@MacUsers: Nope! no use. It stopped working again  :'( As the slot warms up, it may relax into a non-working state - hence the zip tie "lasso" around the base of the slot. In one scenario the ram slot was mostly broken on the side next to the good ram slot - I resolved that one with a lollipop stick folded in half to provide pressure on the broken side of the slot supported against the good ram slot. You may be able to obtain the same result with a large binder clip that fits over the ram module and still has room for the lid to close  :-\

yes, exactly! I completely misunderstood the diagram. As you said, it was completely opposite to what it's supposed to me. Once corrected all started working as expected.

@opjohnny: @Jason: The Celeron 1037U is actually faster than the Atom D2550.  Significantly so in some workloads. Isn't it really only faster on memory-bound applications?  The Atom has a higher clock out of the box, but the Celeron has better memory bandwidth and more instructions. Clock speed is a poor measure of performance.

Why do I just want to scream TROLL!!!! Seriously, by all reasonable accounts the Wiki page referenced seems to be referring to any variant of *NIX, including any of the BSD variants.  Are you going to write out every variant of *nix every time you reference this stuff?  Your posts will end up two miles long on the first mention. Now, if you have specific knowledge as to why this particular issue is not an issue with BSD variants, or pfSense in particular, I'd love to hear it.

PFW600 I resolved the issue. Apparently something happened to either the LAN switch (HP Procurve) or the firewall. Half the ports on the HP were not communicating as well as the firewall being unresponsive on LAN and WAN ports. Removing the ethernet cable between them and power cycling restored operation on both. Everything appears to be working normally again. Not sure what happened there.

ok silly there were some its of paper that came in the box - doh… so its pfsense,org/activate thanks again GX

:) Solved it myself: Diagnostics -> Edit File -> "/usr/local/pkg/lcdproc.inc" change "localhost" for "127.0.0.1" to fix bug  (where LCD screen only showed "Cli:0 and A=Scr:0" or "-> options") Works, but still has syslog error messages when shutting down: "LCDd: sock_send: socket write error" "LCDd: sdeclcd: cannot release IO-permission for 0x378!"

Wow lots posts.  Well at idle nothing connected default Mbufs setting and 8 NICS all igb driver. MBUF Usage 68% (17976/26584) Memory Usage 3% of 8129MB Same as above with 1M Mbufs reports MBUF Usage 3% (26326/1000000) Memory Usage 4% of 8129MB I think will just leave it cranked up like virgiliomi said monitor it and adjust accordingly. Will report back my results after some time.  Deploying router tomorrow. It will be running 70-100 devices, dual wan combine 200/150Mbps, 3 VLAN's and 3 dual band UniFi AP's. Thanks a lot for all your help.

@stewie2016: @JimPhreak - may I PM you? Of course.  I'll be home in 45 min and will answer then.

@Sebastiii: @BlueKobold: @thiagomespb: the pfSense 2.2.2 has support for the edge router? it would be interesting, as the freebsd 10 mips already runs .. anyone know if we can test .. That there is a fork from FreeBSD 10.x for mips means not that  there is also a pfSense fork. Hi :) Can you link to both fork ? Thanks :P For the fork of the mips based FreeBSD you might be asking @thiagomespb, he was telling about that FreeBSD 10 is running on mips based UBNT EdgeRouter and not me, and for the pfSense there is not a fork up now, as I am right informed!

Yes, it's covered in the thread above. It was a hardware defect. The pfSense folk replaced the unit and all is fine. @stewie2016: Did you determine what was causing your issues?

@clugo633: theres also this one for $20 more…. http://www.amazon.com/Samsung-850-mSATA-2-Inch-MZ-M5E120BW/dp/B00TGIVQ4G/ref=sr_1_10?ie=UTF8&qid=1430176491&sr=8-10&keywords=Toshiba+mSATA Necessary?  Could use that 20 on something else. It depends on what you want to do with your firewall.  I have 50 users on network running through an APU with squid caching and logging, squidguard blocking and logging, and lightsquid keeping an eye on all of them.  It has 4GB RAM and the 30GB SSD.  I only use about 4GB for squid's cache.  It works very well.  I never run into out of space issues. If you want to save the $20, then save the $20.  If you want to spend it on a bigger SSD, then do so.  There really isn't a wrong answer.

Thanks Evad! That might be the case. I'll give it a try to see if that fixes the issue at all. Best!

I have the APU1C on a Comcast 50/10 pipe.  Works like a charm, even with squid and an openVPN back to the office.

Hi. I set it to maximum. # sysctl -a | egrep -i 'hw.machine|hw.model|hw.ncpu' hw.machine: amd64 hw.model: Intel(R) Celeron(R) D CPU 3.33GHz hw.ncpu: 1 hw.machine_arch: amd64 # cat /var/log/dmesg.boot | grep CPU: CPU: Intel(R) Celeron(R) D CPU 3.33GHz (3333.54-MHz K8-class CPU) I have done the same to my test machine x3650. I'll be looking over a packet capture in the near futur and post back. Thanks for your interrest. I'm still looking for a solution.

@valnar: Yup.  http://www.cincinnatibell.com/internet/#plans Never before have I wanted to live in Cincinnati…

matthackett, what setup did you go with?

Funny you say that. I been researching Wimax and maybe with that going tits up here in the US we will get that band one day??? Doubt it with those money grubbers in charge.

Looks like I got the nmedia display working finally, you have to make some edits to allow 38400 as an option for the port speed.  Here are the steps I used: From: https://forum.pfsense.org/index.php?topic=83747.15 Install lcdproc-dev 0.5.6 from pakages drop to freebsd console (or use Diagnostics\Command Prompt in the GUI) mount -uw / pkg install sysutils/lcdproc (For editing you can use console or Diagnostics\Edit File) Edit /usr/local/pkg/lcdproc.xml -Search for 19200 you should find: <option><value>19200</value> <name>19200 bps</name></option> -Under it add: <option><value>38400</value> <name>38400 bps</name></option> -Save this file Edit /usr/local/pkg/lcdproc.inc -Again search for 19200 you should find: case "19200": -Under it add: case "38400": -This will enable 38400 option for port speed, but looks like if you pick the hd44780 no matter what it is going to add a Speed=0 to the configuration file. -Search for: $config_text .= "Speed=0\n"; -Right above that line add: if ($lcdproc_config[port_speed] == '' || $lcdproc_config[port_speed] == '0') -Now save the file. -In the Services LCDProc choose these options: Comp Port: mine is /dev/cuaU0  set this depending on your config Display Size: 2 rows 20 columns Driver: hd44780 Connection Type: Mplay Blast from VLSystem Port speed: 38400 -Save and restart the system, this should get it working. -For some reason mine required a reboot or restart of LCDd (one extra time) to work properly.