Glossary of Networking Terms

 

Access Glossary
Alignment Errors
Auto Partitioning
Broadcast Frames
Buffer Full Errors
Carrier Sence
Carrier Sence Loss Errors
Number of times lost
Collisions
Continual during transmit
Collision Rate
Cyclic Redundancy Check
Cyclic Redundany Check(CRC) Errors
(# of packets w/ an incorrect CRC)
Data Rate Mismatch Errors
Excessive Collision Errors
Frame
Frame Check Sequence(FCS) Errors
Framing Errors
(# of frames lost)
Framing Length Errors
Frame Too Long Errors
Giant Frame Errors
Heartbeat Errors
HST - Host Address
If - Interface #
Input Discards (ID)
Input Discard Errors
Input Drop Errors
Input Errors
(# of errors)
Input FIFO Overruns Errors
Input Overrun Errors
Input Packets
Input Queue (buffer) Full
Input Traffic (IKPk)
Counted in Megabytes
Inter Packet Gap(IPG) Errors
Jabber Condition
Jams
Late Collision Errors
Lost Frame Errors
Maintenance Fail(MT Fail)
Maintenance Media test (MT) (# of times failed)
Media Access Control(MAC) Frame
Miscellanous errors
Missing Start Frame Delimiter(SFD) Errors
Multicast Frames
Non-Unicast Packets
Octets
Out Success With Retry Errors
Out Of Window Collision Errors
Detecting
Output discards (OD)
Output Discard Errors
Output Errors
(# of errors)
Output FIFO Overrun Errors
Output Packets
Output Traffic(OKPk)
Counted in Megabytes
Packet Framing Error
Percent Collisions
Preamble Field
Readable Frames
Readable Octets
Runt Errors
Self-Test Fail (ST Fail)
(# of times failed)
Short Event Errors
Signal Quality Error(SQE) Test
Start Frame Delimiter(SFD) Field
Transmit Collision Errors
Transmitted Frames
Transmitted Octets
Unicast Packets

 

Glossary of Ethernet Terminology

Alignment Errors
Frame misalignments are input error and indicate damaged frames. A frame has an alignment error if the number of bits received is not a multiple of eight; since there are eight bits in one byte, frames must be received on byte boundaries. A misaligned frame also has a Frame Check Sequence (FCS) error. The causes of alignment errors include bad wiring, broken Ethernet cards, and cable runs that are out of specification.

Auto Partitioning
Auto Partitioning is an IEEE 802.3 specification which may be optionally implemented by the vendor. Auto Partitioning means that the hardware has automatically separated a port from the rest of the network. In accordance to the specifications, the hardware attempts to recover automatically when the condition clears. The port may be partitioned because too many collisions are occurring from the port (greater than 31 consecutive collisions) or because a single collision from the port is too long (greater than 1024 bit times in duration). Auto Partitioning can occur when a station is disconnected from a port or the wrong type of wiring is used, drastically increasing the number of collisions. Auto Partitioning may also occur if the attached device is sending a jamming signal which could bring down the entire network.

Broadcast Frames
Broadcast frames are frames that are directed to the broadcast group address. This does not include received frames which contained a framing error. Multicast packets are not included.

Buffer Full Errors
Buffer Full errors are input errors which indicate that the hardware has asserted an overwrite warning, signaling that the receive buffer is full. It does not mean, however, that a frame was necessarily lost.

Carrier Sense
Carrier Sense is the process used by devices on Ethernet to determine whether the cable is currently being used by a transmitting station. If electrical signals are detected on the cable, then carrier has been detected and a station is currently transmitting on the cable.

Carrier Sense Loss Errors
Carrier Sense Losses are output errors which indicate that carrier sense was lost or never appeared when the device was attempting to transmit a packet. A station listens on the wire as it transmits a frame in order to detect collisions. Carrier sense loss occurs when the hardware is transmitting a frame onto the wire and does not see its own carrier wave on the Ethernet. Carrier sense loss can occur if the optical power is marginal to a Fiber Optic Transceiver (FOT) on a router, for example. Loss of carrier sense may also be the result of plugging and unplugging the cable connections on a network.

