more than one subnet mask.
• VLSM = Subnetting a Subnet
• “If you know how to subnet, you can do VLSM!”
VLSM Example using /30 subnets
207.21.24.0/24 network subnetted into eight /27 (255.255.255.224)
subnets
207.21.24.192/27 subnet, subnetted into eight /30
(255.255.255.252) subnets
• This network has seven /27 subnets with 30 hosts each AND eight
/30 subnets with 2 hosts each.
• /30 subnets are very useful for serial networks.
207.21.24.192/27 207.21.24. 11000000
/30 Hosts Bcast 2 Hosts
0 207.21.24.192/30 207.21.24. 110 00000 01 10 11 .193 & .194
1 207.21.24.196/30 207.21.24. 110 00100 01 10 11 .197 & .198
2 207.21.24.200/30 207.21.24. 110 01000 01 10 11 .201 & .202
3 207.21.24.204/30 207.21.24. 110 01100 01 10 11 .205 & .206
4 207.21.24.208/30 207.21.24. 110 10000 01 10 11 .209 & .210
5 207.21.24.212/30 207.21.24. 110 10100 01 10 11 .213 & .214
6 207.21.24.216/30 207.21.24. 110 11000 01 10 11 .217 & .218
7 207.21.24.220/30 207.21.24. 110 11100 01 10 11 .221 & .222
207.21.24.192/30 207.21.24.204/30
207.21.24.216/30
207.21.24.96/27 207.21.24.128/27
207.21.24.64/27
207.21.24.208/30 207.21.24.212/30
207.21.24.196/30 207.21.24.200/30
207.21.24.160/27 207.21.24.224/27 207.21.24.32/27 207.21.24.0/27
• This network has seven /27 subnets with 30 hosts each AND seven /30 subnets
with 2 hosts each (one left over).
• /30 subnets with 2 hosts per subnet do not waste host addresses on serial
networks .
VLSM and the Routing Table
Displays one subnet mask for all child routes.
Classful mask is assumed for the parent route.
Routing Table without VLSM
RouterX#show ip route
207.21.24.0/27 is subnetted, 4 subnets
C 207.21.24.192 is directly connected, Serial0
C 207.21.24.196 is directly connected, Serial1
C 207.21.24.200 is directly connected, Serial2
C 207.21.24.204 is directly connected, FastEthernet0
Each child routes displays its own subnet mask.
Routing Table with VLSM
RouterX#show ip route
Classful mask is included for theparent route.
207.21.24.0/24 is variably subnetted, 4 subnets, 2 masks
C 207.21.24.192 /30 is directly connected, Serial0
C 207.21.24.196 /30 is directly connected, Serial1
C 207.21.24.200 /30 is directly connected, Serial2
C 207.21.24.96 /27 is directly connected, FastEthernet0
• Parent Route shows classful mask instead of subnet mask of the child
routes.
• Each Child Routes includes its subnet mask.
Final Notes on VLSM
• Whenever possible it is best to group contiguous routes together so they can
be summarized (aggregated) by upstream routers. (coming soon!)
• Even if not all of the contiguous routes are together, routing tables use
the longest-bit match which allows the router to choose the more specific
route over a summarized route.
• Coming soon!
• You can keep on sub-subnetting as many times and as “deep” as you want to
go.
• You can have various sizes of subnets with VLSM.
Discontiguous subnets
• “Mixing private addresses with globally unique addresses can create
discontiguous subnets.” – Not the main cause however…
• Discontiguous subnets, are subnets from the same major network that are
separated by a completely different major network or subnet.
• Question: If a classful routing protocol like RIPv1 or IGRP is being used, what do the
routing updates look like between Site A router and Site B router?
Discontiguous subnets
• Classful routing protocols, notably RIPv1 and IGRP, can’t support discontiguous
subnets, because the subnet mask is not included in routing updates.
• RIPv1 and IGRP automatically summarize on classful boundaries.
• Site A and Site B are all sending each other the classful address of 207.21.24.0/24.
• A classless routing protocol (RIPv2, EIGRP, OSPF) would be needed:
• to not summarize the classful network address and
• to include the subnet mask in the routing updates.
Discontiguous subnets
• RIPv2 and EIGRP automatically summarize on classful boundaries.
