TCP/IP Suite Error and Control

Messages (ICMP)
Overview

• Knowledge of ICMP control messages is an essential part

of network troubleshooting and is a key to a full
understanding of IP networks.
• This module will:
– Describe ICMP
– Describe the ICMP message format
– Identify ICMP error message types
– Identify potential causes of specific ICMP error
messages
– Describe ICMP control messages
– Identify a variety of ICMP control messages used in
networks today
– Determine the causes for ICMP control messages

2
Overview

IP is a best effort delivery system.

• Data may fail to reach its destination for a variety of reasons, such
as hardware failure, improper configuration or incorrect routing
information.
• IP does not have a built-in mechanism for sending error and control
messages.
• IP also lack a mechanism for host and management queries.
Internet Control Message Protocol (ICMP) was designed to handle
these issues.
3
ICMP

• ICMP messages can be divided into categories (depending

upon the author.
• The Cisco curriculum divides it into:
– Error-Reporting Messages
– Suite Control Messages
4
Internet Control Message Protocol (ICMP)

• IP is an unreliable method for delivery of network data.

• Nothing in its basic design allows IP to notify the sender that a data
transmission has failed.
• Internet Control Message Protocol (ICMP) is the component of the
TCP/IP protocol stack that addresses this basic limitation of IP.
• ICMP does not overcome the unreliability issues in IP.
• Reliability must be provided by upper layer protocols (TCP or the
application) if it is needed. .
5
Error reporting and error
correction
• When datagram delivery errors
occur, ICMP is used to report
these errors back to the source
ICMP X
of the datagram. msg
source
destination

Example
• Workstation 1 is sending a datagram to Workstation 6
• Fa0/0 on Router C goes down
• Router C then utilizes ICMP to send a message back to Workstation 1
indicating that the datagram could not be delivered.
• ICMP does not correct the encountered network problem.
• Router C knows only the source and destination IP addresses of the
datagram, not know about the exact path the datagram took to Router
C, therefore, Router C can only notify Workstation 1 of the failure
• ICMP reports on the status of the delivered packet only to the source
device.
6
 Format of an ICMP
Message
http://www.iana.org/assignments/icmp-parameters

Type Field
Type Name Type Name
---- ------------------------- ---- -------------------------
0 Echo Reply 17 Address Mask Request
1 Unassigned 18 Address Mask Reply
2 Unassigned 19 Reserved (for Security)
3 Destination Unreachable 20-29 Reserved (for Robustness Experiment)
4 Source Quench 30 Traceroute
5 Redirect 31 Datagram Conversion Error
32 Mobile Host Redirect
6 Alternate Host Address
33 IPv6 Where-Are-You
7 Unassigned
34 IPv6 I-Am-Here
8 Echo
35 Mobile Registration Request
9 Router Advertisement 36 Mobile Registration Reply
10 Router Solicitation 37 Domain Name Request
11 Time Exceeded 38 Domain Name Reply
12 Parameter Problem 39 SKIP
13 Timestamp 40 Photuris
14 Timestamp Reply 41-255 Reserved
15 Information Request
16 Information Reply

7
 Format of an ICMP
Message
http://www.iana.org/assignments/icmp-parameters

Code Field Many of these ICMP types have a "code"

field.
Type 3: Destination Unreachable Here are the assigned code fields for Type 3
Destination Unreachable.
Codes
0 Net Unreachable Codes 2 and 3 are created only by the
1 Host Unreachable Destination Host, all others are created only
2 Protocol Unreachable by routers.
3 Port Unreachable
4 Fragmentation Needed and Don't Fragment was Set
5 Source Route Failed
6 Destination Network Unknown
7 Destination Host Unknown
8 Source Host Isolated
9 Communication with Destination Network is Administratively Prohibited
10 Communication with Destination Host is Administratively Prohibited
11 Destination Network Unreachable for Type of Service
12 Destination Host Unreachable for Type of Service
13 Communication Administratively Prohibited
14 Host Precedence Violation
15 Precedence cutoff in effect
8
ICMP Error Messages
Unreachable Destinations

• Network Unreachable
– generated by router lacking any route to destination
• Host Unreachable
– last hop router cannot contact destination
• Port Unreachable
– no process bound to port

10
Unreachable
networks

Network communication depends upon certain basic conditions being met:

• Sending and receiving devices must have the TCP/IP protocol stack
properly configured.
 proper configuration of IP address and subnet mask.
 A default gateway must also be configured if datagrams are to travel
outside of the local network.
• A router also must have the TCP/IP protocol properly configured on its
interfaces, and it must use an appropriate routing protocol.
If these conditions are not met, then network communication cannot take
place.
11
Unreachable
networks

Examples of problems:
• Sending device may address the datagram to a non-existent IP
address
• Destination device that is disconnected from its network.
• Router’s connecting interface is down
• Router does not have the information necessary to find the destination
network.

