This document provides answers to some of the more frequently asked
questions about IP Routing.
Note: For information on document conventions, refer to
Technical Tips Conventions.
What does it mean to have fast or autonomous switching "enabled" and
"disabled" on the same interface?
A. Look at this example:
Ethernet 6 is up, line protocol is up
Internet address is 220.127.116.11, subnet mask is 255.255.255.0
Broadcast address is 18.104.22.168
Address determined by non-volatile memory MTU is 1500 bytes
Helper address is 22.214.171.124
Outgoing access list is not set
Proxy ARP is enabled
Security level is default
Split horizon is enabled
ICMP redirects are always sent
ICMP unreachables are always sent
ICMP mask replies are never sent
IP autonomous switching is enabled
IP autonomous switching on the same interface is disabled
Gateway Discovery is disabled
IP accounting is disabled
TCP/IP header compression is disabled
Probe proxy name replies are disabled
If you enable fast or autonomous switching on an interface, packets
coming from any other interface on the router are fast-switched (or
autonomous-switched) to that interface. If you enable same-interface fast or
autonomous switching, packets whose source and destination address are the same
are fast or autonomous switched.
You can use same-interface fast or autonomous switching in cases where
you have Frame Relay or Asynchronous Transfer Mode (ATM) WAN links configured
as subinterfaces on the same main interface. Another situation is when you are
using secondary networks on LAN interfaces, as during IP address migration. In
order to enable same-interface fast switching, use the
How is the load shared between two parallel lines of equal capacity when
these lines are configured for load balancing?
A. For IP, if the router is fast switching, it load balances on a
per-destination basis. If the router is process switching, it load balances on
a per-packet basis. For more information, refer to
Does Load-Balancing Work? Cisco IOS® Software also supports both per
packet and per destination load balancing with Cisco Express Forwarding (CEF).
For more information, refer to
Balancing with CEF and
Load Balancing Over Parallel Links Using Cisco Express Forwarding.
What does route summarization mean?
A. Summarization is the process by which we collapse many routes with a
long mask to form another route with a shorter mask. Refer to
and Route Summarization and the "Summarization" section of
Interior Gateway Routing Protocol for more information. The
command works only if you have contiguous subnets. If you work with
discontiguous subnets, you need to use the
interface configuration command on every
interface that participates in the routing process where you want to configure
When does a Cisco router generate a source quench?
A. Prior to Cisco IOS® Software Releases 11.3 and 12.0, a Cisco router
generates a source quench only if it does not have the buffer space needed to
queue the packet. If the router can't queue the routed packet onto the output
interface's queue, it generates a source quench and registers an output drop
against the output interface. If the router isn't congested, it won't generate
a source quench.
You can look at the
command output for source quenches sent. Also
to see if there are any drops. If there are
none, then you should not see any source quench.
Cisco IOS Software Releases later than 11.3 and 12.0 do not include the
source quench feature.
When does a Cisco router initiate a routing request out its interfaces?
A. A Cisco router that runs a distance vector routing protocol initiates a
routing request out its interfaces if any of these conditions are met:
The interface goes down.
There is any change to the router global
There is any change to the metric
EXEC command is used.
interface configuration command is used.
The router is booted.
There is any change to the
The request is sent out to all interfaces configured for that
particular protocol no matter which interface triggers the request. The request
is sent out to one interface only if that is the only interface configured for
ip igrp events
ip igrp transactions
command is enabled, you see this in
any of these situations:
IGRP: broadcasting request on Ethernet0
IGRP: broadcasting request on Ethernet1
IGRP: broadcasting request on Ethernet2
IGRP: broadcasting request on Ethernet3
What is the difference between the ip default-gateway, ip
default-network, and ip route 0.0.0.0/0 commands?
command is used when IP routing is
disabled on the router. However,
are effective when IP routing is enabled
on the router and they are used to route any packets which do not have an exact
route match in the routing table. Refer to
a Gateway of Last Resort Using IP Command for more information.