Collisions
On an Ethernet, only one station may transmit at a time. Any computer wishing to transmit checks whether the cable is busy by using carrier sense. If carrier is not detected and the cable has not been busy for a specified period of time, then the station begins transmitting its data. While it is transmitting a frame, a computer listens to ensure that no other station starts transmitting. If any other station begins transmitting at the same time, a collision occurs. In other words, a collision is a condition where two devices detect that the network is idle and try to send packets at approximately the same time. Since only one device can transmit at a time, both devices must back off and attempt to retransmit again. The retransmission algorithm requires each device to wait a random amount of time, so the two stations are very likely to retry at different times; thus, the second station will sense that the network is busy and wait until the frame is finished being transmitted. If the two devices retry at the same time again, they will colliding once more. This process repeats until either the packet finally makes it onto the network without collisions, or 16 consecutive collision occur and the packet is aborted. (See Excessive Collisions.)

Collisions which occur due to two stations transmitting at the same time are normal events on an Ethernet. As the amount of traffic on a network increases, the number of collisions on a network also increases. However, collisions due to two stations transmitting at different times is not considered normal. (See Late Collisions.) Unusual numbers of collisions on a network may indicate a pathological problem and should be examined. Oftentimes, these collisions are the result of either a misconfigured network or a device transmitting when it should not be and stepping on other packets. Most likely, the collisions are the result of reflections or other problems with the physical plant, such as a bad transceiver. A ground loop is another possible cause of collisions if the cable is grounded at two points. Collisions can also happen if there is a bad cable or a station is incorrectly connected to the network.

Collision Rate
The collision rate is calculated by dividing the total number of packets by the total number of collisions and multiplying the result by 100. What is an acceptable level of collisions? This depends on your application and protocol along with the physical design of your network. In some cases, collision rates of 50% will not cause a large decrease in perceived throughput. If your network is slowing down and you notice the percentage of collisions is consistently above 10%, you may want your network examined.

Cyclic Redundancy Check
The Cyclic Redundancy Check (CRC) is a an error-checking algorithm used to check for the validity of a frame. A transmitting station uses the algorithm to compute a CRC value based on the bits in the frame and sets the Frame Check Sequence field of the frame to the calculated CRC value. A station receiving the frame can then use the same algorithm to determine whether or not the frame has been corrupted. If the computed CRC value of a frame received by a station does not match the value in its Frame Check Sequence field, then the frame is bad.

Cyclic Redundancy Check (CRC) Errors
See Frame Check Sequence (FCS) Errors.

Data Rate Mismatch Errors
Data Rate Mismatches are input errors which occur when the incoming data rate is not with in the tolerance level of 10MHZ + or - 0.01%. In other words, an Ethernet device on the network is not transmitting at the correct 10 million/bits per second on the network and is breaking the specifications.

Excessive Collision Errors
Excessive Collisions are output errors which indicate that frames were not transmitted due to jams on the network. Excessive collision errors are the number of times that a frame destined for the network incurred 16 collisions and was discarded. Excessive collision errors are cases where the router, for example, tries several times to transmit a frame, but is collided with each time; eventually the router gives up and throws the frame away. The usual cause of jams is an exceedingly busy network. If the jam count fluctuates between being zero and a huge number, an intermittent problem is occurring on the network which causes the devices to repeat collisions at a high rate, such that the router transmit any frames.

Frame
See Media Access Control (MAC) Frame.

Frame Check Sequence (FCS) Errors
Frame Check Sequence (FCS) errors are input error that are aligned but fail the Cyclic Redundancy Check (CRC). FCS errors indicate damaged frames. There can be many causes for FCS errors, including flaky in-room wiring, broken Ethernet cards, and cable runs which are too long.

Framing Errors
See Alignment Errors.

Frame Length Errors
Frame Length errors are input errors due to frames that have a length field value greater than the maximum allowed data size. Unfortunately, if both IEEE 802.3 Ethernet and Ethernet Type 2 coexist on the same network, the Frame Length errors will show a misleadingly high error rate. This is because Ethernet Type 2 frames do not have a length field similar to IEEE 802.3 Ethernet frames.

Frame Too Long Errors
Frame Too Long errors are input errors where the frame has a valid Start Frame Delimiter (SFD) and Frame Check Sequence (FCS), but it is longer than the maximum allowable size of an Ethernet frame.

Giant Frame Errors
See Frame Too Long Errors.

Heartbeat Errors
Heartbeat errors are output errors which indicate that a heartbeat was not detected after transmitting a frame. In Ethernet Version 2, heartbeat is a test of the collision functionality of the transceiver. The term heartbeat is often used interchangeably with Signal Quality Error (SQE). See Signal Quality Error Test.

