• Not Answered

NDCC Port: Excessive packets discarded on port

Hello,

I just wanted to ask some opinion regarding an alarm that keeps coming and going intermittently. From NDCC we keep getting this Message for Primary Connection saying "Port: Excessive packets discarded on port 2.4" the port sometimes changes to another port but currently the port is 2.4 it is a primary for one of our CIOC. If we leave that alarm like that it will go away but after 24hrs or so it will disappear and will come again. It was happening before but less frequent like once or twice per 2-3 months but now it is happening in few days. Just wanted to know what does it mean? Thank you for anyone who will answer. 

Below is the alarm that is showing, it is coming from a Emerson smart switch RM100

dolatrec

• 3 Sep 2025 7:54 PM
• Cancel

3 Replies

• Andre Dicaire

  

  • 3 Sep 2025 8:06 PM

  The short answer is the switch is detecting malformed packets that cannot be processed. They are discarded. There are several reasons, poor network connection, incorrect media type (Multimode fiber in a Single mode port), or a faulty device sending problematic packets. Check physical network connections, verify cable. Trouble shoot the switch. What is connected to this port? is it controller, workstation, IO node?
  
  Move the cable to a different port on the switch and see if the issue follows the cable (Swap with another port). If issue moves with cable, it is not the switch. If it remains on the same port, issue is in the switch.

  Andre Dicaire

  • Cancel
• dolatrec

  

  • 3 Sep 2025 10:29 PM

  In reply to Andre Dicaire:

  Thanks for the reply, Andre, it is a CIOC card primary which is connected on that port 2.4 that is being reported. Where can I check or verify that "multimode fiber in a single mode port"? Is it in NDCC? I'll probably check the physical cable (swap port etc.) during our planned plant shutdown as I cannot risk when production is online unless it breaks down totally.

  • Cancel
• Andre Dicaire

  

  • 4 Sep 2025 3:31 PM

  In reply to dolatrec:

  check out BOL section on Hardware Alarm condition parameters (Smart Swich).

  Excessive Packets Discarded On Port

  Unusually high levels of packets discarded (due to CRC errors, packets too small (runts) or packets too large (giants) can be an indicator of loose or deteriorating network cable connections, damaged network cables, or failing network interface electronics.

  If HAC_DISCPKTS_PWA.CVI has the value 1, 2, or 3 the workstation interrogates the switch for:

  • The list of all ports installed on the switch.
  • Whether each port is enabled or disabled.
  • Information about packets discarded (due to CRC errors, runts, and giants) in the last 60 seconds on each installed port.

  For each enabled port, the number of packets discarded in the last 60 seconds is written to the corresponding array element in HAC_DISCPKTS_PRT[n]. 0 is written for ports that are not installed or that are disabled.

  The sum from all ports of the packets discarded in the last 60 seconds is written to HAC_DISCPKTS_VAL.

  If the port is enabled, and if the number of discarded packets on this port in the last 60 seconds is greater than HAC_DISCPKTS_LIM the raw DISCPKTS condition is true for that port.

  The Smart Switch Module can be  viewed on line and the HAC_DISCPKTS_PRT[n] array can be viewed as well.  The port identification 2.4 aligns with physical numbering on the switch, but the array is a set of 32 Float values.  So 2.4 might mean Float number 12.  If one port has issues, the overal HAC_DISCPKTS_VAL will align.  If multiple ports have discarded packets, VAL would show the sum.  Alerts would report affected ports, so looks like must 2.4 is seeing issues.

  The Smart Switch module Class explains the assignments:

  Array index n = (switch module m - 1) * PPM + (switch module port p)

  where PPM (Ports per Module):

  • Is 8 for FP20 and RM100 switches.
  • Is 4 for MD20 and MD30 switches.
  • Is 24 for RM104 switch

  Thus for an RM100:

  • Port 1.1 information goes into HAC_XXXXXXXX_PRT[1]
  • (1 - 1) * 1 = 1
  • Port 4.8 information (if present) goes into HAC_XXXXXXXX_PRT[32]
  • (4 - 1) * 8 = 32

  For an MD30:

  • Port 7.4 information (if present) goes into HAC_XXXXXXXX_PRT[28]
  • (7 - 1) * 4 = 28

  For an RM104:

  • Port 1.1 information goes into HAC_XXXXXXXX_PRT[1]
  • (1 - 1) * 1 = 1
  • Port 1.24 information (if present) goes into HAC_XXXXXXXX_PRT[24]
  • (1 - 1) * 24 = 24

  Andre Dicaire

  • Cancel