How do I use the ip helper-address command to forward Bootstrap Protocol
command takes an argument of either the IP
address of the BOOTP server or a directed broadcast address for the segment on
which the BOOTP server resides. You can also have multiple instances of the
command with different IP addresses if you have more than one BOOTP server.
ip helper-address command can also be used on individual sub
Enhanced Interior Gateway Routing Protocol (EIGRP) redistributes with the
IGRP IP routing protocol automatically. Does EIGRP also interact with the
Routing Information Protocol (RIP) IP routing protocol?
A. EIGRP can interact with RIP using the
commands. Because RIP and EIGRP are so fundamentally different, automatic
interaction would probably produce unpredictable and undesirable results.
However, automatic interaction is possible between EIGRP and IGRP because of
their architectural similarities. Refer to
Routing Protocols for more information.
How do I configure my router to prefer an Open Shortest Path First (OSPF)
route over an EIGRP route when the route is learned from both sources?
A. The short answer is to use the distance
command under the routing process. OSPF has a default administrative distance
of 110 and EIGRP has a default administrative distance of 90 for internal
routes. If the same route prefixes are learned under both routing protocols,
EIGRP-learned routes will be installed into the IP routing table because of the
lower administrative distance (90 is less than 110). The key to having OSPF
routes installed in the Routing Information Base (RIB), instead of EIGRP
routes, is to make the administrative distance of OSPF less than that of EIGRP
that uses the
command. To learn more about administrative distance,
Is Administrative Distance?
Does the use of extended IP access control lists (ACLs) filter regular
routing updates (such as OSPF)? Do I need to explicitly permit the multicast
IPs used by routing protocols (such as 126.96.36.199 and 188.8.131.52, in the case of
OSPF) for updates to ensure the proper working of routing protocols?
A. Any IP ACL on an interface is applied to any IP traffic on that
interface. All IP routing updates packets are handled as regular IP packets at
the interface level, and, thus, they are matched with the ACL defined at the
interface using the
command. To ensure that the routing updates are not denied by ACLs, permit them
using the following statements.
To permit RIP use:
access-list 102 permit udp any any eq rip
To permit IGRP use:
access-list 102 permit igrp any any
To permit EIGRP use:
access-list 102 permit eigrp any any
To permit OSPF use:
access-list 102 permit ospf any any
To permit Border Gateway Protocol (BGP) use:
access-list 102 permit tcp any any eq 179
access-list 102 permit tcp any eq 179 any
For more information on ACLs, refer to
IP Access Lists and
Commonly Used IP ACLs.
Does the interface subcommand no arp arpa disable the Address Resolution
Protocol (ARP) function for a router interface?
A. By Advanced Research Projects Agency (ARPA) ARP, you mean "Ethernet
interfaces" and, by default, ARP ARPA is set with no
snap. This means that ARPA style ARPs are sent, but both ARPA and
Subnetwork Access Protocol (SNAP) are answered. By setting no arp
arpa, ARP requests are disabled, although null entries are
created for every station to which an ARP request is attempted. You can enable
SNAP alone, ARPA alone (the default), both SNAP and ARPA together (send two
ARPs every time), or neither SNAP nor ARPA (which is what happens if you set
no arp arpa without setting up any other ARP).
Would it be possible to configure a router for a 255.255.254.0 Ethernet
and a 255.255.252.0 serial subnet? Does IGRP/RIPv1 support variable subnetting?
A. Yes it is possible to configure these subnet masks. In order to subnet
on a Cisco router, the subnet bits must be contiguous, so 255.255.253.0 would
not be valid (11111111.11111111.11111101.00000000) while 184.108.40.206 would be
valid (11111111.11111111.11111100.00000000). Subnetting by borrowing all but
one bits from the host portion is not allowed. Also, traditionally, subnetting
with a single bit was not allowed. The masks above satisfy these conditions.
Addressing and Subnetting for New Users for more information.
IGRP RIP version 1 do not support variable length subnet masking
(VLSM). A single router running any of these protocols would work fine with
variable length subnetting. An incoming packet destined for one of the
configured subnets would be routed properly and delivered to the correct
destination interface. However, if the VLSM and discontiguous networks are
configured across multiple routers in IGRP domain, then it will lead to routing
problems. Refer to
Doesn't RIP or IGRP Support Discontiguous Networks? for more
The newer IP routing protocols, EIGRP, ISIS, and OSPF, as well as RIP
version 2, support VLSM, and they should be preferred in your network design.
