@NineX: @w0w: @dotdash: I would contact Netgate. I'd guess it had the same Atom bug as the 2440 and larger Rangely boxes. If you didn't update the bios to fix it, it may indeed be bricked. IIRC, they are covering Atom failures for three years from purchase, but contact them to confirm. There is no any BIOS fix regarding AVR54, it's a hardware bug. You are only partially right. as ADI boards don't use SERIQ bus , AVR54 can be mitigated by powering off this bus. and that exacly bios update i think .12 is doing. It's not solving problem permanentny but prolongs life of affected CPU. Exactly. You are right, the description of BIOS update was not clear enough and I've missed the point.

It might be able that more the one issue will be invited in that case. First of all PPPoE usage of the Internet account, this is then more single CPU core threaded and  nothing can help you out with it either another more powerful or strong CPU. The gateway device from the ISP if it will be not a real or pure modem, but more a router will be also doing NAT at the WAN port, and then you will be also able to ask you ISP what to do or how to solve this problem. Perhaps he will send you another "box" that is doing not NAT or is acting as a pure modem. The SG-4860 unit is doing for someone called @gonzopancho at a 1Gbit/s symmetric Internet connection nearly or around ~900 MBit/s and together with the TCP/IP overhead and passing firewall rules through it, it might be the nearly 1 GBit/s real throughput. If you now gets only 500 - 600 MBit/s I personally would do a fresh and full install without any additional packets and have a look what is going on then. The ADI 2.4.1 serial Image is able to download and pretty new, so it could be if you are using VLANs at the WAN port that the version 2.4.0 will be a better option at this time for you. Can you set this up for us and your self? Before doing so please do a configuration backup but don´t play it back on the fresh installation to prevent a typo or earlier set up false in the test. Normally it will take 30 minutes for you. And then you will be able to see what is going on in real. Would be my way in that case.

@SpaceBass: Ugh! People doing  construction down the road sliced a line.  My UPS ran dead in my system did not shut down gracefully. Now when I watch it boot on the console, it tries to boot with a network boot from each of the interfaces and then shows a “booting from CBFS “screen. It doesn’t seem to want to use any of the internal devices - almost like they aren’t there. Does this mean the internal storage is toast? If you're on console when the 2440 is powered up - and select a boot disk - does it come up? Been there myself, and posted a thread this evening for a similar situation - it was stuck at the bootloader. Not sure what your config is - but I have an mSATA installed on my 2440, and prior to last time around, I still had good partitions on the eMMC. See my thread for possible corrective actions if mSATA is installed.

You can find the update content information here https://github.com/ADIEngineering/adi_coreboot_public/tree/master/releases/ADI_RCCVE-01.00.00.17

Definitely is a hardware issue, though it's not clear what it is. The cpuid isn't necessarily related there. How old is the device? Even if it's out of the Netgate warranty it may still be under warranty from Supermicro

haha - OK.  Now I understand how this came to be.  An accident with benifits.

Alright, i figured as much.  :-\ Thanks.  8)

The images available in the member portal don't seem to support VGA console after the bootloader. You may be downloading the wrong image. Please contact our support and they will help you out. https://customercare.netgate.com/

Thanks Jim.  I will probably move back to pihole until this is fixed…hopefully soon, as I like being able to consolidate everything on my netgate.

There are two M.2 slots on the SG-3100 capable of taking an SSD. One is 22x42mm and the other is 22x80mm. However the 42mm slot is the only one that can take a 3G/4G modem so if you ever wanted to use that you want to use an 80mm SSD. Steve

Gold doesn't include support, it's for supporting the pfSense project and getting something in return. https://www.netgate.com/our-services/gold-membership.html

@rudivd: Hi all, Just received my 3100. (2.4.1-RELEASE) I need ravd to work reliably for it to be deployed. I have not yet tested the box with ravd, but is this bug acknowledged (and reproduceable ? ) ? Is there an estimate for 2.4.2 where a fix could | should be implemented ? Thanks ! Rudi See if the report matches your issue. https://redmine.pfsense.org/issues/8022

Cool, I ordered a couple of those power cables from the guy and I'll test this & report back!

Please allow me to investigate all the tickets and communication you've had with us and I'll follow up. Thank you!

Sounds good! We already have your information from the previous tickets so expect an email from us. Thank you!

The gui home page with all the widgets yes is a bit of cpu hog ;)

I don't think so.  I think you have to do it the way you did it.  However, you could probably boot and run the box off that usb thrumdrive. Never seen anyone try.

I received my SG-3100 on Friday, I'm still going through the pfsense handbook, but from other stuff I've read on this board and the subreddit board, it looks like I may have to wait until 2.4.2 is out in order to use it because it's going to replace a centurylink router with PPPOE and a vlan of 210.  However, at the PfSense portal, there are no ARM images or beta releases for ARM.  Also, I haven't got to this part yet, but when I get to the point where this is set up and connected to the WAN, will I be able to see all the packages available for the SG-3100 with "pkg search -g *", at least from the command line or somewhere in the GUI?

@valnar: I think this is similar to my issue. The new defaults with ZFS seem to use more (idle) CPU than before.  I haven't had time to test that theory yet with UFS. https://forum.pfsense.org/index.php?topic=138477 Yes, sounds similar. I'm not terribly worried if I know it's normal, but it does seem to be substantially more than the ~1% I am used to.