• When using RIPv2 and EIGRP, to disable automatic summarization (on both routers):
Router(config-router)#no auto-summary
• SiteB now receives 207.21.24.0/27
• SiteB now receives 207.21.24.32/27
Short Term Solutions: IPv4 Enhancements
• CIDR (Classless Inter-Domain Routing) – RFCs 1517,
1518, 1519, 1520
• VLSM (Variable Length Subnet Mask) – RFC 1009
• Private Addressing - RFC 1918
• NAT/PAT (Network Address Translation / Port Address
Translation) – RFC
Private IP addresses (RFC 1918)
If addressing any of the following, these private addresses can be used instead of globally unique addresses:
• A non-public intranet
• A test lab
• A home network
Global addresses must be obtained from a provider or a registry at some expense.
Short Term Solutions: IPv4 Enhancements
• CIDR (Classless Inter-Domain Routing) – RFCs 1517,
1518, 1519, 1520
• VLSM (Variable Length Subnet Mask) – RFC 1009
• Private Addressing - RFC 1918
• NAT/PAT (Network Address Translation / Port Address
Translation) – RFC
Network Address Translation (NAT)
NAT: Network Address Translatation
• NAT, as defined by RFC 1631, is the process of swapping one address for
another in the IP packet header.
• In practice, NAT is used to allow hosts that are privately addressed to access
the Internet.
Network Address Translation (NAT)
TCP Source Port 1026 2.2.2.2 TCP Source Port 1923
TCP Source Port 1026 2.2.2.2 TCP Source Port 1924
• NAT translations can occur dynamically or statically.
• The most powerful feature of NAT routers is their capability to use port address translation
(PAT), which allows multiple inside addresses to map to the same global address.
• This is sometimes called a many-to-one NAT.
• With PAT, or address overloading, literally hundreds of privately addressed nodes can access
the Internet using only one global address.
• The NAT router keeps track of the different conversations by mapping TCP and UDP port
numbers.
Classless Routing Protocols
RIPv2
Classless routing protocols
• The true defining characteristic of classless routing protocols is the capability
to carry subnet masks in their route advertisements.
• “One benefit of having a mask associated with each route is that the all-zeros
and all-ones subnets are now available for use.”
• Cisco allows the all-zeros and all-ones subnets to be used with classful
routing protocols.
Classless Routing Protocols
“The true characteristic of a classless routing protocol is the ability to carry
subnet masks in their route advertisements.” Jeff Doyle, Routing TCP/IP
Benefits:
• All-zeros and all-ones subnets
• - Although some vendors, like Cisco, can also handle this with classful
routing protocols.
• VLSM
• Can have discontiguous subnets
• Better IP addressing allocation
• CIDR
• More control over route summarization
Classless Routing Protocols
Classless Routing Protocols:
• RIPv2
• EIGRP
• OSPF
• IS-IS
• BGPv4
Note: Remember classful/classless routing protocols is different than classful/classless
routing behavior. Classlful/classless routing protocols (RIPv1, RIPv2, IGRP, EIGRP,
OSPF, etc.) has to do with how routes get into the routing table; how the routing table
gets built. Classful/classless routing behavior (no ip classless or ip classless) has to do
with the lookup process of routes in the routing table (after the routing table has been
built). It is possible to have a classful routing protocol and classless routing behavior
or visa versa. It is also possible to have both a classful routing protocol and classful
routing behavior; or both a classless routing protocol and classless routing behavior.
RIP version 1
• Classful Routing Protocol, sent over UDP port 520
• Does not include the subnet mask in the routing updates.
• Automatic summarization done at major network boundaries.
• Updates sent as broadcasts unless the neighbor command is used
which sends them as unicasts.
0 1 2 3 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| command (1) | version (1) | must be zero (2) |
+---------------+---------------+-------------------------------+
| address family identifier (2) | must be zero (2) |
+-------------------------------+-------------------------------+
| IP address (4) |
+---------------------------------------------------------------+
| must be zero (4) |
+---------------------------------------------------------------+
| must be zero (4) |
+---------------------------------------------------------------+
| metric (4) |
+---------------------------------------------------------------+
RIP version 2
• Classless Routing Protocol, sent over UDP port 520
• Includes the subnet mask in the routing updates.
• Automatic summarization at major network boundaries can be disabled.
• Updates sent as multicasts (224.0.0.9) unless the neighbor command is
used which sends them as unicasts.