12
Destination unreachable message

ICMP Destination Unreachable

Type = 3

• If datagrams cannot always be forwarded to their destinations,

ICMP delivers back to the sender a destination unreachable
message indicating to the sender that the datagram could not be
properly forwarded.
• A destination unreachable message may also be sent when packet
fragmentation is required in order to forward a packet.
– Fragmentation is usually necessary when a datagram is forwarded
from a Token-Ring network(4500-18000) to an Ethernet network.
– If the datagram does not allow fragmentation, the packet cannot be
forwarded, so a destination unreachable message will be sent.
– More a little later on fragmentation and MTU Path Discovery!
• Destination unreachable messages may also be generated if IP
related services such as FTP or Web services are unavailable.
13
ICMP Echo (Request) and Echo Reply

Echo = Type 8
Echo Reply = Type 0
Notice that the code is 0 for both
Ethernet Header IP Header ICMP Message Ether.
(Layer 2) (Layer 3) (Layer 3) Tr.
Ethernet Ethernet Frame Source IP Add. Type Code Check- ID Seq. Data FCS
Destination Source Type Dest. IP Add. 0 or 8 0 sum Num.
Address Address Protocol field
(MAC) (MAC)

• IP Protocol Field = 1
• The echo request message is typically initiated using the ping
command .
14
 Detecting excessively long routes
IP Header
0 15 16 31
4-bit 4-bit 8-bit Type Of
Version Header Service 16-bit Total Length (in bytes)
Length (TOS)
3-bit
16-bit Identification Flags 13-bit Fragment Offset

8 bit Time To Live 8-bit Protocol 16-bit Header Checksum

TTL

32-bit Source IP Address

32-bit Destination IP Address

ICMP Time Exceeded
Options (if any)
Type = 11
Data

• A TTL value is defined in each datagram (IP packet).

• As each router processes the datagram, it decreases the TTL value by
one.
• When the TTL of the datagram value reaches zero, the packet is
discarded.
• ICMP uses a time exceeded message to notify the source device that
the TTL of the datagram has been exceeded.

15
http://www.switch.ch/docs/ttl_default.html
TTL Overview - Disclaimer:
The following list is a best effort overview of some widely used TCP/IP stacks. The
information was provided by vendors and many helpful system administrators. We would
like to thank all these contributors for their precious help ! SWITCH cannot, however,
take any responsibility that the provided information is correct. Furthermore, SWITCH
cannot be made liable for any damage that may arise by the use of this information.

+--------------------+-------+---------+---------+
| OS Version |"safe" | tcp_ttl | udp_ttl |
+--------------------+-------+---------+---------+
AIX n 60 30 Assigned Numbers (RFC
DEC Pathworks V5 n 30 30
1700, J. Reynolds, J.
FreeBSD 2.1R y 64 64
HP/UX 9.0x n 30 30 Postel, October 1994):
HP/UX 10.01 y 64 64
IP TIME TO LIVE
Irix 5.3 y 60 60
Irix 6.x y 60 60 PARAMETER
Linux y 64 64
MacOS/MacTCP 2.0.x y 60 60
The current
OS/2 TCP/IP 3.0 y 64 64 recommended default
OSF/1 V3.2A n 60 30 time to live (TTL)
Solaris 2.x y 255 255 for the Internet
SunOS 4.1.3/4.1.4 y 60 60 Protocol (IP) is 64.
Ultrix V4.1/V4.2A n 60 30
VMS/Multinet y 64 64
VMS/TCPware y 60 64
VMS/Wollongong 1.1.1.1 n 128 30 Safe: TCP and UDP
VMS/UCX (latest rel.) y 128 128 initial TTL values
MS WfW n 32 32
MS Windows 95 n 32 32
should be set to a
MS Windows NT 3.51 n 32 32 "safe" value of at
MS Windows NT 4.0 y 128 128 least 60 today.
16
IP Parameter Problem

ICMP Parameter Problem

Type = 12

• Devices that process datagrams may not be able to forward a

datagram due to some type of error in the header.
• This error does not relate to the state of the destination host or
network but still prevents the datagram from being processed and
delivered.
• An ICMP type 12 parameter problem message is sent to the source of
the datagram.

