I live in a southern climate that's not exactly cool, especially during the summer. I have a garage which is not airconditioned, just passive vents to the outside. There I have an enclosed 16U cabinet only with a 120mm exhaust fan at the top. In there I have a 1U machine with Intel motherboard with Atom CPU, I believe D410, 16GB SSD, PicoPSU and dual gigabit Intel NIC (PCI) consumption is a little over 10 watts at the plug.

There is also no fan in the 1U system not an intake/exhaust fan or a CPU fan. My intent is to place a fan where the traditional PSU has been placed, but I have not gotten around to it yet. This has been like this for almost a year and never had an issue with overheating, lockups or shutdowns.

Like I said I "need" to add a fan to the pfSense and I also have an additional 120mm fan to add the the enclosure and also a temperature controller so the fans don't run 24/7 (the power consumption of the cabinet fans and also the system fans quickly adds up, I want to say it's around 5 to 10 watts per fan) but so far it has not given me any problems.

How many active users/devices?
Prevailing weather conditions? (temperature range)
What type of wifi clients (2.4GHz or also 5GHz)?
Noisy environment or in the middle of nowhere?

Nearly all major manufacturers offer outdoor APs, but feature-sets can be quite different from one another…

@MMacD

Installation was not a problem. All authors describing their problems with the serial console have managed a succussful install, e.g. I did boot from CD and USB to manage a full install, flashed nanobsd, nanobsd_vga to obtain an embedded install. Instead, activating the serial console after a successful install was not fully possible. And the straight nanobsd images (e.g. without VGA) did not boot at all on my machine. This affects obviously at least some motherboards. Please find further details in the current thread and all referenced ones if desired.

A working serial console is valuable in particular, if you run your system without monitor and keyboard (even if you could), to

interact when booting, e.g. to select boot slice see boot messages and possible errors access your system, if accidently locked out from SSH or webGUI due to rule error

Furthermore, at least me would like to exclude any misconfiguration before reporting this behavior as a bug.

Hope this helps to clarify :).

Peter

Try getting a hold of the Pantech (UML 290) Extended AT commands document and look into setting the frequency bands properly.
In many cases these are not properly configured so when a dissociation happens it could fail over to another frequency band which might not be supported by the wireless provider.
Also look into the reset commands, they could potentially bring back the modem to the initial connection state.
You can also try getting an usb sniffer for windows and check the AT commands upon a successful connection, if you're lucky you might be able to see some useful information. (in case the connection is not done through QMI, depending on QMI support and implementation)

That's more like it. I wrote a quick patch for the pfSense image build process:

–- ./sys/dev/usb/serial/u3g.c 2012-03-06 14:37:00.000000000 +0100
+++ ./sys/dev/usb/serial/u3g.c.new 2012-06-03 18:19:07.000000000 +0200
@@ -460,2 +460,3 @@
U3G_DEV(SIERRA, E6892, 0),

U3G_DEV(XXXXX, XXXXX, 0),
U3G_DEV(SIERRA, E6893, 0),
--- sys/dev/usb/usbdevs 2012-06-03 18:26:29.000000000 +0200
+++ sys/dev/usb/usbdevs.new 2012-06-03 18:26:22.000000000 +0200
@@ -2992,2 +2992,3 @@
product SIERRA E6892 0x6892 E6892
+product XXXXX XXXX XXXXX XXXXX
product SIERRA E6893 0x6893 E6893

Replace the XXXXX with your device information. Save as broadband.diff and place it under /home/pfsense/tools/patches/RELENG_8_3
Edit /home/pfsense/tools/builder_scripts/patches.RELENG_8_3 and add the following line:
~~broadband.diff
below ~~huawei_k4505.diff.
run ./build_iso.sh (the last stage) to build the images. (The build process will fail the first time, perform a cleanup and re-run ./build_iso.sh.)
That will re-fetch the kernel sources, re-apply the patches and build the images.
It should work for all Sierra and Novatel devices, and for the majrority of HUAWEI and ZTE.

This applies to the pfSense build image process which is described at: http://devwiki.pfsense.org/DevelopersBootStrapAndDevIso

Well, can confirm that Intel 82574L is working well with pfSense 2.0.1. This chip is attached on my Jetway NF99FL-525.

you wont find those 2 files (u3g.c or usbdevs) in a regular pfSense image. a regular image has a pre-compiled kernel, kernel source files not included.  the only way to have access to those files is to compile pfSense from source, process which requires a regular FreeBSD install and source code downloaded (for both FreeBSD and pfSense). read that post, it tells you how to do it.

