OPNsense Forum
English Forums => General Discussion => Topic started by: firewalled101 on October 04, 2022, 05:36:10 am
-
I have a TOPTON device with 4 ports. I followed this guide https://docs.opnsense.org/manual/how-tos/lan_bridge.html to create a bridge interface close to the attached diagram. I noticed that I have errors out on the bridge interface and VLAN10. The physical interfaces OPT3 and OPT4 are enabled but unused at present. I tried to disable unused interfaces and it did not solve the issue. I attached a diagram for the interfaces and screenshot from my OPNSense instance. Shall I worry about these errors at all?
Thanks.
-
The same problem. TOPTON device, intel i226-v. LAN-bridge configurated like in manual
13296 errors out : 21,543,583 packets out
I tried to reconnect all devices, it does not help
-
I have a similar device, but I'm running Proxmox on it, and OPNsense inside a container. Proxmox is providing the bridge.
It's not stable. I've stopped connecting 2.5 Mbps devices to it and run them through a GbE switch first. Even so, periodically one of the NICs will lock up, cause some kernel messages and it require a reboot. There's a lot of worry that there's a hardware flaw in the Intel i226-v drivers and it's not certain if it'll be fixable by a driver update.
All of this may or may not be related to the particular issue you two are seeing, however.
-
I have CWWK device with 6 Intel i226-V NICs, with 5 of those configured as a bridge. I don't have VLANs in my setup and also see the Errors Out on the bridge interface. Network conditions do not seem to affect it and it constantly around 0,4% - 0,5% of the total packets out. Any hints on how to debug it would be much appreciated
-
I hope you realize that what you own is not a switch. you are going to lose tons of frames and you should be ok with it. :)
-
I know that it's not a switch and I'm accepting potential downfalls of using software bridge as a switch. However I don't understand why should I expect lost frames regardless of the load? Also I don't see actual lost frames anywhere. When I stress test using iperf I'm yet to see a single dropped frame. That's why I'd like to know where does those errors actually come from.
-
I've been able to solve it on my part. This turned out to be an user error. I have WiFi Mesh on the same network with the wired backhaul. In some very specific conditions I've noticed that sometimes there had been up/down events on the interface connected to the unmanaged ethernet switch with loop detection. One of the WiFi APs had been connected via this switch. This pointed me towards the probable cause of the drops, and I've enabled Spanning Tree Protocol on the bridge interface. Since then I have 0 errors on the bridge.
I just wish there had been an easy way to just know what is the reason of the error in the first place.
-
The i225 and i226 chips are known to have issues, particularly as embedded devices. Try connecting to a 1g port and see if the issues go away. I've seen some reports that changing out the device you connect to or adding a switch between it and the device(in the case of modems, etc) can fix things.
Unfortunately, I can't really provide more info than that as my PCIe i225 has been solid.
-
To me this stopped happening when I connected unused LAN port.