Routing Protocols Technical Support Page for more information on all IP
Can an interface have more than one ip access-group statement in its
A. In Cisco IOS versions 10.0 and later, you can have two
commands per interface (one for each
interface ethernet 0
ip access-group 1 in
ip access-group 2 out
One access-group is used for inbound traffic
and one for outbound traffic. Refer to
Commonly Used IP ACLs and
IP Access Lists for more information on ACLs.
Can I configure two interfaces in the same subnet (t0 = 220.127.116.11/24
and t1 18.104.22.168/24)?
A. No. For the routing to work, each interface should be on a different
subnet. However, if you are only bridging, and not doing IP routing, then you
can configure the two interfaces on the same subnet.
Is it possible to have duplicate ip addresses for two serial interfaces
that belong to the same router?
A. Yes, duplicate ip addresses are allowed on serial interfaces. It is a
more efficient way of bundling links together (ie. MLPPP) and also a better way
to preserve address space. Change the encapsulation from the default HDLC to
PPP in order to assign duplicate ip addresses.
I have primary and secondary IP addresses configured on an Ethernet
interface and my router is running RIP (a distance vector routing protocol).
How does split-horizon affect the routing updates?
A. Refer to
Split Horizon Effects RIP/IGRP Routing Updates when Secondary Addresses Are
Is there a performance advantage when using the the IP access list
keyword established on an extended ACL? Does using
"established" make the access list more vulnerable? Do you have specific
examples of the usage?
A. There is no real performance advantage. The keyword
simply means that packets
with the acknowledgment (ACK) or reset (RST) bits set are let through. To learn
more about ACLs in general, refer to
IP Access Lists.
the internal hosts to make external TCP connections and to receive the return
control traffic. In most scenarios, this type of ACL would be essential on a
firewall configuration. The same result can also be achieved either by using
Reflexive ACLs or Context-Based Access Control. Refer to
Commonly Used IP ACLs for some sample configurations.
I have four equal cost parallel paths to the same destination. I am doing
fast switching on two links and process switching on the other two. How will
the packets be routed in this situation?
A. Assume that we have four equal cost paths to some set of IP networks.
Interfaces 1 and 2 fast switch (ip route-cache
enabled on the interface) , 3 and 4 do not (
). The router first establishes the four
equal cost paths in a list (path 1, 2, 3, and 4). When you do a
show ip route x.x.x.x, the four "next hops" to
The pointer is called interface_pointer on interface 1.
Interface_pointer cycles through the interfaces and routes in some orderly
deterministic fashion such as 1-2-3-4-1-2-3-4-1 and so on. The output of
ip route x.x.x.x has a "*" to the left of the "next hop"
that interface_pointer uses for a destination address not found in the cache.
Each time that interface_pointer is used, it advances to the next interface or
To illustrate the point better, consider this repeating loop:
A packet comes in, destined for a network serviced by the four
The router checks to see if it is in the cache. (The cache starts off
If it is in the cache, the router sends it to the interface stored in
the cache. Otherwise, the router sends it to the interface where the
interface_pointer is and moves interface_pointer to the next interface in the
If the interface over which the router just sent the packet is
running route-cache, the router populates the cache with that interface ID and
the destination IP address. All subsequent packets to the same destination are
then switched using the route-cache entry (thus they are fast-switched).
If there are two route-cache and two non-route-cache interfaces, there
is a 50 percent probability that a uncached entry will hit an interface that
caches entries, caching that destination to that interface. Over time, the
interfaces running fast switching (route-cache) carry all the traffic except
destinations not in the cache. This happens because once a packet to a
destination is process-switched over an interface, the interface_pointer moves
and points to the next interface in the list. If this interface is also
process-switched, then the second packet is process-switched over the interface
and the interface_pointer moves on to point to the next interface. Since there
are only two process-switched interfaces, the third packet will route to
fast-switched interface, which, in turn, will cache. Once cached in the IP
route-cache, all the packets to the same destination will be fast-switched.
