Losing WAN overnight

jerryaycock · Dec 5, 2014, 3:22 PM

I get up almost every morning and have no WAN connection.
I am on a new install and no packages other than CHRON .
Very BASIC setup. When it is up it is awsome , How can I see what is happening ?
A quick reboot clears it up.

2.1.5-RELEASE (i386)
built on Mon Aug 25 07:44:26 EDT 2014
FreeBSD 8.3-RELEASE-p16

Dell optiplex GX 270
st-1023 gigabit ethernet cards

Thanks
Jerry

stephenw10 · Dec 5, 2014, 3:25 PM

Check the system logs. Check the ppp logs if you're using PPPoE.
What WAN type is it (PPPoE, DHCP, static)?
What device is connected upstream? Does that have logs?

Steve

jerryaycock · Dec 5, 2014, 4:01 PM

WAN is DHCP
Cable modem No Logs it appears to be up Never had problems with it before.
I have cleared logs will copy next failure.

Thanks
Jerry

KOM · Dec 5, 2014, 4:12 PM

Also check the Gateways log to see if apinger is biting you.

kejianshi · Dec 5, 2014, 4:42 PM

Yeah - On two seperate systems I tried two solutions.

System 1 that gave issues, I turned off gateway monitor.

System 2 I raised the thresholds of everything in the advanced gateway settings.

Both solutions worked fine. The second one allowed me to keep seeing quality graphs in RRD data.

stephenw10 · Dec 5, 2014, 4:44 PM

Some cable modems will hand out a local private IP if they loose their upstream connection in order to access them and fix whatever problem exists. That's fine but pfSense receives that IP and never rejects it favour of a public IP when the connection comes back. What does Status: Interfaces: show when the WAN connection is lost?

Steve

jerryaycock · Dec 5, 2014, 7:04 PM

Dec 5 18:00:27 apinger: ALARM: WAN_DHCP(68.0.32.1) *** down ***

Steve I will check status next time.

Thanks Jerry

stephenw10 · Dec 5, 2014, 7:37 PM

Ok, so apinger is marking the connection as down. Check the system log for any interface UP of DOWN events like:

kernel: fxp0: link state changed to DOWN

That would indicate a local physical connection failure like if your modem rebooted.

If you don't see that, and it's not the cable modem local address problem, then it's probably apinger causing the problem. It may be a symptom rather than a cause but it's easy to test by disabling it. Go to System: Routing: Gateways: and edit the existing gateway. Check 'disable gateway monitoring'.

Steve

jerryaycock · Dec 8, 2014, 1:36 PM

I turned off gateway monitor and has not failed .
Hope I can find time to learn more about PFSense.
Jerry

stephenw10 · Dec 8, 2014, 2:15 PM

Hmm, well there are some issues with apinger as you may have seen while searching for this error. Personally I've never had any problems with it. Hopefully this will be resolved when 2.2 is released.

Steve

kejianshi · Dec 8, 2014, 3:29 PM

Hmmmm…

Well - For me at least raising the thresholds very high worked just as well as turning it off.
That way when you really do have an outage it will do what its supposed to and you get to keep your pretty rrd quality graphs.

stephenw10 · Dec 8, 2014, 6:53 PM

It may work just as well normally but will it come back up after an outage? With the threasholds set to their standard values if you have a laggy or lossy connection, such as a wifi or 3g/4g link, then you would exepct to see lots of alarms and down times but you still expect it to come back up by itself which this isn't. :-
I'll not pretend to know too much about apinger though it seems to behave in it's own special way. ;) I once read through all the config scripts and still came away with conflicting ideas.

Steve