Mastering

Network
Protocols

an Networking eBook
 Mastering Network Protocols

S ometimes the simplest elements of modern networking are the easiest to forget, whether you're talking about
calculating a subnet or understanding performance bottlenecks at the protocol level.

This guide is written with two audiences in mind: Network technicians who could use a refresher course on con-
cepts they haven't needed since they crammed for their first certification, and IT managers with a background in
disciplines outside networking who could use a cheat-sheet when they're dealing with this involved and complex
field.

Whether we're covering new ground or introducing new concepts, we hope you find helpful this look at the founda-
tion of networking.

Understanding IP Addresses
Networks don't work without addresses: Whenever you are sending something, you need to specify where it should
go and where it came from. To be an effective network engineer or administrator, you need to understand IP
addresses backward and forward: you need to be able to think on your feet. If something breaks, likely as not some
address assignment has been screwed up. And spotting the problem quickly could be the difference between being
the hero, or the person who "takes a long time to fix the problem." Thoroughly understanding IP addresses in their
primal form is crucial to understanding subnets.

IPv4 Addresses and 32-bit Numbers

IP addresses are just 32-bit binary numbers, but they're important binary numbers: you need to how to work with
them. When working with subnet masks, new network administrators generally get confused with the ones they
haven't memorized. All the subnet mask amounts to is moving the boundary between the part of the address that
represents a "network" and the part that represents a "host." Once you're comfortable with this method of thinking
about IP addresses and masks, you've mastered IP addressing.

Binary is quite simple. In binary the only numbers are zeros and ones, and a 32-bit number holds 32 zeros and
ones. We're all used to base-10 numbers, where each place in a number can hold any number from 0-9. In binary
each place holds either a zero or a one. Here's the address 255.255.255.0 in binary:

11111111.11111111.11111111.00000000

For convenience, network engineers typically break IP addresses into four 8-bit blocks, or octets. In an 8-bit num-
ber, if all of the bits are set to 1, then the number is equal to 255. In the previous address, 11111111 represents
255 and 00000000 represents zero.

The way binary really works is based on powers of two. Each bit represents a different power-of-two. Starting at the
left-hand side, the most significant bit, numbers form in the following manner:

Power-of-two 232 ... 27 26 25 24 23 22 21 20

Decimal 4,294,967,296 ... 128 64 32 16 8 4 2 1

The result is additive, meaning that if all bits are set, you simply add the power-of-two value up for each place. For
example, if we have an 8-bit number, 11111111 , we simply add: 27 + 26 + 25 + 24 + 23 + 22 + 21 + 20 = 255

Try a non-trivial example now: 11110000

We can see that four bits are "set" in the above 8-bit number. Summing the power-of-two values in those places
yields: 27 + 26 + 25 + 24 = 240

1 Mastering Network Protocols, an Internet.com Networking eBook. Copyright 2006, Jupitermedia Corp.
 Mastering Network Protocols