17
ICMP Control Messages
Introduction to ICMP Control Messages

• Unlike error messages, control messages are not the

results of lost packets or error conditions which occur
during packet transmission.
• Instead, they are used to inform hosts of conditions such
as:
– Network congestion
– Existence of a better gateway to a remote network

19
ICMP Redirect
3

ICMP Redirect 2

Type = 5 Code = 0 to 3 1 2
4

• ICMP Redirect messages can only be sent by routers

• Host H sends a packet to Host 10.1.1.1 on network 10.0.0.0/8.
• Since Host H is not directly connected to the same network, it forwards
the packet to its default gateway, Router R1 at 172.16.1.100.
• Router R1 finds the correct route to network 10.0.0.0/8 by looking in its
route table.
• It determines that the path to the network is back out the same
interface the request to forward the packet came from to Router R2 at
172.16.1.200.
• R1 forwards the packet to R2 and sends an ICMP redirect/change
request to Host H telling it to use Router R2 at 172.16.1.100 as the
gateway to forward all future requests to network 10.0.0.0/8.
20
ICMP Redirects

ICMP Redirect
Type = 5 Code = 0 to 3

• Default gateways only send ICMP redirect/change request messages if

the following conditions are met:
– The interface on which the packet comes into the router is the
same interface on which the packet gets routed out.
– The subnet/network of the source IP address is the same
subnet/network of the next-hop IP address of the routed packet.
– The datagram is not source-routed.
– The route for the redirect is not another ICMP redirect or a default
route.
– The router is configured to send redirects. (By default, Cisco
routers send ICMP redirects. The interface subcommand no ip
redirects will disable ICMP redirects.)

21
Clock synchronization and transit time
estimation Replaced by

ICMP Timestamp Request

Type = 13 or 14

• The TCP/IP protocol suite allows systems to connect to one another

over vast distances through multiple networks.
• Each of these individual networks provides clock synchronization in its
own way.
• As a result, hosts on different networks who are trying to
communicate using software that requires time synchronization
can sometimes encounter problems.
• The ICMP timestamp message type is designed to help alleviate this
problem.
• The ICMP timestamp request message allows a host to ask for the
current time according to the remote host.
• The remote host uses an ICMP timestamp reply message to respond
to the request.
22
 Clock synchronization and transit time
estimation Replaced by

ICMP Timestamp
Type = 13 or 14

• All ICMP timestamp reply messages contain the originate, receive and
transmit timestamps.
• Using these three timestamps, the host can estimate transit time across
the network by subtracting the originate time from the transit time.
• It is only an estimate however, as true transit time can vary widely based
on traffic and congestion on the network.
• The host that originated the timestamp request can also estimate the
local time on the remote computer.
• While ICMP timestamp messages provide a simple way to estimate time
on a remote host and total network transit time, this is not the best way
to obtain this information.
• Instead, more robust protocols such as Network Time Protocol (NTP)
at the upper layers of the TCP/IP protocol stack perform clock
synchronization in a more reliable manner.
23
 Information requests and reply message
formats
ICMP Information Request/Reply
Type = 15 or 16

Replaced by
• The ICMP information requests and reply
messages were originally intended to
allow a host to determine its network
number.
• This particular ICMP message type is
considered obsolete.
• Other protocols such as BOOTP and
Dynamic Host Configuration Protocol
(DHCP) are now used to allow hosts to
obtain their network numbers.

