CompTIA’s product-neutral Network+ exam tests your skills outside the Microsoft mold. Here’s your guide to preparation.
Network Detailing
CompTIA’s product-neutral Network+ exam tests your skills outside the Microsoft mold. Here’s your guide to preparation.
- By Greg Neilson
- 11/01/2000
Whether you’re an experienced network professional or
somebody new to the IT biz, the Network+ test from CompTIA,
released last year, offers an excellent jumpstart to your
certification efforts. CompTIA is a high-tech vendor consortium
that offers non-product-specific exams. The organization
cut its teeth on the A+ test, which is aimed at proving
the expertise of help desk technicians. Network+ does
the same thing for those in the networking profession.
And some vendors—including Microsoft—consider it a great
foundation for continuing onto premium certifications,
such as the MCSE. I recently tried the beta version of
this exam. Why a beta version, you ask? By testing new
questions, CompTIA can seed its live exam with new items
or come out with a new version altogether.
The certification is intended to test the working knowledge
of a networking professional with 18 to 24 months of experience.
As well as some basic theory—the OSI reference model layers
and the 802.x standards—much of the exam is practical
in nature. You’ll need to be conversant in all aspects
of TCP/IP: addressing, subnetting, standard ports, configuration,
standard applications, and also trou-bleshooting tools.
In the article I highlight the main areas you need to
master in your preparation for the Network+ test.
CompTIA
Network+ (NK-N10-001) |
Reviewer’s
Rating: “You still need to know
your stuff, but this is a very practical
exam on network fundamentals.”
Title:
Network+
Current Status:
Already live; version reviewed was beta
for the purpose of substituting new
questions in the current exam.
Number of Questions:
168 questions for beta; the regular
exam has 65 questions.
Time Allowed:
Three hours for beta: the regular exam
is 90 minutes.
Who should take
it? Anyone interested in obtaining
a non-vendor networking certification.
Some vendors, such as Novell, accept
the exam in place of other company-
specific tests.
What course prepares
you? No CompTIA classes available,
but various training centers and programs
offer their own classes.
|
|
|
Knowledge of Networking Technology
How well do you understand what each of the seven layers
of the OSI Reference model does? Let’s review them.
- Physical. Defines the
electrical characteristics of network communication.
- Data Link. Defines the
standard frame layouts.
- Network. Breaks the
network up into logical network segments, through which
data can be routed to reach other segments.
- Transport. Ensures
the reliability of data; it handles flow control, acknowledgments,
and sequence numbering.
- Session. Establishes
and maintains the dialog between two communicating systems.
RPCs reside here.
- Presentation. Works
with the data formats, for example, character-set conversions,
compression, expansion, and encryption/decryption.
- Application. The interface
with the user.
One other theoretical area you need some understanding
of is the Institute of Electrical and Electronic Engineers
(IEEE) 802.x specifications, which covers the Physical
and Data Link layers. There are many of these that cover
things such as voice/data integration and wireless networking,
but the three most important are:
- 802.2, Logical Link
Control (LLC), which subdivides the Data Link layer
in two: LLC and the lower MAC (Media Access Control)
sub-layer. This means that the LLC functionality can
be shared regardless of the underlying topology such
as Ethernet or Token-Ring.
- 802.3, Ethernet.
- 802.5, Token-Ring.
Tip: As well as these layers,
a network professional should understand the options available
for network connectivity and the layers of the model these
operate within. For example, a bridge operates at the
data link layer whereas a router operates at the network
layer. A brouter is a combination bridge and router; it
can operate at both the data link and network layers.
Another technical topic we all hold dear to our hearts:
the different classes of IP addressing and the default
subnet mask each has. Table 1 shows the details you should
know. The subnet mask tells you which parts of the IP
address in the x.x.x.x dotted decimal notation represent
the network address and which represent the node address.
The rule is that when the subnet is converted to binary,
the network address is in ones and the node address is
in zeros. So you can see that a class A address uses the
first octet, or byte, for the network address, with the
remainder being the node address. Similarly, a class C
address uses three bytes for the network address and leaves
one byte for the node address.
Table 1. Understanding TCP/IP address ranges
is a must-know for the network pro. |
Class of address |
Address range for first octet |
Subnet mask |
A |
0 - 127 |
255.0.0.0 |
B |
128 - 191 |
255.255.0.0 |
C |
192 - 223 |
255.255.255.0 |
The TCP/IP protocol suite has a number of standard ports
assigned to applications, many of which you should know.
A common problem that can occur in practice is that these
ports are blocked by a router, preventing a user from
getting to a given application. Table 2 shows the main
ports you need to be aware of.
Table 2. Ports used by the main TCP/IP applications.
|
Port number |
Application name |
20 |
FTP (data) |
21 |
FTP |
23 |
Telnet |
Also make sure you know how to work with the main TCP/IP
utilities: ARP, NBTSTAT, TRACERT, NETSTAT, IPCONFIG/WINIPCFG,
and PING. Spend time reviewing what these commands do,
when to use them, and the purpose for the various optional
parameters for each command.