It's just that simple. If you can convert a binary number to

decimal form, you can easily figure out subnet masks and
network addresses. The International Standards Organization (ISO) devel-
oped the OSI (Open Systems Interconnection) model. It
Focusing on 32-bit IPv4 addresses themselves now, there are divides network communication into seven layers.
a few different types that need to be understood. All IP Layers 1-4 are considered the lower layers, and most-
addresses can be in the range 0.0.0.0 to 255.255.255.255, ly concern themselves with moving data around.
but some have special uses. Layers 5-7, the upper layers, contain application-level
data. Networks operate on one basic principle: "pass
Loopback it on." Each layer takes care of a very specific job, and
Packets that will not leave the host (i.e., they will not traverse then passes the data onto the next layer.
an external network interface). Example: 127.0.0.1
Layer one is simply wiring, fiber, network cards, and
Unicast anything else that is used to make two network
Packets that are destined for a single IP address. Example: devices communicate. Even a carrier pigeon would be
2.2.2.2 considered layer one. Network troubleshooting will
often lead to a layer one issue.
Multicast
Packets that will be duplicated by the router, and eventually Layer two is Ethernet, among other protocols. The
routed by multicast routing mechanisms. Example: 226.0.0.2 most important take-away from layer 2 is that you
Limited Broadcast should understand what a bridge is. Switches, as
they're called nowadays, are bridges. They all operate
A broadcast packet, sent to every host, limited to the local at layer 2, paying attention only to MAC addresses on
subnet. Example: 255.255.255.255 Ethernet networks. If you're talking about MAC
address, switches, or network cards and drivers,
Directed Broadcast you're in the land of layer 2. Hubs live in layer 1 land.
Packets that are routed to a specific subnet, and then broad-
cast. Example, assuming we are not on this subnet: If you're talking about an IP address, you're dealing
1.1.1.255 with layer 3 and "packets" instead of layer 2's
"frames." IP is part of layer 3, along with some rout-
There are also some special cases of IP addresses, including ing protocols, and ARP (Address Resolution Protocol).
private and multicast addresses. Addresses in the range
224.0.0.0 to 239.255.255.255 are reserved for multicast. Layer four, the transport layer, handles messaging.
Everything below that range is fair game on the Internet, Layer 4 data units are also called packets, but when
excluding addresses reserved by RFC 1918 and a few other you're talking about specific protocols, like TCP,
special-purpose assignments. These "1918 addresses" are they're "segments" or "datagrams" in UDP.
private addresses, meaning Internet routers will not route
them. The ranges include: And arriving at layer 7, we wonder what happened to
layer 5 and 6. In short: They're useless.
• 10.0.0.0 to 10.255.255.255
• 172.16.0.0 to 172.31.255.255 A few applications and protocols live there, but for
• 192.168.0.0 to 192.168.255.255 understanding networking issues talking about these
provides zero benefit. Layer 7, our friend, is "every-
These IP addresses can be assigned locally to as many com- thing." Dubbed the "Application Layer," layer 7 is
puters as you want, but before those computers access the application-specific. If your program needs a specific
Internet, the addresses must be translated to a globally format for data, you invent some format that you
routable address. This is commonly done via Network expect the data to arrive in and you've just created a
Address Translation (NAT). The 1918 addresses aren't the layer 7 protocol. SMTP, DNS and FTP are all layer 7
only reserved spaces, but they are defined to be "site local." protocols.
Multicast also has a reserved range of addresses that aren't – Charlie Schluting, Enterprise Networking Planet
designed to escape onto the Internet: 224.0.0.0 to
224.0.0.255 are multicast "link-local" addresses.

2 Mastering Network Protocols, an Internet.com Networking eBook. Copyright 2006, Jupitermedia Corp.
 Mastering Network Protocols

Let's review the concept of a local subnet. Once we have assigned a valid IP address to a computer, it will be able
to speak to the local network, assuming the subnet mask is configured properly. The subnet mask tells the operat-
ing system which IP addresses are on the local subnet and which are not.

If an IP we wish to talk to is located on the local subnet, then the operating system can speak directly to it without
using the router. In other words, it can ARP for the machine, and just start talking. IP address and subnet mask con-
figuration is fairly straightforward for general /24 networks. The standard 255.255.255.0 mask means that the first
three octets are the network address, and the last part is reserved for hosts. For example, a computer assigned the
IP of 10.0.0.1 and a mask of 255.255.255.0 (a /24, or 24-bits if you write it out in binary) can talk to anyone inside
the range 10.0.0.1 to 10.0.0.255.

Understanding Subnets and CIDR

Let's get one thing straight: there is no "Class" in subnetting. In the olden days, there were Class A, B, and C net-
works. These could only be divided up into equal parts, so VLSM (Variable Length Subnet Masks) were introduced.
The old Class C was a /24, B was a /16, and A was a /8. That's all you need to know about Classes. They don't
exist anymore.