0 1 2 3 3
0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| command (1) | version (1) | must be zero (2) |
+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
| Address Family Identifier (2) | Route Tag (2) |
+-------------------------------+-------------------------------+
| IP Address (4) |
+---------------------------------------------------------------+
| Subnet Mask (4) |
+---------------------------------------------------------------+
| Next Hop (4) |
+---------------------------------------------------------------+
| Metric (4) |
+---------------------------------------------------------------+
Issues addressed by RIP v2
The following four features are the most significant new features added to RIP v2:
• Authentication of the transmitting RIP v2 node to other RIP v2 nodes
• Subnet Masks – RIP v2 allocates a 4-octet field to associate a subnet mask to a
destination IP address.
• Next Hop IP addresses – A better next-hop address, than the advertising router, if one
exists.
• It indicates a next-hop address, on the same subnet, that is metrically closer to the
destination than the advertising router.
• If this router’s interface is closest, then it is set to 0.0.0.0
• Multicasting RIP v2 messages – Multicasting is a technique for simultaneously
advertising routing information to multiple RIP or RIP v2 devices.
RIP v2 message format
• All the extensions to the original protocol are carried in the unused fields.
• The Address Family Identifier (AFI) field is set to two for IP. The only
exception is a request for a full routing table of a router or host, in which case
it will be set to zero.
Authentication
• RFC 1723 describes only simple password
authentication
• Cisco IOS provides the option of using MD5
authentication instead of simple password
authentication.
Same limitations of RIPv2 as with RIPv1
• Slow convergence and the need of holddown timers to
reduce the possibility of routing loops.
Note: See CCNA 2 for review if needed.
Same limitations of RIPv2 as with RIPv1
• RIP v2 continues to rely on counting to infinity as a means of
resolving certain error conditions within the network.
• Dependent upon holddown timers.
• Triggered updates are also helpful.
Note: See CCNA 2 for review if needed.
Same limitations of RIPv2 as with RIPv1
• Perhaps the single greatest limitation that RIP v2 inherited from RIP is that its
interpretation of infinity remained at 16.
Basic RIPv2 configuration
Other:
For RIP and IGRP, the passive interface command stops the router from sending
updates to a particular neighbor, but the router continues to listen and use routing
updates from that neighbor. (More later.)
Router(config-router)# passive-interface interface
Default behavior of version 1 restored:
Router(config-router)# no version
Compatibility with RIP v1 NewYork
interface fastethernet0/0
ip address 192.168.50.129 255.255.255.192
ip rip send version 1
RIPv2 ip rip receive version 1
interface fastethernet0/1
ip address 172.25.150.193 255.255.255.240
ip rip send version 1 2
• Interface FastEthernet0/0 is
configured to send and receive
RIP v1 updates. interface fastethernet0/2
• FastEthernet0/1 is configured ip address 172.25.150.225 225.255.255.240
to send both version 1 and 2
updates.
router rip
• FastEthernet0/2 has no special version 2
configuration and therefore
sends and receives version 2 network 172.25.0.0
by default. network 192.168.50.0
Discontiguous subnets and classless routing
router rip
version 2
no auto-summary
• RIP v1 always uses automatic summarization.
• The default behavior of RIP v2 is to summarize at
network boundaries the same as RIP v1.
Configuring authentication (EXTRA)
Router(config)# key chain Romeo
Router(config-keychain)# key 1
Router(config-keychain-key)# key-string Juliet
The password must be the same on both routers (Juliet), but the name of the key (Romeo) can be
different.
Router(config)# interface fastethernet 0/0
Router(config-if)# ip rip authentication key-chain Romeo
Router(config-if)# ip rip authentication mode md5
• If the command ip rip authentication mode md5 is not added, the interface will use the
default clear text authentication. Although clear text authentication may be necessary to communicate
with some RIP v2 implementations, for security concerns use the more secure MD5 authentication
whenever possible.
Show commands
show ip rip database
Router# show ip rip database
172.19.0.0/16 auto-summary
172.19.64.0/24 directly connected, Ethernet0
172.19.65.0/24
[1] via 172.19.70.36, 00:00:17, Serial1
[2] via 172.19.67.38, 00:00:25, Serial0
172.19.67.0/24 directly connected, Serial0
172.19.67.38/32 directly connected, Serial0
172.19.70.0/24 directly connected, Serial1
172.19.86.0/24[1] via 172.19.67.38, 00:00:25, Serial0
[1] via 172.19.70.36, 00:00:17, Serial1
• The show ip rip database command to check summary address
entries in the RIP database.
