Windows Server Troubleshooting - Ethernet

Click here to start saving with ING DIRECT!

Home | Up | Methodology | Architecture | Tools | Memory | Processor | Registry | File System | Active Directory | Contents

Get the Book

Major Topics
Methodology
Architecture
Tools
Memory
Processor
Registry
File System
Network
Active Directory
Contents
Other Topics
Counters
Browser
Bindings
Network Monitor
Commands
Name Resolution
Ethernet
Protocols
More Detail

eXpert Genealogy

Memory from Crucial.com


2003-2006 Team Approach Limited
All rights reserved


Common Ethernet Problems

  • Cable faults such as broken cables or connectors
  • Segments that are too long result in signals that are too weak to decode
  • Too many hubs cause collision detection failure because of excessive propagation delays
  • Excessive traffic cause excessive Ethernet collisions and slow throughput
  • Bad FCS checksum usually indicates a bad NIC
  • Jabbers are packets longer than 1518 bytes
  • Runts are packets shorter than 64 bytes
  • Mismatched frame types

Speed and Duplex Mismatches

Normally NICs and switch/hub ports autoSense speed and duplex settings. If the autoSense does not work properly, the resulting mismatch will result in either extremely slow or no communications. To correct this problem you can manually set the speed and duplex settings. Both the NIC and the switch/hub port must have the same settings. The following dialog shows the NIC properties where speed and duplex can be set.

NICs support different speed settings to be compatible with old and new faster switch ports. Classical Ethernet is half-duplex which means that the NIC cannot send and receive data at the same time. Newer switches support full-duplex operation which supports sending and receiving data at the same time.

                   
Half-Duplex hub                                                 Full-Duplex switch

Frame Types

Ethernet sends octet bytes on a LAN local area network in a grouping called an Ethernet frame. Some software supports organizing the frame in different formats. This may lead to communications problems if a client and server use a different frame format. Consistency is important for successful communications. This is rarely a problem for TCP/IP which normally used only the original DIX Ethernet frame.

Original DIX Dec/Intel/Xerox Ethernet frame

6 bytes 6 2 46-1500 4
Destination Source Type Data FCS

IEEE 802.3 Standard frame

6 bytes 6 2 46-1500 4
Destination Source Length Data FCS

IEEE 802.2 Standard frame

6 bytes 6 2 46-1500 of Data 4
Destination Source Length
DSAP SSAP
FCS

SNAP Subnetwork Access Protocol frame

6 bytes 6 2 46-1500 of Data 4
Destination Source Length
DSAP SSAP + EtherType
FCS

The FCS Frame Check Sequence is a checksum calculated from data in the rest of the frame. The receiver verifies the checksum and discards it if the checksum does not match the expected value. Invalid checksums occur when data becomes corrupted by problems such as electrical interference.

The Length field identifies the number of bytes 46-1500 in the Data field. The Type/EtherType field is a coded value that indicates the Transport layer protocol within the Data field. Examples are IP or IPX. The numeric Type values are greater than 1500. The value of the Type/Length field identifies the frame format. Values greater than 1500 identify the IEEE 802.3 frame type and lower values identify the original DIX Ethernet frame type.

The IEEE Institute of Electrical and Electronics Engineers sets international standards in data communications. The DSAP and SSAP fields are the IEEE replacement for the EtherType field. The SNAP frame has both the EtherType and the IEEE DSAP and SSAP fields so that it is compatible with software that needs either standard.