Thus, there is a 50 percent probability that a uncached entry will hit an
interface that caches entries, caching that destination to that interface.
In case of a failure of a process-switched interface, the routing table
is updated and you would have three equal cost paths (two fast-switched and one
process-switched). Over time, the interfaces running fast switching
(route-cache) carry all the traffic except destinations not in the cache. With
two route-cache and one non-route-cache interfaces, there is a 66 percent
probability that a uncached entry will hit an interface that caches entries,
caching that destination to that interface. You can expect that the two fast
switched interfaces will carry all the traffic over time.
Similarly when a fast switched interface fails, you would have three
equal cost paths, one fast-switched and two process-switched. Over time the
interface running fast switching (route-cache) carries all the traffic except
destinations not in cache. There is 33 percent probability that a uncached
entry would hit an interface that cached entries, caching that destination to
that interface. You can expect that the single interface with caching enabled
will carry all of the traffic over time in this case.
If no interface is running route-cache, the router
round-robins the traffic on a packet-by-packet basis.
In conclusion, if multiple equal paths to a destination exist, some are
process-switched while others are fast switched, then over time most of the
traffic will be carried by the fast-switched interfaces only. The load
balancing thus attained is not optimum and might in some cases lower the
performance. Therefore, it is recommended that you do one of the following:
What is Unicast Reverse Path Forwarding (uRPF)? Can a default route
0.0.0.0/0 be used to perform a uRPF check?
A. Unicast Reverse Path Forwarding, used for preventing source address
spoofing, is a "look backward" ability which allows the router to check and see
if any IP packet received at a router interface arrives on the best return path
(return route) to the source address of the packet. If the packet was received
from one of the best reverse path routes, the packet is forwarded as normal. If
there is no reverse path route on the same interface from which the packet was
received, the packet is dropped or forwarded, depending on whether an access
control list (ACL) is specified in the
verify unicast reverse-path list interface
command. For more information, refer to the
Unicast Reverse Path Forwarding chapter of the
IOS Security Configuration Guide, Release 12.2
Default route 0.0.0.0/0 can not be used to perform a uRPF check. For
example, if a packet with source address 10.10.10.1 comes on Serial 0 interface
and the only route matching 10.10.10.1 is the default route 0.0.0.0/0 pointing
out Serial 0 on the router, the uRPF check fails and it drops that packet.
Who does load-balancing when there are multiple links to a destination,
Cisco Express Forwarding (CEF), or the routing
A. CEF does the switching of the packet based on the routing table which
is being populated by the routing protocols such as EIGRP, RIP, Open Shortest
Path First (OSPF), and so forth. CEF does the load-balancing once the routing
protocol table has been calculated. For more details on load balancing, refer
to How does
What are the maximum number of secondary IP addesses that can be
configured on a router interface?
A. There are no limits on configuring secondary IP addresses on a router
interface. For more information, refer to
What is the Pause control counter?
A. The Pause control counter indicates the number of times the router
requests another router to slow the traffic. For example, two routers, Router A
and Router B, are connected through a link with flow control enabled. If Router
B faces a traffic burst, Router B sends a Pause output packet to inform Router
A to slow the traffic because the link is oversubscribed. At that time, Router
A receives a Pause input packet that informs it of the request sent by Router
B. Pause output / input packets are not a problem or an error. They are simply
flow control packets between two devices.
Can a VLAN interface and a tunnel interface have the same IP
A. No. Bridging over tunnel is not supported, as the tunnel requires IP
traffic to be encapsulated in a GRE header, and you cannot encapsulate the
layer 2 traffic.
What is Virtual Routing and Forwarding (VRF)?
A. Virtual Routing and Forwarding (VRF) is a technology included in IP
network routers that allows multiple instances of a routing table to exist in a
router and work simultaneously. This increases functionality because it allows
network paths to be segmented without the use of multiple devices. Because
traffic is automatically segregated, VRF also increases network security and
can eliminate the need for encryption and authentication. Internet Service
Providers (ISPs) often take advantage of VRF in order to create separate
Virtual Private Networks (VPNs) for customers. Therefore the technology is also
referred to as VPN routing and forwarding.
