The Fast Ethernet (ETH100) ports

[15:03:33] ABILIS_CPX: D P PO:ETH100

PO:58  - Not Saved (SAVE CONF), Not Refreshed (INIT) --------------------------
ETH100 ------------------------------------------------------------------------
       LOG:NO            MODE:AUTO         DUPLEX:HALF   TPPOL:AUTO
       dma-rxbuf:25      dma-txbuf:25      ip-rxbuf:25   arp-rxbuf:5
       pppoed-rxbuf:5    pppoes-rxbuf:25   ieee-rxbuf:25

Usually this parameter makes possible to activate/deactivate logging functionalities of meaningful events of the port as well as the detection and signalling of alarms in case of critical events.

The following table shows the available options and the related functionalities usable by the parameter:

Option	Meaning
D	Recording of the driver state changes and/or the meaningful events in Debug Log
S	Recording of the driver state changes and/or the meaningful events in the System Log
A	Periodic detection of possible alarms. The detected alarms can be displayed the command ALARM VIEW or by the analogous command available on the UTILITY of the LCD display on the front panel
L	On alarm detection, acoustic signal generation plus a message on the LCD display. This function depends on activation of alarms detection by the "A" option
T	Generation by the Agent SNMP of Abilis CPX of SNMP traps corresponding to any change of the driver state and/or occurring of meaningful events

Beside the already described options the following values are also allowed:

Option	Meaning
NO	It means that all the logging functionalities, alarms detection and generation, above mentioned, are disabled.
ALL	It means that all the logging functionalities, alarms detection and generation, above mentioned, are enabled.
+E	This option added to one or more of the previous ones, extends its (their) set of meaningful events. The value "ALL+E" activates all the options and extends the set of meaningful events. The value "NO+E" is meaningless so it is ignored.

Options can be combined together.

Some examples:

• setting "LOG:DS+E", activates the extended logging functions for Events Log and System Log
• setting "LOG:STA", activates the extended logging functions for System Log, SNMP traps generation and periodic detection of alarm states;

By using the characters "+" and "-" as prefix of one or more options is possible to add or delete one or more functionalities without setting from the scratch the value of the parameters.

Some examples:

• Suppose the current value of the parameter is "LOG:DSTA", by setting "LOG:-A", the periodic detection of eventual alarm states is removed, leaving unchanged all the remaining options; in such way the final value of the parameter will be "LOG:DST";
• Suppose the current value of the parameter is "LOG:ST", by setting "LOG:+DA", the logging function of the events on the Events Log and the periodic alarm detection are added to the already activated options; in such way the final value of the parameter will be "LOG:DSTA".

The changes made on this parameter are immediately activated, without the need of initialization commands.

It sets the link speed or the negotiation ("AUTO") mode.

Negotiation is made according to the Nways protocol, however if the counterpart does not support that protocol, adjusting is carried out trying alternatively the use of 10TP and 100TX. The faster will be choose.

The Nways protocol negotiates also the half/full duplex more, adjustment obtained by attempt, always uses the "HALF" duplex mode.

It sets the transmission mode: "HALF" correspond to alternating bidirectional; "FULL" to simultaneous bidirectional.

Connections to HUBs always require the "HALF" mode, while the ones to SWITCHs (depending on the SWITCH itself) could accept the "FULL" mode.

This parameter is meaningful only if MODE: is set to "10TP" or "100TX"; while it is meaningless if MODE: is set to "AUTO".

It sets the Fast Ethernet board to the real polarity of signals in the TP cable.

For a correct installation the polarity will always be normal, however it is possible to correct a wiring error via software.

The value "AUTO" activates the automatic recognition; while with values "NORMAL" and "REVERSE" the adjustments forced manually.

It sets the number of receive buffers for DMA.

It sets the number of transmit buffers for DMA.

It sets the number of receive buffers for IP.

It sets the number of receive buffers for ARP.

It sets the number of receive buffers for PPPOE Discovery protocol.

It sets the number of receive buffers for PPPOE Session protocol.

It sets the number of receive buffers for IEEE Bridge.

