What's new

Solved WAN DHCP renewal + VPN drops connection

  • Thread starter Deleted member 83610
  • Start date
  • SNBForums Code of Conduct

    SNBForums is a community for everyone, no matter what their level of experience.

    Please be tolerant and patient of others, especially newcomers. We are all here to share and learn!

    The rules are simple: Be patient, be nice, be helpful or be gone!

That was right after I ran your commando.
You're in a different time zone to me and saying "10 minutes ago" or "right after" doesn't help me understand the log entries. You need to tell me at what time something happened with reference to the times in the log file. So did you run the command at 22:32:56 or was it some other time?
 
Okay. I just cleared the log this time around. Then I ran your killall command again right before 23:15 my time. This is everything that happened in the system log right after:

syslog.txt

Also, I could try and disable the VPN client. Not sure if it will help though since it seems the lease just expires and the connection drops beforehand anyways.
 
Last edited by a moderator:
That was right after I ran your commando.

Also, I've stopped using the guest networks a while back for what you're mentioning.
I always have a blast watching @ColinTaylor go commando as well... ;) LOL
 
:D

This is my favourite commando:

1668464992284.png


Yes, I happen to love movies.
 
I'll try and stop the VPN connection beforehand this time around:

1668466827620.png
 
Does Predator count?
Perhaps not. I just got carried away. :)

Without the VPN connection running I saw as before that the lease expired and I lost one packet while pinging Google:

1668468083698.png


This is the system log from the time I stopped the VPN connection and then the lease expired approximately five minutes after:

syslog.txt

It automatically started up the VPN connection again (probably because of the killswitch).
 
This is all very strange. I'm beginning to think this is something specific to the RT-AC86U firmware. With my RT-AX86U and RT-AC68U when you do a killall -s USR1 udhcpc it's meant to do a renew, as seen in the logs. But in your case it appears to make no attempt to renew and instead deconfigures the interface. It's only meant to do that after a leasefail condition.

I'm wondering whether this happens on all RT-AC86U's and people don't notice the brief outage because their lease is much longer.


Let off some steam Bennett.
 
Last edited:
I think you're right on.

The effect when the lease expires is pretty unfortunate if you're playing Diablo on hardcore or some other competitive game. :eek:

The only way I started noticing was on the traffic graphs on some software I'm running. I noticed these dips straight to the floor with nothing coming in or going out with hourly intervals.
 
See if you can get your ISP to increase the lease time to a day or more. Then you can probably ignore it and hope that it's fixed in a future firmware update.
 
I think you're right on.

The effect when the lease expires is pretty unfortunate if you're playing Diablo on hardcore or some other competitive game. :eek:

The only way I started noticing was on the traffic graphs on some software I'm running. I noticed these dips straight to the floor with nothing coming in or going out with hourly intervals.

Have you tried playing with the DHCP Query Frequency setting - normal/aggressive/continuous?
 
See if you can get your ISP to increase the lease time to a day or more. Then you can probably ignore it and hope that it's fixed in a future firmware update.
Yes, I will give them a call tomorrow. It would be really nice if @RMerlin could look into it if he can find the time some day though.

Meanwhile... Thanks a lot for the help so far @ColinTaylor. :)

Have you tried playing with the DHCP Query Frequency setting - normal/aggressive/continuous?
Yes, I have tried all of them unfortunately without it making any difference. Thanks for the suggestion though.
 
If you can install Entware and then tcpdump, you can try to capture the traffic and see what is happening.
Code:
tcpdump -vnes0 -i eth0 port 67 or port 68
Leave it running for the full hour.
 
@raritywrt Here's a temporary change that might provide some more information.

Issue the following command and wait a few seconds for it to complete.
Code:
killall udhcpc; /sbin/udhcpc -i eth0 -S -p /var/run/udhcpc0.pid -s /tmp/udhcpc_wan -O33 -O249

Then either wait for the next renewal or force it with the following command:
Code:
killall -s USR1 udhcpc
See what information you get in the syslog. In my case I see this:
Code:
Nov 15 02:01:32 udhcpc[20797]: performing DHCP renew
Nov 15 02:01:32 udhcpc[20797]: sending renew
Nov 15 02:01:32 udhcpc[20797]: lease of 80.70.xxx.yyy obtained, lease time 326442
Nov 15 02:01:32 custom_script: Running /jffs/scripts/dhcpc-event (args: renew)

I'm wondering whether it's this problem noted in the source code. In which case this command might reveal something:
Code:
netstat -anp | egrep ':67|:68'
 
Last edited:

Latest threads

Sign Up For SNBForums Daily Digest

Get an update of what's new every day delivered to your mailbox. Sign up here!
Top