0537 - Ethernet error threshold exceeded

There are two types of error that can generate UAM 0537:
  • Too much data received on the port.

  • An Ethernet interface experiences excessive errors at the physical layer, such as CRC or framing errors. This error is issued when the Ethernet statistics indicate errors occurring in any 15-second window. The alarm will clear when no errors have occurred in the previous 15 second window. In full-duplex mode, collisions will not be ignored but when configured for half-duplex mode, they will be ignored, since collisions are expected in half-duplex mode. When collisions occur repeatedly for the same packet more than 16 times then the “excess collisions” error count is pegged.

Example
    RLGHNCXA21W 06-12-07 12:01:43 EST  EAGLE 35.6.0 
**  2315.0537 ** ENET 1201,B             Ethernet error threshold exceeded

Alarm Level: Major

Recovery

This error may be due to
  • Too much incoming data for the port to handle.
  • Faulty hardware; an EAGLE card, cable, or immediate IP switch/router.
  • Configuration mismatch problems; the following items should match on both ends:
    • Duplex
    • Speed
    • Ethernet type
    • Autonegotiate used on EAGLE or immediate IP switch/router (it is recommended to lock down your IP connection and do not use AUTONEGOTIATE)
  1. Perform this step if UAM 0537 is issued and the card boots; otherwise, skip to 2. If the card boots, the problem is probably too much incoming data on the port. Deactivate the port that has too much data using the dact-ip-lnk:loc=xxxx:port=xxxx command. Fix the network problem, then reactivate the port using the act-ip-lnk:loc=xxxx:port=xxxx.

    Note: An external hardware failure can cause IP overload on the port resulting in card boot. This problem is uncommon, but can occur when a problem with the customer network overwhelms the Ethernet port with a storm of traffic. The solution is to resolve the network problem. The dact-ip-lnk:loc=xxxx:port=xxxx also makes it possible to deactivate a bad Ethernet port so the other port can work.

  2. Use the following command to determine the current EAGLE configuration for the IP card reporting the errors.

    rtrv-ip-lnk:loc=xxxx

    wherexxxx is the card location identified in the error message.

  3. Use the netstat -d pass command to view the driver statistics for the local interface.

    • For the A ethernet interface, pass:loc=XXXX:cmd="netstat -d 0"

    • For the B ethernet interface, pass:loc=XXXX:cmd="netstat -d 1"

      where xxxx is the card location identified in the error message.

    Table 1 and Table 2 show the errors that show up on the DCM/DSM/EDCM/SSEDCM and E5-ENET/E5-ENET-B card types. Because the boards use different ethernet chips, the statistics that are available are different.
    See (DCM/DSM/EDCM/SSEDCM) Ethernet Error Statistics

    Statistic Peg

    Description

    Half Duplex threshold count

    Full Duplex threshold count

    overflow

    Number of times the RX FIFO overflowed for frames received.

    1

    1

    CRC errors

    Number of frames received or discarded with CRC errors but no framing errors.

    1

    1

    short frame

    Number of frames received or discarded with carrier sense or RX-D.V. activity less than the “ShortEventMaxTime” (74-82 bit times).

    1

    1

    oversize frame

    Number of receive frames with greater than the 1518 byte maximum frame size.

    1

    1

    terminal count

    Receive DMA tried to receive more than the buffer capacity.

    1

    1

    excess collisions

    Number of times a frame collided 16 times without successful transmission.

    1

    1

    underflow

    Count of transmit underflow errors.

    1

    1

    CS error

    Number of times the transmitter had transmit data available and was ready to transmit but had to defer transmission due to carrier sense going HIGH. (Tx defer count in the See data sheet).

    n/a

    1

    alignment error

    Number of frames received or discarded with both a framing error and a CRC error.

    1

    1

    very long event

    Number of times the transmitter is active for greater than the MAU Jabber Lockup Protection Timer allows ([4-7ms] at 10 Mbit and [0.4 – 0.75ms] at 100 Mb.11 rxerrorNumber of times RXERR is asserted by the Ethernet PHY.

    1

    1

    num_job_q_full

    Number of times the net task job queue was full.

    1

    1

    rxerror

    Number of times RXERR is asserted by the Ethernet PHY.

    1

    1
    GEI (E5-ENET/E5-ENET-B) Ethernet Error Statistics

    Statistic Peg

    Description

    Half Duplex threshold count

    Full Duplex threshold count

    crcerrs

    Number of receive frames with CRC errors.

    1

    1

    algnerrc

    Number of receive frames with alignment errors (the frame is not an integer number of bytes in length).

    1

    1

    rxerrc

    Number of frames received in which I_RX_ER was asserted by the PHY.

    1

    1

    ecol

    When 16 or more collisions have occurred on a frame, this register increments, regardless of the value of collision threshold.

    1

    1

    tncrs

    This register counts the number of successful frame transmissions in which the internal carrier sense signal from the PHY was not asserted within one slot time of start of transmission.

    n/a

    1

    tuc

    Transmit underrun count.

    1

    1

    rlec

    This register counts receive length error events.

    1

    1

    rnbc

    The number of times that frames were received when there were no available buffers in host memory to store those frames.

    1

    1

    ruc

    This register counts the number of received frames that passed address filtering, and were less than minimum size (64 bytes from <Destination Address> through <CRC>, inclusively), and had a valid CRC.

    1

    1

    roc

    This register counts the number of received frames with valid CRC field that passed address filtering, and were greater than maximum size.

    1

    1

  4. If the error is due to configuration problems, correct the configurationso the EAGLE and the IP switch/router match.
  5. If the configuration matches on both ends of the IP segment, replace the EAGLE card as identified in the error message.

    If replacing the card does not fix the issue, begin local procedures to verify the local IP segment.