Input Discard Errors
Input Discards are input errors, which occur when the hardware attempts to receive a packet, but the local receive packet buffer is full. Input discard errors usually indicate that the network has more traffic than it can handle. They may also indicate a software discard such as an unhandled protocol.

Input Drop Errors
See Buffer Full Errors.

Input Errors
Input Errors is the sum of all the input errors on an interface.

Input FIFO Overruns Errors
Input FIFO Overruns are input errors indicating that the input queue overflowed while copying received frames. The problem is that the hardware is unable to store bytes in the local packet buffer as fast as they come off the wire.

Input Overrun Errors
See Frame Too Long Errors.

Input Packets
See Readable Frames.

Inter Packet Gap (IPG) Errors
The Inter Packet Gap (Inter Frame Gap or IFG) is an enforced quiet time between transmitted Ethernet frames. The Inter Frame Gap is required so that listening stations are provided with enough time to detect that there is no carrier sense on a wire. Inter Packet Gap errors are usually the result of a device that is not adhering to the Ethernet specifications.

Jams
If a workstation detects a collision while it is transmitting, the station does not immediately stop transmitting. Instead, the workstation puts out a jam so all other stations will detect the collision. When a repeater detects a collision on one of its ports, it also puts out a jam on all its ports. This ensures that the transmitting stations detect the collision and that the stations on the other repeater ports see the collision. (See excessive collisions.) The purpose of a workstation to continue transmitting after detecting a collision for a short period of time ensures that every station on the network has time to see the collision.

Late Collision Errors
Late Collisions are input errors due to a collision which occurs after a station has been transmitting for some period of time. The device has transmitted enough of a frame that every other station on the wire should know that the media has been acquired. Late Collisions indicate that the time to propagate the signal from one end of the network to another is longer than the time to put the entire packet on the network; therefore, the two devices that cause the Late Collision never detect that the other station is sending data until after both stations put entire frames on the network. Late collisions are detected by the transmitter after the first Òslot timeÓ of 64 byte times. Late Collisions, as a result, are only detected during transmissions of frames longer than 64 bytes. Late Collisions are similar to normal collisions; it just happens too late.

Typical causes of late collisions are that either the entire network is temporally larger in diameter than is allowed by the IEEE 802.3 specification; the cable lengths of segments may be in excess of the maximum length permitted for the cable type or that there are excessive numbers of repeaters between network devices. Another cause of late collisions is that there are defective Ethernet transceivers or controllers. Some damaged frames on the net look enough like collision fragments to fool the hardware`s controller chip. One bad thing about late collisions is that collisions involving small frames cannot be detected by the transmitter. A network suffering a measurable rate of late collisions (on large frames) is also suffering loss of small frames. Some higher protocols do not cope well with packet losses, no matter how small. A 1% packet loss is enough to reduce the speed of NFS by 90% with the default retransmission timers. Additionally, Ethernet controllers do not retransmit frames lost to late collisions.

Lost Frame Errors
Lost Frames are input errors which indicate that frames have been lost due to insufficient buffer space. Lost frames are packets missed by the station because it was either too busy to take the interrupt in time or because the queue of packets in memory waiting for processing overflowed. Both indicate that the device is being swamped with traffic. If the traffic on the network is not overloaded, then there may be a machine causing sporadic packet floods of the network.

Maintenance Fail (MT Fail)
Every few seconds, the router sends out a signal on its interfaces checking for an open cable. After a certain number of failed maintenance checks, the interface is brought down for self-testing.

Media Access Control (MAC) Frame
A frame is a packet that is transmitted on the physical wire. IEEE 802.3 (IEC 8802-3) protocol standard is an ANSI specification for Ethernet. An IEEE 802.3 frame consists of a Preamble, Start Frame Delimiter, Destination Hardware Address, Source Hardware Address, Data, and Frame Check Sequence:

Preamble (7 Octets)
Start Frame Delimeter (1 Octet)
Destination Address (6 Octets)
Source Address (6 Octets)
Length Field (2 Octets)
Payload (46 to 1500 Octets)
Frame Check Sequence (4 Octets)

If the protocol being used is Ethernet Type 2, then the frame has a field indicating the packet type instead of a length field. Ethernet Type 2 is an older protocol developed by Xerox which is still in common use today.