24
 Address Masks

ICMP Address Mask Request/Reply

Type = 17 or 18

• This new subnet mask is crucial in Replaced by

identifying network, subnet, and host bits in
an IP address.
• If a host does not know the subnet mask,
it may send an address mask request to the
local router.
• If the address of the router is known, this
request may be sent directly to the router.
• Otherwise, the request will be broadcast.
• When the router receives the request, it will
respond with an address mask reply.
• Somewhat obsolete, was used with
diskless workstations that used RARP for
the IP address and ICMP for the subnet
mask.
25
Router Solicitation and Advertisement

ICMP Router Solicitation

Type = 10

ICMP Router Advertisement

Type = 9
Replaced by
• When a host on the network boots, and the host
has not been manually configured with a
default gateway, it can learn of available
routers through the process of router discovery.
• This process begins with the host sending a
router solicitation message to all routers,
using the multicast address 224.0.0.2 as the
destination address. (May also be broadcast).
• When a router that supports the discovery
process receives the router discovery
message, a router advertisement is sent in
return.
• Routers may also periodically advertise router
advertisement messages.
26
IRDP

• Some newer IP hosts use ICMP Router Discovery Protocol

(IRDP) (RFC 1256 ) to find a new router when a route
becomes unavailable.
• A host that runs IRDP listens for hello multicast messages
from its configured router and uses an alternate router
when it no longer receives those hello messages.
• The default timer values of IRDP mean that it's not suitable
for detection of failure of the first hop.
• The default advertisement rate is once every 7 to 10
minutes, and the default lifetime is 30 minutes.

27
ICMP source-
quench messages
ICMP Source Quench
Type = 4

• Congestion can also occur for various reasons including when traffic
from a high speed LAN reaches a slower WAN connection.
• Dropped packets occur when there is too much congestion on a
network.
• ICMP source-quench messages are used to reduce the amount of data
lost.
• The source-quench message asks senders to reduce the rate at which
they are transmitting packets.
• In most cases, congestion will subside after a short period of time, and
the source will slowly increase the transmission rate as long as no
other source-quench messages are received.
• Most Cisco routers do not send source-quench messages by
default, because the source-quench message may itself add to the
network congestion. (See TCP)
28
ICMP source-
quench messages
ICMP Source Quench
Type = 4

• IP has no mechanism for flow control

• Some issues with ICMP Source Quench:
– A router or destination host (buffers full) will send one source-
quench message for each discarded packet.
– No mechanism to tell the source that the congestion has been
relieved and source can resume sending at previous rate.
• Remember, TCP/IP uses TCP mechanisms for flow control and
reliability including sliding windows.

29
 ICMP Path MTU Discovery

Information from:
Marc Slemko
Path MTU Discovery and Filtering ICMP
http://alive.znep.com/~marcs/mtu/
and
Cisco Systems
Path Maximum Transfer Unit (MTU) Discovery
http://www.cisco.com/en/US/products/sw/iosswrel/ios_abcs_ios_the_abcs_
ip_version_60900aecd800c1126.html
Path MTU
Discovery

Problem:
• How path MTU discovery (PMTU-D) combined with filtering ICMP
messages can result in connectivity problems.
• Path MTU discovery allows a node to dynamically discover and adjust
to differences in the MTU size of every link along a given data path.
• In IPv4, the minimum link MTU size is 68 octets and the recommended
minimum is 576 octets, which is the minimum reassembly buffer size.
• So, any IPv4 packet must be at least 68 octets in length.
• (In IPv6, the minimum link MTU is 1280 octets, but the recommended MTU value for
IPv6 links is 1500 octets. The maximum packet size supported by the basic IPv6 header
is 64,000 octets. Larger packets called jumbograms could be handled using a hop-by-
hop extension header option.)
31
Path MTU Discovery - Terms

• MTU: The maximum transmission unit is a link layer restriction on the

maximum number of bytes of data in a single transmission (ie. frame,
cell, packet, depending on the terminology).
– The table above shows some typical values for MTUs, taken from
RFC-1191.
• Path MTU : The smallest MTU of any link on the current path between
two hosts.
– This may change over time since the route between two hosts,
especially on the Internet, may change over time.
– It is not necessarily symmetric and can even vary for different types
of traffic from the same host.

32
Terms

• Fragmentation: When a packet is too large to be sent across a link as a single

unit, a router can fragment the packet.
– This means that it splits it into multiple parts which contain enough
information for the receiver to glue them together again.
– Note that this is not done on a hop-by-hop basis, but once fragmented a
packet will not be put back together until it reaches its destination.
– Fragmentation is undesirable for numerous reasons, including:
• If any one fragment from a packet is dropped, the entire packet needs
to be retransmitted. This is a very significant problem.
• It imposes extra processing load on the routers that have to split the
packets.
• In some configuration, simpler firewalls will block all fragments
because they don't contain the header information for a higher layer
protocol (eg. TCP) needed for filtering.
33
Terms
3 4