An IP address consists of a host and a network portion. Coupled with a subnet mask, you can determine which part
is the subnet, how large the network is, and where the network begins. Operating systems need to know this infor-
mation in order to determine which IP addresses are on the local subnet and which addresses belong to the outside
world and require a router to reach. Neighboring routers also need to know how large the subnet is, so they can
send only applicable traffic that direction. Divisions between host and network portions of an address are completely
determined by the subnet mask.

Classless Internet Domain Routing (CIDR), pronounced "cider," represents addresses using the network/mask style.
What this really means is that an IP address/mask combo tells you a lot of information:

network part / host part

0000000000000000/0000000000000000

The above string of 32 bits represents a /16 network, since 16 bits are masked.

Throughout these examples (and in the real world), certain subnet masks are referred to repeatedly. They are not
special in any way; subnetting is a simple string of 32 bits, masked by any number of bits. It is, however, helpful for
memorizing and visualizing things to start with a commonly used netmask, like the /24, and work from there.

Let's take a look at a standard subnetting table, with a little bit different information:

Subnet Number of Number of Bits

mask bits /24 subnets addresses stolen
/24 1 256 0
/25 2 128 1
/26 4 64 2
/27 8 32 3
/28 16 16 4
/29 32 8 5
/30 64 4 6
/31 128 2 7

3 Mastering Network Protocols, an Internet.com Networking eBook. Copyright 2006, Jupitermedia Corp.
 Mastering Network Protocols

Because of the wonders of binary, it works out that a /31 has two IP addresses available. Imagine the subnet:
2.2.2.0/31. If we picture that in binary, it looks like:

00000010.00000010.00000010.00000000 (2.2.2.0)

11111111.11111111.11111111.11111110 (31)

The mask is "masking" the used bits, meaning that the bits are used up for network identification. The number of
host bits available for tweaking is equal to one. It can be a 0 or a 1. This results in two available IP addresses, just
like the table shows. Also, for each additional bit used in the netmask (stolen from the network portion), you can see
that the number of available addresses gets cut in half.

Let's figure out the broadcast address, network address, and netmask for 192.168.0.200/26. The netmask is sim-
ple: that's 255.255.255.192 (26 bits of mask means 6 bits for hosts, 2^6 is 64, and 255-64 is 192). You can find
subnetting tables online that will list all of this information for you, but we're more interested in teaching people how
to understand what's happening. The netmask tells you immediately that the only part of the address we need to
worry about is the last byte: the broadcast address and network address will both start with 192.168.0.

Figuring out the last byte is a lot like subnetting a /24 network, but you don't even need to think about that if it
doesn't help you. Each /26 network has 64 hosts. The networks run from .0 to .64, .65 to .128, .128 to .192,
and from .192 to .256. Our address, 192.168.0.200/26, falls into the .192 to .256 netblock. So the network
address is 192.168.0.192/26. And the broadcast address is even simpler: 192 is 11000000 in binary. Take the last
six bits (the bits turned "off" by the netmask), turn them "on", and what do you get? 192.168.0.255. To see if you
got this right, compute the network address and broadcast address for 192.168.0.44/26. (Network address:
192.168.0.0/26; broadcast 192.168.0.63).

It can be hard to visualize these things at first, and it helps to start with making a table. If you calculated that you
wanted subnets with six hosts in each of them, (eight, including the network and broadcast address that can't be
used) then you can start making the table. The following is 2.2.2.0/29, 2.2.2.8/29, 2.2.2.16/29 and the final subnet
of 2.2.2.249/29.

Subnet Network First IP Last IP Broadcast

Number Address Address
1 2.2.2.0 2.2.2.1 2.2.2.6 2.2.2.7
2 2.2.2.8 2.2.2.9 2.2.2.14 2.2.2.15
3 2.2.2.16 2.2.2.17 2.2.2.22 2.2.2.23
32 2.2.2.249 2.2.2.250 2.2.2.254 2.2.2.255

In reality, you're much more likely to stumble upon a network where there's three /26's and the final /26 is divided
up into two /27's. Being able to create the above table mentally will make things much easier.