[16:05:04] ABILIS_CPX: D S PO:ETH100

PO:58  ------------------------------------------------------------------------
ETH100 STATE:READY          MAC-ADDR:00-80-48-CD-05-7B   CARD:FL100TX
       MODE:AUT-100TX       DUPLEX:AUT-HALF              TPPOL:AUT-NORMAL
       IEEE-FILTER:OFF
       -----------|---INPUT---|--OUTPUT---|-----------|---INPUT---|--OUTPUT---|
       SUCC       |      16320|        571|CHAR       |    1522126|      35033|
       BROADCAST  |      18100|         61|MULTICAST  |          0|          0|
       NOBUF      |         41|           |LOST-IP    |          0|           |
       LOST-ARP   |          0|           |LOST-RARP  |          0|           |
                  |           |           |N-COLL     |           |          7|
       FCS        |          0|           |L-COLL     |          0|          0|
       RX-OVR     |          0|           |E-COLL     |           |          0|
       TX-UND     |           |          0|DEFERRED   |           |         37|
       CARR       |          0|           |ALIGN      |          0|           |
       TBE        |         41|        632|SHORT      |          0|          0|
       QED        |           |          0|LONG       |          0|          0|
       WDG-TOUT   |          0|           |RUNT       |          0|           |
       JBR-TOUT   |          0|           |HB-FAIL    |          0|           |
       GPT-TOUT   |          0|           |NEG-FAIL   |          0|           |
       BUS-ERR    |          0|           |LINK-FAIL  |          3|           |
       DSC-ERR    |          0|           |FILT-FAIL  |          0|           |
       MII-ERR    |          0|           |CARD-RESET |          2|           |
       ------------------------------------------------------------------------

[16:15:38] ABILIS_CPX: D SE PO:ETH100

PO:58  ------------------------------------------------------------------------
ETH100 --- Cleared 000:03:57:43 ago, on 22/07/2004 at 12:19:16 ----------------
       -----------|---INPUT---|--OUTPUT---|-----------|---INPUT---|--OUTPUT---|
       SUCC       |      16320|        571|CHAR       |    1522126|      35033|
       BROADCAST  |      18100|         61|MULTICAST  |          0|          0|
       NOBUF      |         41|           |LOST-IP    |          0|           |
       LOST-ARP   |          0|           |LOST-RARP  |          0|           |
                  |           |           |N-COLL     |           |          7|
       FCS        |          0|           |L-COLL     |          0|          0|
       RX-OVR     |          0|           |E-COLL     |           |          0|
       TX-UND     |           |          0|DEFERRED   |           |         37|
       CARR       |          0|           |ALIGN      |          0|           |
       TBE        |         41|        632|SHORT      |          0|          0|
       QED        |           |          0|LONG       |          0|          0|
       WDG-TOUT   |          0|           |RUNT       |          0|           |
       JBR-TOUT   |          0|           |HB-FAIL    |          0|           |
       GPT-TOUT   |          0|           |NEG-FAIL   |          0|           |
       BUS-ERR    |          0|           |LINK-FAIL  |          3|           |
       DSC-ERR    |          0|           |FILT-FAIL  |          0|           |
       MII-ERR    |          0|           |CARD-RESET |          2|           |
       ------------------------------------------------------------------------

It shows the current state of the ETH100 port driver.

Driver	States	Meaning	Values shown in:
			System Log	Debug Log	Display LCD
ETH100	DOWN	Port is not active	DN
	READY	Port is active	RD
	TXDISABLED	The transmission has been disabled due to a fatal error	TD
	HALTED	Port halted because of a fatal error	HL
	FAILED	Port is not working because of an initialisation error	FL
	NEGOTIATING	Port is negotiating	NG
	ERR	Software error. Contact Abilis assistance.	NA

The parameter MAC-ADDR: shows the MAC (Media Access Control) address used by the ETH100 port driver. MAC addresses are usually made of a sequence of 6 octets (48bits) and they are displayed as a sequence of 6 hexadecimals separated by the character ":" (colon).

It shows the Fast Ethernet board type in use.

The prefix "AUT-" means that the speed mode, currently used, is the result of the negotiation procedure.

The prefix "FIX-" means that the speed mode, currently used, was selected manually.

The value UNKNOWN means that the negotiation procedure is active but no valid signal has been yet detected, owing to, for example, the link absence.