ICMP Destination Unreachable

Fragmentation needed, but DF Set

• DF (Don't Fragment) bit: This is a bit in the IP header that can be set to
indicate that the packet should not be fragmented by routers.
– If the packet needs to be fragmented, an ICMP "can't fragment" error is
returned sent to the sender and the packet is dropped.
• ICMP Can't Fragment Error:
– This error is a type 3 (destination unreachable), code 4 (fragmentation
needed but don't-fragment bit set)
– Returned by a router when it receives a packet that is too large for it to
forward and the DF bit is set.
– The packet is dropped and the ICMP error is sent back to the origin host.
– Normally, this tells the origin host that it needs to reduce the size of its
packets if it wants to get through.
– Recent systems also include the MTU of the next hop in the ICMP
message so the source knows how big its packets can be.
– Note that this error is only sent if the DF bit is set; otherwise, packets are
just fragmented and passed through.
34
Terms

• MSS: The MSS is the maximum segment size.

– It can be announced during the establishment of a TCP
connection to indicate to the other end the largest
amount of data in one packet that should be sent by the
remote system.
– MSS is beyond the scope of this discussion.

35
Path MTU Discovery (PMTU-D)

• Now you know that Path MTUs vary.

• You know that fragmentation is bad.
• The solution?
– Well, one solution is Path MTU Discovery.
– The idea behind it is to send packets that are as large
as possible while still avoiding fragmentation.

36
 PMTU-D

• A host does this by starting by sending packets that have a maximum

size of the lesser of the local MTU or the MSS announced by the remote
system.
• These packets are sent with the DF bit set.
• If there is some MTU between the two hosts which is too small to pass
the packet successfully, then an ICMP can't fragment error will be sent
back to the source.
• It will then know to lower the size; if the ICMP message includes the next
hop MTU, it can pick the correct size for that link immediately, otherwise
it has to guess.
37
PMTU-D

• The exact process that systems go through is somewhat more

complicated to account for special circumstances. See, RFC 1191.
• A good indication of if a system is trying to do PMTU-D is to watch the
packets it is sending with something like tcpdump or snoop and see if
they have the DF bit set; if so, it is most likely trying to do PMTU-D.
• Most Windows and Linux/Unix OS’s default to using PMTU-D.
• Adjusting IP MTU, TCP MSS, and PMTUD on Windows and Sun
Systems - http://www.cisco.com/warp/public/105/38.shtml
38
The problem with ICMP filtering and PMTU-D

• Many network administrators have decided to filter ICMP at a router or

firewall.
• There are valid (and many invalid) reasons for doing this, however it
can cause problems.
• ICMP is an integral part of the Internet and can not be filtered without
due consideration for the effects.
• In this case, if the ICMP can't fragment errors can not get back to the
source host due to a filter, the host will never know that the packets it is
sending are too large.
• This means it will keep trying to send the same large packet, and it will
keep being dropped--silently dropped from the view of any system on
the other side of the filter.
• While a small handful of systems that implement PMTU-D also
implement a way to detect such situations, most don't and even for
those that do it has a negative impact on performance and the network.

39
The Symptoms

• If this is happening, typical symptoms include the ability for

small packets (eg. request a very small web page) to get
through, but larger ones (eg. a large web page) will simply
hang.
• This situation can be confusing to the novice administrator
because they obviously have some connectivity to the
host, but it just stops working for no obvious reason on
certain transfers.

40
The Fix

• There is one solution, and several workarounds, for this problem.

• The Fix:

• Fix your filters!

– The real problem here is filtering ICMP messages without
understanding the consequences.
– Many packet filters will allow you to setup filters to only allow
certain types of ICMP messages through.
– If you reconfigure them to let ICMP can't fragment (type 3, code 4)
messages through, the problem should disappear.
– If the filter is somewhere between you and the other end, contact
the administrator of that machine and try to convince them to fix the
problem.
• We will learn how to do this on Router Access Control Lists (ACLs)

41
Recommended Reading

Where Wizards Stay Up Late

TCP/IP Illustrated, Vol. 1 W. Katie Hafner and Matthew Lyon
Richard Stevens Addison-Wesley ISBN 0613181530
Pub Co ISBN: 0201633469
 Very enjoyable reading and you
• Although, published in 1994, written by the do not have to be a networking
late Richard Stevens, it is still regarded as geek to enjoy it!
the definitive book on TCP/IP.  National Bestseller
42