I suggest you guys to get the 3-eth (big) model. It has more RAM and uses a 15V supply with a standard P4 plug. Other models are powered with 5V, but uses a different (proprietary?) keyed plug. The disadvantage is the absence of a buzzer, I like it. The sound goes out by sound card only. There are plenty of eBox models, so choose it carefully, matching the specs which you want.

Presuming your Intel NIC is using the em(4) driver.  From the Manual Page: 1000baseTX  Sets 1000Mbps operation.  Only full-duplex mode is supported at this speed. I think they mean 1000BaseT, but you can't have everything. When I have gig ports talking to gig ports, I use autoselect.  If I have gig ports talking to who knows what's going to plug in, I use autoselect.  If I have something static like you have and I have a gig port talking to a 100BaseTX port, I usually hard set both sides.

thank you very much

sorry folks, the cause was stupidity! The ip address that the pppoe server was handing out was already in use.  I changed it to an unused address and everything sprang into life! Macsmith

In a quick comparison with my config, I don't have that "lport 0" setting in my clients config. Try without (comment it out by prepending #), it should no longer be necessary to tell the client to use a random port. Restart the openvpn, establish a connection, and issue a ping -t to your bsserver. If it again shows disconnections, check the output from the pfSense log (openvpn)

@stephenw10: Ah, Ok. So when you select option 3 'Boot from USB' what your actually doing is telling the kernel to wait a short time in order for USB devices to power up/initialise themselves so they can be detected. You can make that change permanently by adding the line: kern.cam.boot_delay=10000 to the file: /boot/loader.conf.local You will have to create that file if you haven't had to already. See: https://doc.pfsense.org/index.php/Boot_Troubleshooting#Booting_from_USB Steve I can't edit my old post, but incase anyone else sees this thread. It should be: kern.cam.boot_delay="10000" I just put a new box and I remember I made this thread. Just updating it since someone else may run into this issue.

@x: Hi, OK, I think I know what happened :) /31 is the answer I was looking for. pfSense do not support /31 networks ;(( Best regards, – Peter Not that it solves anything for you right now, but it's on the TODO-list for pfSense 2.2: https://redmine.pfsense.org/issues/1972

I understand. Nothing obvious that would stop that laptop booting. I don't think I've ever seen an 'OK>' prompt. Steve

@orthopteroid: The strange thing is I don't see the file snort-2.9.6.2-amd64.pbi.sha256 listed on https://files.pfsense.org/packages/amd64/8/All/ Packaging error? The very old pfSense version, 2.1.1, wasn't using FreeBSD 8.3, but a lower version. So, probably the download path, used to load basic FreeBSD packages to complete pfSense packages isn't probably valid anymore. Easy and best solution:  get the latest version 2.1.4 and your problems are solved.

Well 2.2 installs but after rebooting the server it's failing to see that PFSense is installed on the machine and it's just caught in an endless reboot cycle. Any ideas here? I've checked in BIOS and everything looks alright to me. Not getting any errors either it's like it's thinking there's nothing to boot from on the HDD.

I've had this working in the past, and I remember using the Target Category to handle the redirect.  I don't remember if I had any problems when using the Common ACL redirect.  Try creating a Target Category for your blocked sites, add the sites you want blocked and then change the redirect message.  Then set Common ACL to Deny that Target Category.

Maybe a nice TXT or PDF report with a section for each service. For example, the DHCP section would list the config settings and any reservations. The Firewall section, would list all the rules.  So on and so forth.

^Exactly. X86 only currently I'm afraid. There is a possibility of MIPS support sometime in the future (after 2.2 is released) but ARM is not even on the map! Steve

Thanks Steve, I would agree the image should be bit for bit and the file format shouldn't make a difference. Its weird, I had this issue whether I used dd or Win Disk Imager. I'm using a transcend 4GB Industrial Grade CF card. I guess that I never actually did anything to the bios at all then, from what you say. For RAM choice it was pretty simple: I took the advice of the pfSense docs: https://doc.pfsense.org/index.php/PfSense_on_Watchguard_Firebox and used the KVR100X64C2/512 module - all in all £10 from fleabay. Its been a learning curve but good fun! I'm now going to convert an XTM 505 as my main box. Thanks, Adam.

Any reason you're using 2.0-release? That's old! There are a large nuber of threads about how to get past 'stuck at 38%' if you find that an issue. Some people found it eventually went past it others recommend disabling and CPU virtualisation features and any multicore features in the bios during the install. They can be re-enabled for normal running. Steve

To anyone that has read this and might have ran into the same issue… The auto detected IPV6 network IP addresses that were assigned to the interface didn't have a gateway that was reachable and so apinger was going bananas trying to constantly reload the configuration. When I disabled iPV6, utilization went back to normal and re-enabled and properly configured the IPV6 addresses on the WAN it fixed the issue... Hope this is at least helpful to someone else as the logs are not very clear as to what's happening.

So it is now online - this connection is setup by the pfsense vm. I preset all the forwarding rules while the router was not connected to the internet but was connected on the LAN side to an open IP address. To put it online I removed the LAN and WAN connections from the old router. Plugged in the WAN line into the unused but configured interface and turned on DHCP and moved the LAN gateway from the open IP address I used to do the configuration to the same IP as the old routers Gateway.