• How come I have these errors In lan? bad NIC?
    My LAN is connected to a Cisco switch SG 200-26.


  • faulty wiring or faulty network card or faulty cisco switch

    i'd try to swap them in that order ;)


  • 678 packets out of 743070326?  That is a 0.00009124304608551951% error rate.  Don't worry about it.

  • LAYER 8 Netgate

    Ethernet error counters should be zero.  You might want to be sure it wasn't a specific event like making a network change, replacing a cable, etc before you chase it down.  Is it incrementing?  Steadily?


  • @Derelict:

    Ethernet error counters should be zero.  You might want to be sure it wasn't a specific event like making a network change, replacing a cable, etc before you chase it down.  Is it incrementing?  Steadily?

    Yes but it's not every day, the only network change is clients on and off

  • LAYER 8 Netgate

    I would want to find it.  See if there are any errors in the switch.  Replacing the cable is the easiest place to start.

    Regarding changes I'm talking things that touch the physical layer.  Plugging and unplugging, rebooting, etc.  And things on other switch ports shouldn't matter - just the specific switch port and the connected LAN port.


  • Interesting to note on my PFSense.

    WAN	LAN	ADMIN
    Packets In 1458563363	1353600630	1397530
    Packets Out 1345103593	1457907316	122592
    Bytes In 968.42 GB	1090.12 GB	310.61 MB
    Bytes Out 1083.66 GB	972.47 GB	8.90 MB
    Errors In 42543	36801	0
    Errors Out 0	0	6
    Collisions 0	0	0
    

    Switch statistics

    Port	Received Packets w/o Error	Received Packets with Error	Broadcast Received Packets	Transmitted Packets w/o Errors
    WAN	2170231073	0	721104	1345095203
    LAN	1457632636	0	6033	1349808440
    

    Kind of funny that the NIC has errors on both ports, inbound, but the switch has none. PFSense has 90 days of uptime, same as my switch because of a power outage, , maybe I was messing around with something during the past 3 months that indirectly caused "errors". That's an average of 0.00292%. I just stopped a ping against YouTube that has been running for 296.74 hours, which only had 318 losses over 2136490 packets, which is 0.014884%.

    Seeing that the ping was a full round trip, to make it comparable, I have to split the average across all interface. 8 hops, which means 16 interfaces, there and back. 1.014884^(1/16) = 1.0009238 or 0.0009238% loss per hop. Seems my per hop loss to youtube is about 1/3 the loss from my switch to my firewall. Something else is at play, it's not randomly distributed.

    I'm not trying to hijack, just trying to add other view points with an example.


  • over a terabyte in 43 days??!?  :o

    Dude!  8)

    But seriously-  Id wonder why as well…

    Put a simple switch between the two ports and see if the errors stop.

  • Netgate Administrator

    So all the errors are between the switch and the LAN NIC. You swapped that cable I assume?
    It could be something obscure inducing interference in the cable. Does that cable run past any appliances/machinery etc? You could put an STP cable in there.
    Is it a Gigabit connection? Does it need to be?

    Steve


  • Think I have found the error, it was the onboard NIC (LAN) on my PfSense box.
    After I had tried to switch cable and lan port on the switch.
    I threw a new second PfSense box on and the error disappeared. had unfortunately not a new NIC card to test with,
    but think I'll try that sometime or maybe a fiber NIC to my switch. Does anyone have a recommendation on a 1Gb /s fiber NIC?


  • i wouldn't bother with fiber nic's. They are expensive ($200-$2000) and offer no benefits over short distances.

  • LAYER 8 Netgate

    Fiber can take errors too if dirty, etc.  Sometimes they even just stop working.  Copper should be clean to 100m.


  • Thanks

    I think i will stick with copper cable.