That's really all you need to know. It gets a little trickier with larger subnets in the /16 to /24 range, but the principal
is the same. It's 32 bits and a mask. Do, however, realize that there are certain restrictions governing the use of sub-
nets. We cannot allocate a /26 starting with 10.1.0.32. If we utter the IP/mask of 10.1.0.32/26 to most operating
systems, they will just assume we meant 10.1.0.0/26. This is because the /26 space requires 64 addresses, and
they must start at a natural bit boundary for the given mask. In the above table, what would 2.2.2.3/29 mean? It
means you meant to say 2.2.2.0/29.

Those tricky ones do demand a quick example. Remember how the number of IP addresses in a subnet gets

4 Mastering Network Protocols, an Internet.com Networking eBook. Copyright 2006, Jupitermedia Corp.
 Mastering Network Protocols

halved when you take another bit from the network side to create a larger mask? The same concept works in
reverse. If we have a /25 that holds 128 hosts, and steal a bit from the host (netmask) portion, we now have a /24
that holds 256. Google for a "subnet table" to see the relationship between netmasks and network sizes all at once.
If a /16 holds 65536 addresses, a /17 holds half as many, and a /15 holds twice as many. Practice, practice, prac-
tice. That's what it takes to understand how this works. Don't forget, you can always fall back to counting bits.

Understanding TCP, the Protocol

TCP is used everywhere, and understanding how TCP operates enables network and systems administrators to
properly troubleshoot network communication issues.

TCP is wonderfully complex, but don't worry: We aren't going to tell you to go read RFC 793. This is a gentle intro-
duction, or demystification, if you will. You'll be familiar with enough terminology, you'll understand the components
of the TCP header, and then we'll discuss "TCP in the wild," which will focus on examining some common issues
with TCP, including window scaling problems, congestion, and of course, the mechanics of a TCP connection.

We sometimes hear people call it "the TCP/IP protocol suite," which means that they're talking about layers 1-4 plus
7, similar to how we presented layers. TCP lives at layer 4, along with its unreliable friend UDP. TCP stands for
Transmission Control Protocol. When a packet is encapsulated, we have the IP header at layer 3, and immediately
following is the TCP header, which becomes the "data" for the IP header. TCP includes its own jargon, just like
everything else. There were Ethernet frames, IP datagrams, and now TCP segments. You can think of them all as
packets, but be sure to use the correct terms when communicating with others.

While trying to think of other things people say about TCP, it seemed apropos to spend some time explaining the
things people are trying to tell you. There's nothing worse than asking a guru a question, and getting a response like
"Well, it's end-to- end." If you knew TCP you'd know what this meant, but then you wouldn't have asked the ques-
tion in the first place. Let's see what we can do about that.

Yes, TCP is end-to-end. There is no concept of broadcast, or anything like it. To speak TCP with another computer,
you must be connected, like a telephone call, so each end is prepared to talk. "Stream delivery" is also another
phrase you'll hear. This simply means that TCP works with data streams, and out of order packets are OK. In fact,
TCP is even OK with lost or corrupted packets; it will eventually get them again.

More likely you'll be hearing a programmer talking about streams, referring to the fact that it's hard to tell when data
is actually going to be sent, and you can send unstructured data down a TCP stream. TCP can buffer things, in
weird ways that sometimes don't make sense, but neither programmers nor users need to worry about that.