The prefix "AUT-" means that the transmission mode, currently used, is the result of the negotiation procedure.

The prefix "FIX-" means that the transmission mode, currently used, was selected manually.

The value UNKNOWN means that the negotiation procedure is active but no valid signal has been yet detected, owing to, for example, the link absence.

The prefix "AUT-" means that the polarity, currently used, has been adjusted automatically.

The prefix "FIX-" means that the polarity, currently used, was selected manually.

The value UNKNOWN means that the negotiation procedure is active but no valid signal has been yet detected, owing to, for example, the link absence.

In case of negotiated speed or manually set to 100TX, the polarity will be always displayed as UNKNOWN because it cannot be determinated.

.

The counter SUCC (INPUT/OUTPUT) shows the number of frames successfully received/sent.

The counter CHAR (INPUT/OUTPUT) shows the number of characters received/sent.

The counter BROADCAST (INPUT/OUTPUT) is incremented every time a "broadcast" frame is received/sent.

The "broadcast" frames are sent to all the LAN Ethernet stations.

The counter MULTICAST (INPUT/OUTPUT) is incremented every time a "multicast" frame is received/sent.

The "multicast" frames are sent to a group of the LAN Ethernet stations.

The counter NOBUF (INPUT) is incremented every time a frame is discarded because of the unavailability of receiving buffers.

High values of this parameter compared to the overall data exchanged could mean that the Abilis system is overloaded, so that it is convenient to contact the seller and consider to upgrade it.

The counter LOST-IP (INPUT) is incremented every time a received frames is discarded because of the unavailability of IP buffers.

The counter LOST-ARP (INPUT) is incremented every time a received frames is discarded because of the unavailability of ARP buffers.

The counter LOST-RARP (INPUT) is incremented every time a received frames is discarded because of the unavailability of RARP buffers.

The counter N-COLL (OUTPUT) is incremented every time a "normal collision" occurs.

Collisions, in a Ethernet network, are an absolutely normal occurrence, but if their number grows up to the overall amount of data exchanged, then it should be considered the network segmentation, for example through a SWITCH.

LANs highly congested present a high number of "extended collisions" (E-COLL).

The counter L-COLL (OUTPUT) is incremented every time a "late collision" occurs, while a frame is sent or received.

Significative values of this parameter could mean:

• hardware failures of one or more LAN stations;

• wrong configuration of the modality half/full duplex in one or more LAN stations;

• the LAN is too wide, segmentation has to be considered. The LAN collision domain is estimated about 2,5km for a 10Mbut/sec LAN while 250m for a 100Mbit/sec one.

The counter E-COLL (OUTPUT) is incremented every time that 16 consecutive collisions for the same frame occur.

If the number of collisions grows up to the overall amount of data exchanged, then it should be considered the network segmentation, for example through a SWITCH.

The counter FCS (INPUT) is incremented every time the FCS (or CRC) check procedure, of each frame, fails.

High values, especially if growing, of this counter could mean a bad quality of the transmission line, corrupting the received data; it surely points out errors into the received data. A bad quality turns into a performance lowering and into possible breaks out of the link.

The counter RX-OVR (INPUT) is incremented every time the communication board reports an "overrun" due to the received data. It occurs whenever the Ethernet driver, due to an excessive payload within the Abilis CPX, is not fast enough to take data from the communication board. Up to now, no Abilis CPX application gave such "overrun" problems, however if the value of this counter get more and more significative compared to the amount of data exchanged, it should be convenient to contact the seller and consider the Abilis CPX upgrade.

However, if high numbers of FCS are also detected, then it is more probable that the "overload" is caused by the excessive number of errors on the transmission line, which, as consequence, should be repaired.

The counter TX-UND (OUTPUT) is incremented every time the communication board reports an "underrun" due to the sent DATA. In the actual implementation this counter has to always be zero, however if other values would be displayed, please contact the technical support.

The counter DEFERRED (OUTPUT) is incremented every time a frame transmission is deferred.

This counter does not highlight particular problems.

The counter CARR (INPUT) shows the number of "carrier loss" in TP links.

it is normal that this counter increments every time the TP link is missing. Growing values, not due to intentional disconnections, could mean:

• bad quality of the transmission line;

• the HUB, which the Abilis is connected to, doesn't work properly.

