|
| |
Troubleshooting
computer networks requires more than the basic scientific approach.
Is troubleshooting easy or difficult? When we do
troubleshooting, we are trying to determine why something is not working as
intended. Often the troubleshooting is unexpected. Many writers on the topic of
troubleshooting provide the scientific method as the solution to problems. The
flowchart is simple.
Unfortunately, this solution is not very
satisfying. Many people follow these steps instinctively as a common sense
approach to solving problems. Internet Genealogy
Computer network troubleshooting is unique because
of the extensive number of complex components. Because of this difficulty, we
need a more detailed and instructive approach to troubleshooting.
The following flowchart shows the basic approach
presented in this book. The main idea is to divide the big problem into
smaller problems until a solution is discovered. Many of the steps may not be
understood at this time, but will be explained in the appropriate chapter.
|
Identify the problematic
network node |
|
Use commands such as
PING & TraceRt |
|
|
|
ê |
|
|
|
|
|
Is there a problem with
one of the network protocols? |
è |
Isolate the problem to a
protocol layer and fix it |
|
|
|
ê |
|
|
|
|
|
Is there a memory
problem? |
è |
Is there a memory leak? |
è |
Fix or eliminate the
software with the memory leak |
|
ê |
|
ê |
|
|
|
ê |
|
Is there sufficient
memory? |
è |
Add more memory |
|
|
|
|
|
|
|
Does the system freeze? |
è |
Investigate priority and
device driver problems |
|
|
|
ê |
|
|
|
|
|
Is there high processor
utilization? |
è |
Is it caused by
hardware
or
software? |
è |
Provide adequate
processor resources |
|
ê |
|
ê
hardware
|
|
|
|
ê |
|
Can an upgraded device
driver fix the problem? |
è |
Upgrade you hardware to
offload the processor |
|
|
|
|
|
|
|
Is there a disk problem? |
è |
Is there sufficient file
cache? |
è |
Add more memory to
ensure sufficient cache |
|
ê |
|
ê |
|
|
|
ê |
|
Use NTFS and do regular
maintenance |
è |
Use RAID |
|
ê |
|
ê |
|
|
|
ê |
|
Is there a boot record
problem? |
è |
Use FixBoot or FixMBR
from the recovery console |
|
|
|
|
|
|
|
Is there a network
problem? |
è |
Use Network Monitor to
identify top broadcasters |
è |
Eliminate unnecessary
broadcasts |
|
|
|
ê |
|
|
|
|
|
Use Network Monitor to
identify top talkers |
è |
Eliminate unnecessary
network traffic |
|
|
|
ê |
|
|
|
|
|
Correct poor
configuration |
è |
Reorganize & upgrade
network for more capacity |
|
|
|
ê |
|
|
|
|
|
Is there a address or
name resolution problem? |
è |
Examine ARP cache, WINS,
DNS, and NBTstats |
What is the worst job in the world?
Windows troubleshooting
is not the worst job in the world, but it is challenging! What makes it so
difficult is that there are so many components. Windows 2000 is said to have
more than 40 million lines of code. If something goes wrong, then where is the
fault? What makes it worst is that this operating system runs on hundreds of
different types of computer, different types of components, cables, hubs,
switches, routers, firewalls, network printers, SAN Storage Area Networks, and
other components. We not only have Windows Servers, but Windows clients; Windows
Professional, NT Workstation, Windows 95, 98, ME, maybe even Windows for
Workgroups and DOS. We might even have Linux, Unix, RISC workstation,
minicomputers, and mainframes. Potentially, thousands of different application
programs run in this environment.
Every network is unique. No one in the world
manages a network like yours. Even if the equipment is the same, the people, our
customers, make it different. Don't be discouraged, a methodical
approach can solve all system problems. You know the solution, it just hasn't
been revealed to you yet. This book will help you to develop such
an approach and it will teach you about the many troubleshooting tools available
in Windows.
Troubleshooting and Optimization
Troubleshooting can be divided into two
categories;
- Troubleshooting failures
- Optimization = troubleshooting poor
performance
What do you need to understand to troubleshoot Windows?
In this book we will cover each of the
following major topics.
-
The Windows architecture defines the
components that we troubleshoot.
-
Microsoft includes many
troubleshooting tools with Windows.
-
You must monitor the key system resources;
processor, memory, and I/O.
-
Repairing system configuration is
centralized in the Registry database.
This document has information on the important software tools used to troubleshoot
Windows in a networked environment. In addition to the tools, there is
background information on the key resources that must be investigated;
Processor
-
Memory
Disk and Network I/O
In Windows troubleshooting, we need to focus on;
Inadequate
hardware, e.g.. insufficient memory, slow processor, etc.
- We need to
determine which component is a bottleneck
Faulty device drivers, e.g. version 1.0 often does not work
properly
-
Poor configuration, e.g. excessive replication between
Domain Controllers
Inefficient applications, e.g. programs that read more data than necessary.
Gradual growth usually leads to system overload. You need to
monitor your system to avoid problems. Your main monitoring tool is the
System Monitor.
It could be worse!
Troubleshooting the current Windows environment
is easier that troubleshooting older operating system. Windows has many
automatic features that eliminates some of the traditional troubleshooting hot
spots.
- Windows automatically detects plug-and-play devices and
installs and configures the correct device driver without administrative
intervention. You can avoid troubleshooting devices by using plug-and-play
hardware.
- Core services in Windows are self-tuning. There is no need to
configure disk cache size, memory pool size, etc.
| |
|