@doktornotor: @craymore: @doktornotor: These things are simply EOL. Having issues even when updating due to lack of RAM, especially if you hit some bigger package update (python, perl comes to mind.) EOL ?!  :o Yeah, these are certainly EOLed as far as pfSense is concerned (together with rest of x86 arch). x86 is gone with 2.4. I mean, recycle them to run LEDE/OpenWrt or something similar on it and they can keep running for lot more years, but pfSense/FreeBSD is getting a pain to run on these. For details of will no longer be supported from 2.4 onwards: https://forum.pfsense.org/index.php?topic=121255.0

Ah, yes no way I know of to recover from that if you're seeing those errors. A fresh CF card is the safest way anyway. You still have the old one to boot from if there's an issue. Steve

I would try re-assigning LAN to nfe1 to be sure it's not a configuration issue. You can do that from the console menu. Depending on exactly which IC that re NIC is using you may need to disable hardware off loading etc. https://doc.pfsense.org/index.php/Tuning_and_Troubleshooting_Network_Cards Steve

Hi Gertjan, tks for the reply, I think I didn`t express myself very well. When I said that I changed some configurations  I mean that I changed on the webGui interface, not on the XML file. I did that just to have sure that the recovery really has returned all the previous configurations. But, about the second question, I tried to backup just a small part and tried to recovery again, but the error persists. I tried with the DHCP area, Captive Portal, aliases, then I quitted. What I think is really strange is to not receive any message of error, even on the log file. Any new ideas would be great. For now, just to have a way to load the backup configuration again, I logged by ssh terminal and manually replaced the config.xml on /cf/conf/ directory. That works, but I know thats not the right way to do it. Well, again, tks for the suggestions. regards

My Apologies. I thought that transparent mode meant the user not having to enter any proxy info in IE.  My mistake. I will do further reading and investigation as to what NIC I have to bind to the proxy.  At this time, I am binding both of them - but in all reality its my understanding that even if I just bind a single interface, both are being joined by the bridge anyhow. In essence, by selecting a single NIC, in bridge mode, both are being seen as one by pfsense. I am just trying to understand that.

@phil.davis: And we seem to have lost our crystal ball, do let us know if you find it ;) Make sure you have power-cycled everything. Have the front-end modem up first then bring up the pfSense box. If I had a crystal ball I would be banging my head on it right now  :o After changing one setting at a time and having no luck, I did what you suggested (and what seems obvious) … power everything down, bring up the modem, then the Pf box, then the peripherals ... presto! Can you delete this post and we will just act like it never happened. lol

Great thanks. Looking forward to finally giving the install a try this weekend!

How many switchports do you need? The Cisco Small Business stuff, like SG300-10, -20 or -28, is considered capable and price worthy in here. I wouldn't use a PoE switch to power a single AccessPoint but use a PoE injector instead. A cheaper switch that might fit is TP-Link TL-SG3210. Personally I'd stay away from everything with a Netgear sticker on it.

I have never seen that work successfully. I haven't spent long trying though. But yeah, there's no easy way and you shouldn't be using flash for caching anyway. TINC is available in 2.3.3 snapshots BTW. And in 2.4 but that won't run on an X750e. Steve

If you run through the setup wizard keeping the default subnet and then change the LAN subnet afterwards do you hit the same issue? Steve

@Derelict: Netgear switches are complete and utter piles of crap. I would concur! See those ports (01-06) with both untagged and tagged VLANs on 10.0.10.12? Are those your Ubiquitis? Try also manually setting the PVID to 20 on those ports. Not having the PVID shown there makes me think that might be your issue. thanks, will give that a go..

Questions… IPMI has Ethernet cable plugged into switch? Reset (cable) modem? Can you post screenshots? I haven't had any issues with Auto-detect. No issues with Supermicro M/B until X10SLL-S and C2758.

I just tried a new install of 2.4-BETA into a VirtualBox VM on my Windows10 laptop. I had no trouble using the arrow keys, tab and enter to navigate the install menus. So it does work on, at least on my relatively modern hardware/OS.

Thank you for the confirmation - wasn't sure if I had missed it somewhere.

@jack290: As a linux user,  the way I've got round this is to connect the SSD into a linux system and partition the disk using the linux fdisk. I found that when installing, the PFS installer gets upset if the partitions are not aligned on what it thinks is a cylinder boundary,  and I find the way round this is to choose sectors which align with both 63 (the old cylinder system) and also 4K boundaries.  This requires a little use of a calculator.  There is usually plenty of room on an SSD, hence I align to 63 x 4096 boundaries, ie 258048. Typically, start 4644864  ( 18 x 258048)  finish 30965759 ( (120 x 258048) -1 )  is 26320896 sectors, 12.5Gb Mark the partition as bootable and set a free BSD type (a5).  More partitions can be created by the same method.  The partition marked bootable must be used. Put the disk in the PFS box and install. When installing, choose own partitioning, pick this partition, do not format then install the bootblocks (bootloader) and complete the install. EDIT -  BSD calls partitions slices.  The above should read  "do not slice the disk, pick the slice (partition) you created."  I'm working from memory so I hope there is enough here for you to get started. The above is a somewhat blunt approach which works for me because I am familiar with the linux version of fdisk.  It is not much help if you do not have an available linux box to use.    Someone else may offer a better solution,  if you are familiar with the BSD version of fdisk it should be possible to boot the installer to a command line,  fire up the BSD partitioner and then use the approach given above to set the partition start and end points.  ( I'm not familiar with the BSD  version of fdisk and it may insist on cylinder boundaries,  however a little use of a calculator should still allow the problem to be solved.) There are differing views on whether or not to set trim, if you want to do so, after installing, boot to a command prompt at the first menu and run:-     /sbin/tunefs -p /    and      /sbin/tunefs -t enable / to enable trim. Hope some of this helps,  and I might add that 4k is also preferred for many modern spinner disks which use "advanced format". I just do a 16MiB alignment. No thinking involved then and only wastes a minuscule amount of space.

Version upgrades such as 2.3.2 to 2.3.3 are not going to be done with 'pkg update; pkg upgrade' as far as I know because pkg doesn't quite know how to update its own repository configurations. The pkg updates are for minor updates, security fixes etc.