There's some good info in this recent thread:

http://forum.pfsense.org/index.php/topic,53273.0.html

Basically, sure, you can easily build a full pfSense install on a modern SSD.  For the most part, even if you go fairly "cheap" on a modern SSD and you run a fairly robust feature set, it should still last a year or 2.  If you scale back your features and/or raise the bar on your SSD, it should last longer.

Your idea has merit, but with a modern SSD it's probably not as much of a problem, especially if you go with an "Enterprise" or "Industrial" SLC based SSD.  That should last long enough that other parts may be your age limiting factor before the SSD starts having problems.

Back to your idea having merit, though, I could imagine some kind of log buffer/commit cycle could be handy.

You should be able to build a custom pfsense image with the added support for MC7710.
http://technotation.com explains you how to do it.

It's because that script was originally written for an embedded install of pfSense 1.2. The embbeded install (and the Nanobsd install) runs with the file system mounted read-only to minimise writes to flash media. The  script remounts it as RW, installs the various files and then remounts it RO again. This won't work on a full install that you are running.
Besides that as I wrote at the top it's now very out of date.
Try this post instead. I can't say if that will work any better since it's still writtenb for Nano but I changed it to use the correct mount script rather than attempting to mount the filesystem directly. I have never tried it on a full install.
If that still doesn't work that you can just edit the install script, install3.sh, and remove the remount calls.

Steve

How much traffic are you passing at that time?
using top -SH is a better measure in my opinion. If you open an ssh session and run top -SH and then open the webGUI to view the dashboard you'll see the CPU usage go up considerably. I.e. simply viewing the dashboard on a system of that type uses CPU cycles. On my own low end P4 it uses quite a lot!

Which options in System: Advanced: Networking:?

Steve

Opps. sorry @wallabybob.

@wallabybob : DHCP server is at VLAN10, 20,30  in PFSENSE which located at my Macmini VMware.. maybe i should make a new post for this. thanks..

@Metu69salemi: Thanks .. ill try that

@proverbian:

is this the structure?

Pretty much. I would plug ONE pfSense network port into switch port 1 and that then becomes the physical interface for the three WAN VLANs. I would then use the other pfSense network port as the pfSense LAN port and that could be connected to a single computer or to one port of a dumb switch or even to (say) port 5 of the GS108T with ports 5 through 8 all members of VLAN 400 and configured in the switch to add VLAN tags on entry and strip VLAN tags on exit so ports 5 through 8 look just like a non VLAN LAN and three computers can be connected to switch ports 6 through 8.

I think it is useful to have the pfSense LAN port as simply configured as possible (no VLANs) so that if you mess up our pfSense configuration and have to recover by restoring factory defaults you can do so with configuration information and minimal changes to your surrounding infrastructure.

If I had that box I would try, carefully, to remove that heatsink to find out what's underneath it.

You can also look at the CF card for clues as to how Watchguard's OS (Linux) talks to the card.

Steve

http://www.newegg.com/Product/Product.aspx?Item=N82E16833127219
For PCI 32bit.

or

http://www.newegg.com/Product/Product.aspx?Item=N82E16833127218
for PCI-e

Or check the one you want to buy at www.wikidevi.com to make sure it is atheros based.

I make kernel module for monitoring cpu temperature
https://dl.dropbox.com/u/75811597/viatemp.tar

@l2i28v5i:

I'm fine with the config as it is as I don't need DHCP on the WAN side.

Then why did you configure the WAN interface to ask a DHCP server for configuration information?

@l2i28v5i:

Perhaps someone knows why it's doing what it's doing.

My GUESS is that dhclient doesn't get an answer to its DHCP request in "reasonable" time and resets the interface (link down/link up). There might be evidence in the system (see Status -> System Logs) to support this - look for reports from dhclient.

The 2.1 beta version is based on FreeBSD 8.3, and there have been reports of that NIC working there. You can get a snapshot of the beta from http://snapshots.pfsense.org/

A release for that is still a couple months away at the earliest, but it's stable enough that there are many people using it in production, especially for newer hardware.

Thanks, but I am using virtual machines in VirtualBox.  :-[
Something I didn't mention maybe.

It doesn't look good does it.
No worries. I have a spare hard drive so I'll install 2008, borrow a switch and use my laptop as client in your configuration.