Whenever a TCP packet is sent, an acknowledgment, or ACK, is returned. This is really the only way to provide a
reliable protocol: You must let the other side know if you have received things. Of course, people will want to
improve on an inefficient system like this. Enter "piggybacking ACKs" into the picture. People call TCP "full duplex"
because of piggybacking; because it lets both sides send data at the same time. This is accomplished by carrying
the ACK for previous packet received within the current packet, piggybacked. In terms of preserving network utiliza-
tion, this is much better than sending an entirely separate packet just to say "got it." Finally, there's the concept of a
cumulative ACK: ACKing more than one packet at a time, to say "I got all the others, including this one."

TCP is a stream, so there isn't really any other concept to worry about aside from a "connection." Maximum
Segment Size, or MSS, is negotiated at connection time, but almost always changes. The default MSS is 536,
which is 576 (the IP guaranteed minimum packet size) minus 20 bytes for the IP header and 20 bytes for the TCP
header. TCP tries to avoid causing IP-level fragmentation, so it will almost always start with 536.

The sexiest feature of TCP still remains the Sliding Window Protocol. The window is essentially the amount of un-

5 Mastering Network Protocols, an Internet.com Networking eBook. Copyright 2006, Jupitermedia Corp.
 Mastering Network Protocols

ACKed data that has been sent, and it can grow and shrink at will. This gets really interesting, and will be covered
next time.

The header of a TCP packet is 20 bytes, just like an IP's. Both IP and TCP headers can get larger, if options are
used. TCP does not include an IP address; it only needs to know about the port on which to connect. Don't let this
confuse you though; TCP keeps track of end-to-end connections in a state table that includes IP addresses and
ports. It's just that the header for TCP doesn't need the IP information, since it comes from IP.

It is easier to think of a packet as a stream, one byte after the next. Everyone always wants to show a table for the
header, but this can confuse matters more. The TCP header, starting with the first bit, is:

• Source port, 16 bits: my local TCP port that's used for this connection
• Destination port, 16 bits: The remote machine's TCP port that I'm talking to
• Sequence number, 32 bits: the number used to keep track of packet ordering
• Acknowledgment number, 32 bits: the previously received sequence number that we're ACKing
• Header length, 4 bits: the number of 32-bit words in the header. This is set to five, if no options are used
• Reserved, 6 bits: reserved for future use
• Flags, 6 bits total, each flag is one bit (on or off):
• URG: urgent field pointer
• ACK: this packet is (or includes) an ACK
• PSH: push function (not used)
• RST: reset, or terminate the connection
• SYN: synchronization packet, aka Start Connection
• FIN: final packet, start hang-up sequence
• Window size, 16 bits: begins with the ACK field that the receiving side will accept
• Checksum, 16 bits: a checksum of the TCP header and data
• Urgent pointer, 16 bits: an offset from the sequence number that points to the data following URG data
• Options: MSS, Window scale, and more. This is mostly the focus of our next installment on TCP.

Each side of the TCP connection uses the two pairs of IP address and Port to identify the connection, and sends
the data on to the application that is listening on the port.

Let's go ahead and take a look at TCP operational issues, now that we know a little about what TCP actually is.

We said that TCP gets "connected" before any data can be sent. To make that work, the side that initiates a TCP
connection will send a SYN (remember the Flags field) packet first. This is simply a packet with no data, and the
SYN flag turned on. If the other side wants to talk on the port it received the SYN on, it will send back a SYN+ACK:
SYN and ACK fields set, and the ACK number set to acknowledge the first packet. Then, to verify the receipt of the
SYN+ACK, the sender will send one final ACK. The SYN, SYN+ACK, ACK sequence is called the three-way hand-
shake. After that happens, the connection is established. The connection will remain active unless it times out or
until either side sends a FIN.