Missing Start Frame Delimiter (SFD) Errors
Missing Start Frame Delimiters are input errors which occur when a burst of data greater than or equal to 10 bytes in length is received with no SFD. A Start Frame Delimiter is required to indicate to the listening stations that the beginning of a frame is being transmitted.

Multicast Frames
Multicast frames are frames that are directed to an active multicast group address. This does not include frames received which had a framing error. Broadcast packets are also not included.

Non-Unicast Packets
Non-Unicast Packets are frames that are addressed to either broadcast or multicast addresses.

Octet
A byte of data, which contains eight 0 or 1 bits.

Out Success With Retry Errors
Out Success With Retry errors are output errors which indicate that a frame successfully transmitted after one or more collisions. Out Success With Retry Errors are the number of frames that experienced a a collision and retransmit before success. The total collisions divided by the number of Out Success With Retry errors will provide the average number of collisions per packet transmitted with collisions.

Out of Window Collisions Errors
Out of Window Collisions are a type of output error. Out of window collision errors occur when a frame in the process of being transmitted collides with another frame. This error usually occurs either when some interface on the network fails to defer or the network has too many stations. (See Late Collisions.)

Output Discard Errors
Output Discards are output errors that occur when the router has to throw a packet out instead of queuing it for transmission on the Ethernet. Output discard errors usually indicate that the network has more traffic than it can handle. They may also indicate a software discard such as no route to destination.

Output Errors
Output errors are the sum of all the output errors that occur on an interface.

Output FIFO Overrun Errors
FIFO Overruns are a type of output error. Output FIFO Overruns occur when the output queue in the adapter underflowed while putting a frame on the wire. This problem occurs when the interface is not receiving bits of the frame fast enough.

Output Packets
See Transmitted Packets.

Preamble Field
The preamble field synchronizes the timing of all listening devices on the network. The preamble field is 7 bytes and consists of a series of 56 alternating zero and one bits.

Readable Frames
A good or Readable Frame is a frame between 64 and 1518 bytes in length and has an SFD and good FCS. In other words, Readable Frames are the number of frames that are successfully received by the port, including broadcast and multicast frames.

Readable Octets
Readable Octets are the total number of data and padding octets in frames that are successfully received by the device, including broadcast and multicast frames. This does not include frames received with a framing errors. The number of Readable Octets does not include the octets in the address, length/type and CRC fields.

Runt Errors
Runt Errors are input errors in which the received frames are smaller than 64 bytes, but have no CRC or alignment errors. Frames must be a minimum length in order ensure that all stations on the network will detect possible collisions. Runt errors are likely the result of a faulty device on the network, or defective software.

Self-Test Fail (ST Fail)
The total number of self test failures on this interface. A self-test is a internal card diagnostic procedure that is performed when the router is restarted or the interface is brought into a self-test mode.

Short Event Errors
See Missing Start Frame Delimiter (SFD) Errors.

Signal Quality Error (SQE) Test
Signal Quality Error (SQE) is the IEEE term for a collision. The term SQE is often used interchangeably with heartbeat. Signal Quality Error Test is an IEEE 802.3 function that tests the transceiver. The SQE Test is a means of detecting a transceiver`s inability to detect collisions. Without SQE Test, it is not possible to determine if your collision detector is operating properly. SQE Test is implemented by generating a test signal on the collision pair from the transceiver following every transmission on the network. It does not generate any signal on the common medium.

IEEE 802.3 specifications state that IEEE 802.3 compliant repeaters must not be attached to transceivers that generate a heartbeat, because a jam signal prevents redundant collisions from occurring on the network. Therefore, the SQE Test between the transceiver and an 802.3 repeater usually must be disabled. See Heartbeat.

Start Frame Delimiter (SFD) Field
The Start Frame Delimiter field of a frame signals the beginning of a frame being transmitted.

Transmit Collision Errors
Transmit Collision Errors are output errors which indicate the total number of collisions that the device encountered when attempting to transmit frames.

Transmitted Frames
Transmitted Frames are the number of frames that are transmitted. This is incremented when a frame is successfully transmitted, even if a collision occurs. In other words, Transmitted Frames include transmissions that succeeded after some number of collisions. This also includes multicast and broadcast frames.

Transmitted Octets
Transmitted Octets are the number of data and padding octets of frames that are transmitted. This includes octets in broadcast and multicast frames that are transmitted. This does not include the address fields, length/type field, and CRC field of the frame.

Unicast Packets
Unicast Packets are frames that are not addressed to either broadcast or multicast addresses.