OOh thanks I miss that point, it's very closed but I cannot see. Thank alot I will try. @SeventhSon: Couldn't you use Reauthentication every minute? Then you don't have to send anything, pfSense will check if the user still has access, otherwise disconnect the user.

Hmm, well it is hooked up to a UPS. I'll try it without the UPS. EDIT: same thing either way : /

Source can be found under the Development heading of the home page and more pointers can be found in the development sub-forum. Obviously you can make changes, you just need to ensure you comply with the licenses that come with it.

you can always install NRPE, then create your local check inside of the pfsense box, and get the data from Nagios using NRPE

This is a post I wrote on my blog. This is one working solution. http://blog.magiksys.net/node/1 Regards. Alain

If I recall correctly, ntpdate is supposed to run during startup to sync the clock with the specified reference clock. Then ntpd runs to nudge the local clock from time to time to keep it in step with the reference clock. ntpdate won't act if ntpd is running. See the FreeBSD man pages at http://www.freebsd.org/cgi/man.cgifor more information on ntpdate and ntpd. From what you have said it seems likely the hardware clock is way out of spec. I have a recollection a PC clock shouldn't drift by more than about a second a day. The software may not be able to cope with a clock so far wrong. Depending on the CPU and chipset, it is sometimes possible for FreeBSD to choose an alternate clock source which might be more accurate than whatever is being used now.

yea I now suspect that the CF was exactly the failure had issues writing the beta5 to it chucked it and got a new one.. this thread can be killed :)

@wallabybob: @BlueToast: I found a problem with this. The dnsomatic API documentation at http://www.dnsomatic.com/wiki/api says in the section Update Parameters "To update all services registered with DNS-O-Matic to the new IP address, hostname may be omitted or set to all.dnsomatic.com (useful if required by client)." Ok, yes, this worked. Thanks!

It is a highly critical datacenter (healthcare systems with critical patients at the other end of the wire) splitted on two sites (10 minutes by car) Two sites but one network, I have full layer 2 continuity between the sites, over two physical path. Sites are in active/active mode : same network servers (physical or vsphere VM) are in active/active or active/passive failover SANs (mostly VMware datastores) are replicated with on-the-fly failover(datacore) over 8gb/s fiberchannel links (two fabrics per site, two physical path) Internet and mpls access are provided through optic fiber with layer 2 continuity/failover (plus copper backup) between the sites. Internet peering is also secured with BGP failover over another ISP (copper + fiber). At this moment I have only one firewall per site, I'm thinking of how to have two firewall per site.

We have an ALIX board causing problems. It's going back for warranty repair :)

I solved the problem. Disabling CDP on the VLAN interfaces at the Cisco Router fixed the problem.

Only remove

;D Thank you very much. Using root did the trick and allowed me to transparently ssh to machines behind the firewall and using public key as credentials…very neat!

You can choose the monitor type for the load balancer entries (TCP or ICMP, I think, there may be more)

That makes sense…  Glad ya got it working...

Aha! Sorted! I've add the following line to /root/.tcshrc setenv PACKAGESITE "ftp://ftp.freebsd.org/pub/FreeBSD/ports/i386/packages-7-stable/Latest/" Works great.  ;D Thanks again Jim.

Thanks, I had misunderstood how filtering is applied but this cleared it.