Closing a TCP connection can be done from either side, and requires that both sides send a FIN to close their
channel of communication. One side can close before the other, or they can both happen at the same time. When
one side sends a FIN, the other sends FIN+ACK, to start the close of its side, and to ACK the first FIN. The person
who sent the first FIN will then FIN+ACK the second FIN, and the other person knows that the connection is closed.
There is no way for the person who sent the first FIN to get an ACK back for that last ACK. You might want to
reread that now. The person that initially closed the connection enters the TIME_WAIT state; in case the other per-
son didn't really get the ACK and thinks the connection is still open. Typically, this lasts one to two minutes.
And we've come to our first problem. If someone, say an attacker, leaves half-open or half-closed connections on
your Web server, this could be bad news. Memory is used up with each connection, and opening thousands of

6 Mastering Network Protocols, an Internet.com Networking eBook. Copyright 2006, Jupitermedia Corp.
 Mastering Network Protocols

bogus TCP connections could bring a server to its knees. Of course, you can't really adjust the TCP timers without
effecting the proper operation of TCP. If you've ever heard of a TCP SYN attack, this is what it means. To prevent
this, most operating systems opt to limit the number of half-open connections, for example in Linux it's normally 256
by default.

Now, since we promised to talk about the everlasting flow control problems, let's get into windowing. TCP uses
"positive ACK with retransmission" to guarantee reliability. The sender will wait a certain amount of time, and if it
doesn't get back an ACK for the packet it sent, it retransmits it. There are many, many timers in TCP, by the way,
this is just another one. The concept of ACKs is important to flow control, because the TCP sliding window protocol
makes the ping-pong nature of ACKs efficient. If TCP were to send a packet and wait for every ACK, it would
essentially cut the throughput in half.

Ideally, we can send many packets at once, and then get back an ACK for all of them, probably piggybacked on
more data from the other side. But how do we know how much to send? Well, the TCP window size controls how
many packets can be held in the "sent but not ACKed" state. If the window is large, we can send large amounts of
packets without waiting for an ACK. On the surface, this doesn't look like flow control, but it certainly is.

The receiving side is the one that controls the window size. If it says zero, then the sender cannot send any more
data at all. If the window size is one, then we're back to the simple "send and wait for ACK" protocol. If the last win-
dow size was zero, the sender will send a probe to figure out when the window is open again. If the sender never
gets an ACK, it just keeps trying until a timer expires. Remember, the window size is a 16-bit field in the TCP head-
er. And if you want a window size (in bytes) larger than 16-bits will accommodate, there's also a TCP option "win-
dow scale" that allows it to be multiplied by the scale factor. Without an extremely large window size, TCP has no
hope of filling a gigabit link. You should now be better prepared to understand the gigabit-tuning article, too.

On the subject of TCP flow control, we can't neglect mention of the Nagle algorithm. What would happen if you had
a large TCP window over a telnet connection? You'd type a command, then wait and wait and wait for a response.
This is a major problem for real-time applications. Furthermore, telnet can add to congestion, since a 1-byte packet
will include 40 bytes of header. RFC 896 defines the Nagle algorithm to attempt to abolish tiny packets. The idea is
that we should give data a chance to pile up before sending, to be more efficient. It says that we can only have one
unacked small segment, and you can't send more data until you get an ACK. Telnet and interactive ssh connections
turn this off with the TCP_NODELAY socket option, so that you can get an immediate response when you press a
key.

This content was adapted from EarthWeb's EnterpriseNetworkingPlanet Web site and was written by Charlie
Schluting. Copyright 2006, Jupitermedia Corp.

JupiterWeb eBooks bring together the best in technical information, ideas and coverage of important IT trends
that help technology professionals build their knowledge and shape the future of their IT organizations. For more
information and resources on networking, visit any of our category-leading sites:

www.enterprisenetworkingplanet.com
www.instantmessagingplanet.com
www.opticallynetworked.com
www.practicallynetworked.com
www.voipplanet.com
www.wi-fiplanet.com
www.opennetworkstoday.com

For the latest live and on-demand Webcasts on networking, visit: www.jupiterwebcasts.com/networking/

7 Mastering Network Protocols, an Internet.com Networking eBook. Copyright 2006, Jupitermedia Corp.