Tip: It can be useful sometimes
to continuously PING a host, since the default PING behavior
is to only send four packets and stop and report the results.
I use this often when performing remote reboots of servers
since this lets me know when the server has actually finished
shutdown (it stops responding to PING) and when the operating
system has started again (it starts responding again to
a PING). In order to do this, you use the -t option with
PING. In this case, you use Ctrl-C to stop sending PINGs.
Given the practical nature of the exam, you need to understand
how you actually configure TCP/IP on a client and what
each of the main configuration parameters actually means—the
IP address, the subnet mask, and the default gateway.
For example, the client sends to the default gateway all
packets not on the same network for which it doesn’t have
an explicit route already configured.
You also need to know about other common network protocols:
NetBEUI, IPX/SPX, and Appletalk. For example, NetBEUI
uses names to identify resources on the network, and it
can’t be routed—it can only be bridged. In a manner analogous
to DNS and HOSTS files, WINS and the LMHOSTS files map
NetBIOS names to IP addresses.
You’ll need a working knowledge of Windows NT and Novell
NetWare concepts. You aren’t expected to be an expert,
but you should understand the way these operate on a network.
For example, NT 4.0 can use TCP/IP, IPX (which it calls
NWLink), and NetBEUI. IPX is traditionally associated
with NetWare, but now NetWare 5.x can use TCP/IP natively.
IPX can be routed, and each network segment is given an
IPX address. Servers have a unique internal IPX address,
and clients have an individual station address created
by the combination of the burnt-in MAC address from their
network adaptor and the network IPX address.
After I took the exam, I realized that there were a number
of areas in the exam objectives that were covered lightly,
if at all. However, these may well be part of the exam
you take so you can’t afford to ignore them. These include
an understanding of star, bush, mesh, and ring topologies
and also high-availability/fault-tolerance options such
as mirroring, duplexing, striping, volume (sets), and
tape backup.
Knowledge of Networking Practices
How good are you at troubleshooting? If you were presented
with a short scenario and then asked what steps should
be taken next, would you feel confident in your choices?
Much of troubleshooting is common sense—you eliminate
the possibilities in a logical fashion. For example, if
a user has a network problem, you need to determine whether
other users have the same problem or whether it’s limited
to that user, or perhaps limited to users on the network
segment.
If you read carefully through the exam objectives for
the Troubleshooting the Network section, it guides you
through the processes the test creators want you to follow
when answering the questions. For some of us who have
been doing this for a while, this will be old news, but
if you’re just starting out, listen up. Learning how to
find the source of a problem systematically could be a
more useful professional skill than any of the technical
content in the exam. For example, I’ll quote objective
11.3.3 to illustrate this:
Identify the following steps as a systematic approach
to determining whether a problem is attributable to
the operator or the system, and given a problem scenario,
select the appropriate next step based on this approach.
- Have a second operator perform the same task on
an equivalent workstation.
- Have a second operator perform the same task on
the original operator’s workstation.
- See whether operators are following standard operating
procedure. Use this approach when asked in the exam
which problem-solving step should come next in the
situation presented.
Tip: Check through the exam
objectives for the list of hardware that you’re supposed
to be able to recognize. You may have seen many of these
already in your work as a network professional, but some
may be pieces of equipment you’re not familiar with.
There are also very specific details you should know—I’m
assuming those who put the exam together found these all
too common in practice. These are listed in the exam objectives
in section 11.1.4. For example, you need to know that
plugging an analog modem in a digital jack in the wall
will damage the modem, so you need to check first. Similarly,
you need to understand that the length of patch cables
contributes to the length of the cabling segment. That
is, in a typical 100-meter maximum lobe length, to meet
specifications we might need to include the 10-meter length
of patch cable—meaning that the maximum length of cable
that can be layed in this case is actually 90m, not the
100m length that we assume can be used.
A Promising Pursuit
I’ve taken a number of certification exams and found
that this was probably among the most practical and useful.
The emphasis isn’t on memorizing meaningless details but
in ensuring you really have a practical understanding
of the material and can actually do the job.
Here are three tips for taking this exam. Follow them
and you’ll pass with flying colors:
- Focus on the exam objectives. Not only will this
guide you through your preparation and ensure you have
everything covered, but—as we’ve seen already—they prescribe
the methods to use to solve problems within the exam.
- You will need to be very conversant with TCP/IP configuration
and troubleshooting to pass this exam.
- Don’t try to just memorize this stuff—ensure you
really understand the content. Much of it you’ll come
across regularly in your day-to-day work. Get plenty
of hands-on experience where appropriate. (Some of the
better study guides have practice exercises to ensure
you’re familiar with the material.)
Industry certification that cuts across the technologies
of multiple vendors is still in its infancy. Yet, if this
is a guide, it offers us the promise of valuable product-independent
certification that will be of great use for employers
and participants alike. Good luck!