Rc.newwanip triggered every 30 minutes

jhuebner

Hey there,

the following problem is bugging me. Since today I'm using the new 2.2 release and the problem is still there. It was the same with 2.1.5 before. I'm running pfSense on a Soekris net6501-70 behind a Fritzbox 7390 which I cannot get easily rid of.

Every thirty minutes the log shows the following entry:

rc.newwanip: pfSense package system has detected an IP change or dynamic WAN reconnection

or

rc.newwanipv6: pfSense package system has detected an IP change or dynamic WAN reconnection

This also triggers all services like OpenVPN etc. to be restarted which is really annoying. I think it all started when I activated the delegation of IPv6 networks to my pfSense box. The WAN interface is configured to have a fixed IP in the Fritzbox network for IPv4 and DHCP6 for IPv6. My LAN and OPT1(WLAN actually) interface also have static IPv4 adresses (own subnets) for IPv4 connectivity and are configured to track the WAN interface for IPv6 - which is working fine.

Any ideas what might trigger the alleged change of/on my WAN interface and how I can prevent the constant restarting of services? I'm quite experienced with pfSense, but this one really got me stumped.

Thanks in advance,
Regards

cmb

Is 1 hour the lease length on your DHCP6? Should see logs in the system or DHCP log with info on lease length, etc.

jhuebner

It might very well be (thirty minutes is the interval of my problem btw). In this case the pfSense is the client for IPv6 addresses on the WAN interface. I cannot see anything relevant in the DHCP tab of status/system logs/dhcp - it only shows entries of the dhcpd which is serving out addresses to the LAN/WLAN interfaces.

And even if it is the lease length, in 47 cases over the day all addresses stay the same - no need to trigger a services restart just because a lease has been acknowledged..

cmb

You should see some logs from dhcp6c somewhere (I just don't recall offhand whether those end up in the system log or the DHCP log). Need to know the specific cause of the issue to do anything about it.

jhuebner

I grep'ed for it in /var/log, it gives

/var/log/dhcpd.log:Jan 27 00:16:28 soekris dhcp6c[26217]: failed to add an address on em0: Invalid argument
/var/log/dhcpd.log:Jan 27 00:16:28 soekris dhcp6c[26217]: failed to update an address [IPv6-address-of-EM0/WAN]

every thirty minutes. Whatever that invalid argument is..?!

[IPv6-address-of-EM0/WAN] is of course a real address in my logs.

dugeem

@jhuebner:

This also triggers all services like OpenVPN etc. to be restarted which is really annoying. I think it all started when I activated the delegation of IPv6 networks to my pfSense box.

A fix was previously documented in https://redmine.pfsense.org/issues/3669 for restarts. This impacts OpenVPN (even if OpenVPN server is only configured for IPv4).

Note that there are 2 parts to the fix:

First part relates to IPv4 script /etc/rc.newwanip - this fix is rolled into 2.1.4 and later.
Second part relates to IPv6 script /etc/rc.newwanipv6 - appears to be awaiting review (despite ticket being marked Resolved). Also impacts 2.2 release. In 2.2 the fix needs to be applied to rc.newwanipv6 script at line 125.

In the meantime it is easy to apply the fix and stop the restarts.

jhuebner

Wow, just adding the "::" to /etc/rc.newwanipv6 in line 125 resolved all my problems.

Thanks for your time, this is a great community!

Now this trivial change just has to find its way into the releases.

Iahmad

dear,

please can you more specific where to add"::" at line 125