Late events: The number of collisions detected after the allowable detection period. Usually indicates a network topology problem, such as daisy-chaining two Express hubs running at 100 Mbps with UTP cable (instead of Intel Cascade Cables in a stack), daisy-chaining too many hubs running at 10 Mbps, or using cables longer than 100 meters.

Very long events: The number of times MAU jabber lockup protection (MJLP) was detected due to transmission of data that exceeded 5 msec. in duration (octet count greater than maxFramSize). This information can help you identify faulty devices or ports.

Rate mismatch: A count of the occurrences of out-of-specification bit rates. This indicates the number of times the FIFO buffer overruns or underruns due to transmission rate errors. This could indicate an incorrect FIFO setting on a network adapter or a faulty adapter.

Total errors: The total number of errors detected since the port was last reset. Total errors include FCS errors, alignment errors, frame too long, short events, late events, very long event, and rate mismatch errors.

Runts: The number of frames detected that are less than the minimum permitted frame size and have a good FCS. Runts usually indicate collision fragments, a normal network event.

Auto partitions: The number of times this port was automatically partitioned. This condition occurs when 64 consecutive collisions are detected on the port. These collisions are due to excessive traffic on the segment, a malfunctioning port, or a malfunctioning adapter that is jabbering.

35

Page 36
Image 36
Intel EE110MM manual