VRF acts like a logical router, but while a logical router can include
many routing tables, a VRF instance uses only a single routing table. In
addition, VRF requires a forwarding table that designates the next hop for each
data packet, a list of devices that can be called upon to forward the packet,
and a set of rules and routing protocols that govern how the packet is
forwarded. These tables prevent traffic from being forwarded outside a specific
VRF path and also keep out traffic that must remain outside the VRF path.
How do I connect two different ISPs and route different traffic to
A. Policy based routing (PBR) is the feature that allows you to route the
traffic to different ISPs based on the source address.
What is the difference between the two methods to create static
A. There are two methods to create static routes:
The ip route 10.1.1.1 255.255.255.0 eth
0/0 command generates an ARP broadcast that looks for the
next-hop IP address.
The ip route 10.1.1.1 255.255.255.0
172.16.1.1 command does not generate an ARP request. It keeps
Layer 2 out of the routing process.
What is the purpose of ports 2228 and 56506?
A. Ports 2228 and 56506 are not registered port numbers. They can be used
by any application. Some applications initiate a connection with these port
numbers. Because of this, the port numbers are shown in the output of the
show ip sockets command. If the port numbers need to
be blocked, configure an access-list in order to block the ports.
What is the difference between point-to-point subinterfaces and
A. Point-to-point interfaces are used in serial communication. These types
of connections are assumed to transmit solely to the station at the opposite
end. The examples of Point-to-point are EIA/TIA 232, EIA/TIA 449, X.25, Frame
Relay, T-carrier, and OC3 - OC192.
Point-to-multipoint connects one station to several other stations.
Point-to-multipoint are of two types
In Point-to-multipoint Non-broadcast, the communication is replicated
to all the remote stations. Only specific, selected stations hear the
replicated communication. The examples are Frame Relay and ATM.
Point-to-multipoint Broadcast are characterized by a physical medum
that connects to all machines and where all communication is heard by all
Can you configure different MTU for subinterfaces under the same main
interface? How do 7500/GSR/ESR routers behave in this
A. You can configure different IP MTU with the
command on different subinterfaces. When you change
the MTU on a subinterface, the router checks the MTU from the main interface.
If the main interface MTU is set to a lower value than the one configured on
the subinterface, the router changes the MTU on the main interface to match
with the subinterface. Thus, the physical MTU configured with the
mtu command on the main interface needs to be higher
than the IP MTU configured on the sub interfaces.
Packet memory is carved based on the highest MTU configured on
75000/GSR. There is one exception to this; the Engine 4+ linecard does not
require to carve buffers on MTU change. On ESR, the packet memory is carved at
boot time and is not affected by MTU settings. So if you change the MTU, you
should not have any impact on ESR.
How do you limit the number of sessions when a customer accesses the
A. If the customers use the same IP address, then use the
ipcp address unique
command in order to reduce the number
of sessions that the customer uses.
How is accounting data age calculated?
A. The accounting data age increments its value in a 1 minute basis since
the time IP accounting was enabled. This continues until the clear
ip accounting command is issued, which resets it from 0.
What does the term threshold and timeout in IP SLA operation
A. Threshold sets the rising threshold that generates a reaction event and
stores history information for an IP SLAs operation.
Timeout sets the amount of time an IP SLAs operation waits for a
response from its request packet.
What is the significance of Time mentioned in the routing table
A. This is the age of the route in the routing table. It is the time
period for which the route is present in the routing table.
What is Network Descriptor Block (NDB)?
A. It is the network information, which is stored in "Routing table" with
Routing Descriptor Block (RDB). The memory to hold the IP routing table learned
prefixes is divided into NDB and RDB. Each route in Routing information Base
(RIB) requires one NDB and one RDB for each path. If the route is subnetted,
additional memory is required in order to maintain the NDB, and the direct
memory use for IP RIB can be shown with the
ip route summary