CHAPTER 3

NETWORK TROUBLESHOOTING

Upon completing this chapter, you should be able to do the following:

l Describe how to diagnose and isolate problems with LANs. Describe how to troubleshoot network malfunctions. Explain how to test and evaluate the connection of networking system nodes. Explain how to troubleshoot communications line problems. l l l

With any network system, you should have a set of error procedures for personnel to follow to handle errors or malfunctions on the system. These error procedures are the steps to be taken when the system is not operating properly. They are different from the error-detection and diagnostic procedures used to isolate and correct transmission problems.

A complete set of diagnostic procedures is necessary for the system. The system procedures are used to isolate the problem to the system or subsystem level. Since the facilities of a network may not be in the local area, it is necessary to have a set of test software and equipment with replacement components available for diagnosing and correcting problems.

TROUBLESHOOTING LANS

As a communications specialist, more than likely you will be expected to know how to troubleshoot problems on LANs. As a troubleshooter, you must be able to identify a wide range of network problems relating to hardware (the data terminal equipment, the communications link, repeaters, gateways, and so on), software (network operating system, applications, and soon), and peopleware (the end user). It will be your job to identify, isolate, and resolve both the simple and complex problems.

DIAGNOSTIC TOOLS

Normally, a problem can be solved without too much difficulty with the help of diagnostic tools. The best diagnostic tool available is accurate documentation. This documentation should include:

l Workstation and server configurations

l All network related software and equipment

l Location and paths of all wiring

l Updated records of all equipment and configurations changes

With documentation in hand, along with the help of diagnostic software (a network management package or a LAN analyzer), and specialized diagnostic equipment, such as a datascope, a time domain reflectometer (TDR), or a breakout box, the job becomes routine.

Classifying the problems and taking the necessary actions to resolve them are an important part of your job as a troubleshooter. However, it is equally important to remember to log all problems according to your activity's procedures. This will identify recurring problems, provide information for long term solutions, and enhance your command's training program.

ISOLATING PROBLEMS

When isolating a problem, consider the three major areas we discussed earlier-the user, the software, and the hardware, usually in that order. The majority of all network-related problems are caused by the user's actions\operator errors. Users either do not understand how to operate their PC in a networking environment or they are unfamiliar with the application software package they are using. Most of the time you will find yourself responding to user problems and complaints. A user will call, saying such things as the following:

My terminal/PC is hung up, and I cannot get into the system.

My terminal/PC screen suddenly went blank.

My temninal/PC keeps coming up with the same error message.

My terminal/PC will not allow me to access the disk file.

My terminal/PC will not print.

It will be your job to determine if the problem is user, software, or hardware related. Whenever you receive a call about a problem, obtain as much information as possible about the person and the problem. Ask the user's name, phone number, the terminal/PC or node identification number, the nature of the problem, and what, if anything, occurred immediately preceding the problem. In addition, you should ask the user what application he or she was

trying to access or currently working with at the time the problem occured. Ask whether other users are experiencing the same or similar problem, did any error messages appear on the screen, and be sure to ask whether the PC was moved before the problem

occured. Sometimes moving hardware creates problems\the connector cable may not be seated properly.

Once you have received initial information about a problem, it should help you to categorize the problem. Keep in mind most problems are the result of inexperienced users/operators. Because so many different types of errors can occur, it would be impossible for us to list them all, along with the necessary steps to resolve them. However, based on

past experiences, we can provide you with some helpful hints and guidelines to follow. If the problem seems to be isolated to one user, it is probably user error; if the problem occurs with a group of users in a common geographic location, the problem is usually related to the cable; and if the problem is network wide, a close look at the network software is in order. Let us take a look at some of the more common problems that frequently occur and their solutions in connection with these three categories.

PO3 Frost has just called to report he cannot log on to the LAN. You begin solving the problem by asking some preliminary questions. You find this is the first time PO3 Frost has used the LAN and no one else in his area is experiencing any problems. At this point, you should be able to recognize the problem is more than

likely the result of an inexperienced operator (user error). The logical corrective action to take is to walk PO3 Frost through the proper log-on procedures and password security requirements. He follows your instructions and successfully logs onto the LAN. PO3 Frost should have been able to log onto the LAN by following the User's Guide on LAN operations. You might want to review the guide to make sure it is current and available to all users.

A few weeks later PO3 Frost calls again and reports he has been having intermittent problems while logged on to the LAN. Sometimes while he is saving or retrieving data, his machine locks up for no apparent reason. Again, no one in his immediate work area is experiencing problems. After obtaining all the pertinent information available, you believe the problem is faulty hardware. During the save and retrieve operations, a packet is generated and sent through the network interface card, onto the cable, and to its destination. So the two most logical components to check are the cable connections and the network interface card. The diagnostic tools to use are the time domain reflectometer (TDR) to check any breaks on the cable and the diagnostics that come with the interface card. PO3 Frost runs the card diagnostics at the terminal while you check the cable continuity. The network card passes the test, but the TDR detects a continuity break near his location. A LAN technician checks the connectors at the workstation and discovers one of the connectors has worked itself loose. After replacing the connector, the cable is tested again and passes. PO3 Frost logs on to the LAN and experiences no further problems. In this example we eliminated the cable itself because no other user on the cable segment was experiencing problems. Had there been other users also experiencing intermittent failures, then the cable would have immediately been our focal point of testing, since this is the commonality between the users.

You arrive at work Tuesday morning and find a stack of messages waiting for you from users experiencing problems while trying to access the word processing program on the LAN. The only thing these users have in common is they all use the same file server. Immediately you focus your attention on the network operating system and software. You call PO3 Door to ask her a few questions before you begin troubleshooting any further. You learn PO3 Door is able to access all application programs on the LAN with the exception of the word processing program. You immediately log on to the network management program and monitor the data traffic. You discover no user has used the word processing program since

Monday at 1600. The only person authorized to use the LAN after 1600 is PO1 Brush, who is the network administrator. You call PO1 Brush and ask if any changes were made to the word processing program since yesterday. PO1 Brush states he installed a new version of the word processing program on Monday around 2200 to eliminate any work disruptions. You ask him to check the security access to this new version. You find PO1 Brush inadvertently restricted all users from accessing the new version of the word processing program after he removed the old one. He makes the necessary access changes, and everybody is once again happy and able to use the new version.

As you can see, there is a pattern to the various types of errors/problems you will encounter. The problems you will be confronted with will range from the simple to the disastrous. They may be user/operator errors, software problems, or hardware malfunctions. Knowing which is sometimes easy. Under other conditions, it may be difficult for you to determine the source of the problem. The important thing is to learn from your past experiences. Keep a list of symptoms, probable causes, and ways you can use to trace a problem to its cause. This will assist you in diagnosing and troubleshooting problems. You will also find users have a tendency to make the same mistakes again and again, especially while they are learning. You can provide them a great service by explaining some of the more common problems they are likely to encounter, the reasons for the problems, and ways to avoid having them happen to them.