The counter ALIGN (INPUT) shows the number of received frames with an align error.

Non-zero values could stand for problems in LAN station, or for a bad transmission line quality.

Only internal use

Only internal use.

The counter SHORT (INPUT) is incremented every time the length of the received Ethernet frame is shorter than 64 bytes. Non-zero values could stand for problems in LAN station, or for a bad transmission line quality.

The counter SHORT (OUTPUT) is incremented every time the "higher level port", such as IPLAN, requests to send a frame whose length is shorter than 64 bytes.

In the actual implementation this counter has to always be zero, however if other values would be displayed, please contact the technical support.

The counter LONG (INPUT) is incremented every time the length of the received Ethernet frame is longer than 1518 bytes. Non-zero values could stand for problems in LAN station, or for a bad transmission line quality.

The counter LONG (OUTPUT) is incremented every time the "higher level port", such as IPLAN, requests to send a frame whose length is longer than 1518 bytes.

In the actual implementation this counter has to always be zero, however if other values would be displayed, please contact the technical support.

The counter WDG-TOUT (INPUT) shows the number of "watchdog-timeout" events.

The Fast Ethernet has its own "watchdog timer", which checks the frame receiving timings. If the frame receiving time is bigger that the allowed value, generally due to transmission errors, the "watchdog timer" will end the reception and signal this event to the driver.

Significative values could stand, most probably, for a bad transmission line quality, but they could also represent faults of the fast-ethernet adapter.

The counter RUNT (INPUT) is incremented every time while a frame is received, a collision, inside the homonymous window (i.e. within the first 512 bytes), is detected.

The above mentioned frame is discarded.

The counter JBR-TOUT (INPUT) shows the number of "jabber-timeout" events.

The "jobber timeout" is an embedded protection of the board to avoid that its transmission problem could halt all the LAN. Particularly if the frame transmission requires a too long time (for example for sending more than 2048 bytes), The transmitter is forcedly deactivated and the event is signalled to the driver which increments the counter JBR-TOUT.

A significative value of this parameter, especially if growing, could represent a faulty board, please contact the technical support.

This parameter is only valid for 10TP links.

This counter shows the number of "heartbeat-fails", that is the bad recognition of the "link pulses" of the 10TP links.

Non-zero values, especially if growing, could stand, most probably, for a bad transmission line quality, but they could also represent faults of the fast-ethernet board or of the HUB port.

The counter GPT-TOUT (INPUT) shows the number of "general purpose timer timeout" events occurred.

Only internal use.

The counter NEG-FAIL (INPUT) shows the number of negotiation failures.

The counter BUS-ERR (INPUT) shows the number of errors detected on the PCI bus.

A significative value of this parameter, especially if growing, could represent a CPX hardware problem, please contact the technical support.

The counter LINK-FAIL (INPUT) shows the number of "link failure" events occurred.

it is normal that this counter increments every time the TP link is missing. Growing values, not due to intentional disconnections, could mean:

• bad quality of the transmission line;

• the HUB, which the Abilis is connected to, doesn't work properly.

The counter DSC-ERR (INPUT) shows the number of "descriptor errors" occurred.

Only internal use.

The counter FILT-FAIL (INPUT) shows the number of "filtering failure" events.

Only internal use.

The counter MII-ERR shows the number of "MII errors" occurred, that is physical errors detected on the MII interface.

Only internal use.

Actually not in use.

This variable shows the CPX port number which the SNMP trap refers to.

This variable shows either the port type which the trap refers to (in this case ETH100) and the description eventually associated to it.

This variable shows the current state of the ETH100 port driver which the trap refers to. The following table shows the relationship among the state value stored in the SNMP variable and the ones obtained executing the command D S.

Value of the SNMP variable	Correspondent state of the ETH100 driver
down(0)	DOWN
ready(1)	READY
txDisabled(2)	TXDISABLED
halted(3)	HALTED
failed(4)	FAILED
negotiating(5)	NEGOTIATING

This variable stores the elapsed time (in cent of second) from the system start to the SNMP trap event occurrence.

This value is usually shown by the format days:hours:minutes:seconds.