• These entries will appear in the database if there are only relevant child
or specific routes being summarized.
• When the last child route for a summary address becomes invalid, the
summary address is also removed from the routing table.
Router#show ip rip database
Show commands
Debug commands
RIPv2 Example
Scenario:
• Discontiguous subnets
• VLSM
• CIDR
• Supernet to 207.0.0.0/8
With the default
auto-summary on
ISP, it will load
balance for all
packets destined for
172.30.0.0/16
SantaCruz1
router rip
RIPv2 Example
network 172.30.0.0
network 192.168.4.0
version 2
no auto-summary
SantaCruz2
router rip
network 172.30.0.0
network 192.168.4.0
version 2
no auto-summary
ISP
router rip
redistribute static
network 10.0.0.0
network 192.168.4.0
version 2
no auto-summary
ip route 207.0.0.0 255.0.0.0 null0
SantaCruz2#show ip route
Examining a Routing Table
172.30.0.0/16 is variably subnetted, 6 subnets, 2 masks
C 172.30.200.32/28 is directly connected, Loopback2
C 172.30.200.16/28 is directly connected, Loopback1
R 172.30.2.0/24 [120/2] via 192.168.4.21, 00:00:21, Serial0
R 172.30.1.0/24 [120/2] via 192.168.4.21, 00:00:21, Serial0
C 172.30.100.0/24 is directly connected, Ethernet0
C 172.30.110.0/24 is directly connected, Loopback0
192.168.4.0/30 is subnetted, 2 subnets
R 192.168.4.24 [120/1] via 192.168.4.21, 00:00:21, Serial0
C 192.168.4.20 is directly connected, Serial0
R 10.0.0.0/8 [120/1] via 192.168.4.21, 00:00:21, Serial0
R 207.0.0.0/8 [120/1] via 192.168.4.21, 00:00:21, Serial0
Supernet, classless routing protcols
will route supernets (CIDR)
RIPv2: Sending and Receiving Updates
ISP(config)# line console 0
ISP(config-line)# logging synchronous
ISP#debug ip rip
RIP protocol debugging is on
ISP#01:23:34: RIP: received v2 update from 192.168.4.22 on Serial1
01:23:34: 172.30.100.0/24 -> 0.0.0.0 in 1 hops
01:23:34: 172.30.110.0/24 -> 0.0.0.0 in 1 hops
ISP# Includes mask
01:23:38: RIP: received v2 update from 192.168.4.26 on Serial0
01:23:38: 172.30.2.0/24 -> 0.0.0.0 in 1 hops
01:23:38: 172.30.1.0/24 -> 0.0.0.0 in 1 hops multicast
ISP#
01:24:31: RIP: sending v2 update to 224.0.0.9 via Ethernet0 (10.0.0.1)
01:24:31: 172.30.2.0/24 -> 0.0.0.0, metric 2, tag 0
01:24:31: 172.30.1.0/24 -> 0.0.0.0, metric 2, tag 0
01:24:31: 172.30.100.0/24 -> 0.0.0.0, metric 2, tag 0
01:24:31: 172.30.110.0/24 -> 0.0.0.0, metric 2, tag 0
01:24:31: 192.168.4.24/30 -> 0.0.0.0, metric 1, tag 0
01:24:31: 192.168.4.20/30 -> 0.0.0.0, metric 1, tag 0
<text omitted>
Adding a default Routes to RIPv2
ISP
router rip
redistribute static
network 10.0.0.0
network 192.168.4.0
version 2
no auto-summary
default-information originate
ip route 207.0.0.0 255.0.0.0 null0
ip route 0.0.0.0 0.0.0.0 10.0.0.2
etherenet0
Other RIPv2 Commands (EXTRA)
Router(config-router)# neighbor ip-address
Defines a neighboring router with which to exchange unicast routing information. (RIPv1 or
RIPv2)
Router(config-if)# ip rip send|receive version 1 | 2 | 1 2
Configures an interface to send/receive RIP Version 1 and/or Version 2 packets
Router(config-if)# ip summary-address rip ip_address
ip_network_mask
Specifies the IP address and network mask that identify the routes to be summarized.
Authentication and other nice configuration commands and examples:
http://www.cisco.com/en/US/products/sw/iosswrel/ps1831/products_configuration_guide
_chapter09186a00800d97f7.html
