This worked fine - bit daunting not knowing when the upgrade has finished but my faith in pfsense has been restored.

PS, Great work on the DHCP lease deletion script available in this build I just upgraded to Jimp  ;)

The package manager works for me with WAN on VLAN.

1.2.3-RC2 built on Sun Jun 21 09:15:22 EDT 2009
running on Dell PowerEdge R200 - bge(4) NIC on a ZyXEL GS-2024 switch.

It may be worth probing what is going on with Wireshark or similar (it may help if your switch supports port mirroring). It could be an MTU issue.

If you're using a bce(4) NIC, the driver in 7.2-RELEASE has various issues. There's a fixed version in RELENG_7 that could be worth having in pfSense based on 7.2-RELEASE; a backport of the new driver to 7.2-RELEASE is available at http://people.freebsd.org/~delphij/misc/193358.tar.bz2 (untar and apply over the source tree).

I get this on my ALIX too.

http://doc.pfsense.org/index.php/Booting_Options

Disable ACPI from the shell ( Option 8 ) on console and issue following command.

echo "hint.acpi.0.disabled=1" >> /boot/loader.conf

I believe if you do a firmware upgrade you will have to reissue this command as it will be overwritten.

Not sure about #1 and #3, but #2 is normal. Those are populated with AJAX and I believe revert to undefined when the updates are interrupted in some way.

Either way, that one is normal and not unusual.

Just to update on the issue ( in case it might be interesting to others experiencing the same) I must say that the issue is still present in 1.2.3 from 0616.
Yesterday again there was a storm in logs with the stuck beacon message but its hard to tell whether or not AP was down at the moment when it happened.
Luckily it recovered and is working ok for now.
After some googling it looks like the problem is in the driver and most likely won't be solved anytime soon :(
See for example :
http://www.daemonforums.org/showthread.php?p=24538
or
http://markmail.org/message/mvqea7gke52cexpd
or even
http://kerneltrap.org/mailarchive/madwifi-devel/2008/9/9/3237624 (not sure whether its related though).
There is some patch that is supposed to fix this but the patch is for Linux.

I wonder how are things in 2.0 with FreeBSD 8?

I was never able to enable the service in the GUI.  Reverting back to an older snapshot worked.  I tried the new snaps as they were released.  Due to lack of any feedback, I have no idea if there was actually a fix for this problem or if it just mysteriously fixed itself.  ???

This update 'pfSense-Full-Update-1.2.3-20090616-0459' seems to run very well on my XenServer.  I getting ready to upgrade to Citrix XenServer 5.5 it just released.  I give everybody a update once complete.  I doing it this up comming weekend when I get home.  I got ton of work scheduled for the weekend.
RC

Ok, the patches made it into RELENG_1_2 so they should be in the next 1.2.3-RC snapshot, or the one after.

If the builder had already started before the code went in, it may not make it in this run.

Excellent, thanks all. Sounds like things will be just fine.

Even a small performance hit would be acceptable, just worried about something major.

Use 208.67.222.222 and 208.67.220.220 (openDNS)

I have to disable sticky, msn and bank sites not work  :'(

At the moment, you must manually change the auto updater from 7_1 to 7_2, than it works as it should!

I reported this issue nearly 18 months ago - it's ticket 1596 in cvstrac.

I'm hesitant to add any more AJAX stuff to the dashboard package for 1.2.x, it's getting pretty slow and burdensome as it is.

It may already be capable of that in 2.0.

Should be fixed now (not by me, I just passed it along)

My states are also back to normal on my embedded box.

Thank ermal for fixing it, he's the one who put the patch in that corrected it :)

@Darkk:

For months I'd have to re-install twice to get the 2.0 Dashboard back.  Now with the latest snapshots I just reinstall it and it comes back normally.  :)

Yeah, I think I finally got it sorted out. PHP's APC cache was holding the old index.php in memory, and I was removing it with an exec call to rm, which didn't update the cache. I replaced that with a PHP unlink() call and that seems to make it handle the transition.

This sounds a lot like a hardware problem of something failing, rather than FreeBSD not being able to deal with some features enabled in the BIOS.

Have you tried a full pass or two of memtest?  CPU fan(s) are OK?

I just had one machine that turned out to be a bad power supply: it would reboot randomly.  (No kernel faults, no messages, just reboot as if the reset button was hit – Sometimes even in BIOS setup screen).  Did you add the two NICs recently?  Maybe their power draw pushes the board too close to the power margins.

regards, ....... Charlie

Hi,

You can disable the hwtagging (hardware vlan tagging) bij issuing a ifconfig <if>-vlanhwtag, for more information see the FreeBSD manpages of ifconfig.
This can be done with a shellcmd.</if>