Yes, you can divide up the switched ports in whatever way you want, it's managed switch. You can even use port VLAN mode to effectively couple the two internal interfaces to two Eth ports directly. That allows you to treat them as real NIC ports and see all traffic on them. But obviously only at 1G. Steve

Hitting 100% packet loss like that is unlikely to be just latency reloading the rules. Is that the system or gateway log? Is there anything else in the system log at that time? It looks more like, say, the modem rebooting. Does it have an uptime you can check? Steve

@stephenw10 VLAN is exactly what I meant, thanks! Have not had my coffee yet. Steve

The SFP adapter did the trick. The activity LED on the WAN RJ45 socket also still works. I did a factory reset and am currently restoring settings from backup.

@stephenw10, Excellent! Ticket submitted.

@bmeeks - thanks for your furthering. It makes so much sense. Thanks.

This is the procedure I use. The reset button is the upper of the two buttons on the side, the one more recessed. The reset is a two step process: Power on the device. After a few seconds, when the green circle LED changes from orange to blue, hold the reset button for 5 seconds. This 'short-press' initiates the reset. Then, after the drive has mounted, the system recognises the reset has been initiated and asks you to confirm the reset. In 22.01 this is indicated by all three LEDs turning red. Hold the reset button until all three LEDs start flashing, ~13s. The system then resets the config and reboots. Steve

The update steamlined the PXE boot options. Previously it tried to PXE boot on a number of interfaces and had to timeout on all of them. Steve

@rcoleman-netgate yeah it has something to do with the latest Windows updates. Which is always ugly. And unpredictable. And frustrating. Not a hard fix, but it is a manual fix.

Yes, the throughput on the 2100 is limited by the CPU so you will see usage proportionate the rate. You will see though that the CPU usage is not equal on both CPU cores. So the limit is reached when one core is at 100% use as each NIC has one queue. Thus the total CPU usage may not show 100% at max throughput. Steve

@stephenw10 sorry i just noticed that , thanks :) I was sure it's the link to the NIC + mounting kit -- Edit -- unfortunately none of the EU partners are shipping this product and getting it shipped from US costs more than the product itself :)

@stephenw10 Nice, will do some testing and enable that before thunder season. :) Thanks, Steve

@gdzbedfz ACS is Auto Channel Selection on the linked page https://support.eero.com/hc/en-us/articles/115000825206-What-advanced-features-do-I-lose-access-to-if-I-put-my-eeros-into-bridge-mode-

Thank you Steve, i switched from my SG-3100 and am using the modified config.xml. Now I updated to 02.02.00.00t-uc-14 and finally to 02.02.00.00t-uc-15. The appliance starts faster compared to the old firmware version.

Hmm, that's weird indeed. I hate that sort of issue. Glad it's not affecting you any longer but let us know if you see anything further. Steve

Here is my setup, I am using a LAGG bond for dual 10G to my 16-XG: [image: 1651679478838-d35c99e8-5f91-43ce-b00b-22bab9d2357d-image.png] [image: 1651679523652-da233669-286c-428e-9c89-e0426b6129b8-image.png] [image: 1651679506743-86fac11f-721a-454e-942a-89d03cbe92d5-image.png] [image: 1651679697884-f8d0c1ff-22fb-4606-9fa5-212633da03fb-image.png] [image: 1651679673364-d2eacc5d-4adc-4003-be13-40db982978c0-image.png]

We have no way to know for sure without testing it. It may work but the chipset has some limitations that restrict SFP module compatibility and that makes it less likely. Steve

@stephenw10 gotcha. I just posted over there.

Indeed, you've been here almost as long as I have, so you're probably good. Steve

@luckman212