Hello,
Since connecting two interfaces of an opnsense device to the switch - with different vlans - in errors for both interfaces are reported. The monitoring (checkmk) shows a constant error rate - 0,5 per second - regardless of the traffic for each interface. Does anyone have an idea what could cause this issue?
Thanks for any help!
ix0:
Status up
MAC address 00:90:0b:7f:84:b2 - LANNER ELECTRONICS, INC.
MTU 1500
IPv4 address 10.2.0.1/26
IPv4 gateway 10.2.0.1
Media 1000baseT <full-duplex,rxpause,txpause>
In/out packets 79509368 / 79575887 (5.76 GB / 5.48 GB)
In/out packets (pass) 79508679 / 79575887 (5.76 GB / 5.48 GB)
In/out packets (block) 85929 / 0 (689 bytes / 0 bytes)
In/out errors 2199811 / 0
Collisions 0
ix1:
Status up
MAC address 00:90:0b:7f:84:b3 - LANNER ELECTRONICS, INC.
MTU 1500
IPv4 address 10.2.8.1/24
Media 10baseT/UTP <full-duplex,rxpause,txpause>
In/out packets 125031641 / 112825265 (92.06 GB / 84.11 GB)
In/out packets (pass) 124958473 / 112825264 (92.06 GB / 84.11 GB)
In/out packets (block) 7824309 / 1 (71 KB / 60 bytes)
In/out errors 2196895 / 0
Collisions 0
Following as I've spotted many errors myself as well (the LAN interface is insane!!!) and I cannot understand what is the underlying cause of this???
(//)
Maybe a bad/incompatible SFP module or cable? ix1 in your readout shows 10baseT. I find it odd that a 10GB capable card is negotiating at 10baseT?
Thanks for your comments.
@NW4FUN
The main difference to your scenario is that i only log errors on incoming packets. The errors occur at a constant rate of 0,5 per second.
In/out errors 2199811 / 0
@opnfwb
Actually the limited bandwidth was the reason for that setup. Since i considered the two parallel links between gateway and switch as problem, i changed the setup yesterday. Now there is only one physical link with one more vlan and i used traffic shaping rules for bandwidth limitation. I keep getting the errors for the remaining interface.
The switch is an older 3com (now HP) model (3Com Baseline Switch 2952-SFP Plus Software Version Release 1519P06). The used Port is not one of the SFP ports.
Hmm, when you say two parallel links are you running these in a bridge? It's been mentioned before that a bridge with unassigned interfaces seems to cause a constant increment in errors.
https://forum.netgate.com/topic/77845/solved-in-out-errors-on-lan/23
No it is/was not bridged, i used different vlans. At the moment the second interface ist disabled at the switch but still connected to the gateway - all i could do remotely. The remaining interface in use keeps logging the errors. As soon as possible i will remove the cable and see if it changes anything.
I finally could stop the error by disabling stp on that switch port (it is an older 3Com/HP). I got the right lead from this thread:
https://community.intel.com/t5/Ethernet-Products/X520-DA2-FreeBSD-rec-len-errs/td-p/579676
Nice find! Thanks for posting a followup with the solution that worked.
Hey guys,
I implemented your proposed solution by disabling Spanning Tree (not sure is the safest thing to do in any case) on my Meraki Switches (MS-125p), however, interfaces are still showing errors on ax0 for both LAN (physical interface) and its VLANS (LAN being the parent interface).
Any words of wisdom anyone?
This is annoying and only happens on ax interfaces, igb are looking good (with or without STP enabled)
Cheers,
NW4FUN
@NW4FUN
I think my case differs from yours - allmost all errors where lenght errors. So my solution is for a very special case and may not help you - sorry.
dev.ix.0.mac_stats.checksum_errs: 17
dev.ix.0.mac_stats.rec_len_errs: 805211
dev.ix.0.mac_stats.byte_errs: 0
dev.ix.0.mac_stats.ill_errs: 0
dev.ix.0.mac_stats.crc_errs: 0
dev.ix.0.mac_stats.rx_errs: 805228