OpenVPN (beta & prod) frequent disconnects?

[jkuner]

Hi - I've been having frequent disconnects (about every 5 minutes or so), which started a week or two ago (not sure of exact start time).

I'm running FTTN to the gateway box in passthrough mode (DMZ+ I think?) to my RT-N66U router, which is running shibby tomato 1.28. The router is running a persistent connection to ovpn.sonic.net, configured from one of the earlier posts.

The current symptom is short disconnects (DNS & connectivity) about every 5 minutes.

I have a few possible culprits, but I'm not sure which one is the root cause and which are downstream.

One is that my ATT DHCP lease is only 10 minutes, though I seem to get the same IP address every time.

In addition, I have tried turning off the VPN at the router and using the Mac client (Viscosity), the connection timed out in about 5 minutes to the beta VPN. Connecting via Viscosity to the prod VPN disconnected after 15 minutes.

Any advice welcome for how to configure my router so I can prevent these timeouts.

Thanks.

[vbrobert]

Hey,

I don't think that the DHCP lease would be a major factor as the client just checks with the server every 5 minutes (half the time of the lease because science) to see if the IP address is still valid. If it doesn't change or expire, then you are good as far as DHCP goes.

Do you have trouble surfing without any VPN involved what so ever? If your internet connection is unstable in general, the VPN will behave worse. I assume that everything has been rebooted? Is it happening on multiple devices?

[rus]

I'm also getting frequent disconnects using the Mac open vpn client to connect to ovpn.sonic.net over an FTTN service from my laptop. No session lasts longer than 25 minutes, and most last only 5-10 minutes. It does not appear to correlate with anything I am doing. Its not really usable. Any thoughts?

[jkuner]

One other update, I went through and rebooted everything (router, ATT endpoint, etc) to troubleshoot as well as make sure the problem was on the VPN.

I'm more convinced the problem is with the VPN or config side.

Here are my logs from tomato; I saw some answers when I googled the error (specifically the route delete failure) but not sure how to reconfigure tomato to fix it.

Code: Select all

Sep 13 07:14:38 bidibidi daemon.info dnsmasq[19870]: reading /etc/resolv.dnsmasq
Sep 13 07:14:38 bidibidi daemon.info dnsmasq[19870]: using nameserver 208.201.224.33#53
Sep 13 07:14:38 bidibidi daemon.info dnsmasq[19870]: using nameserver 208.201.224.11#53
Sep 13 07:14:38 bidibidi daemon.info dnsmasq[19870]: read /etc/hosts - 2 addresses
Sep 13 07:14:38 bidibidi daemon.info dnsmasq[19870]: read /etc/dnsmasq/hosts/hosts - 7 addresses
Sep 13 07:14:38 bidibidi daemon.info dnsmasq-dhcp[19870]: read /etc/dnsmasq/dhcp/dhcp-hosts
Sep 13 07:14:41 bidibidi daemon.err openvpn[19719]: event_wait : Interrupted system call (code=4)
Sep 13 07:14:41 bidibidi daemon.notice openvpn[19719]: SIGTERM received, sending exit notification to peer
Sep 13 07:14:42 bidibidi daemon.notice openvpn[19719]: /sbin/route del -net x.x.x.x netmask 255.255.255.255
Sep 13 07:14:42 bidibidi daemon.warn openvpn[19719]: ERROR: Linux route delete command failed: external program exited with error status: 1
Sep 13 07:14:42 bidibidi daemon.notice openvpn[19719]: /sbin/route del -net 0.0.0.0 netmask 128.0.0.0
Sep 13 07:14:42 bidibidi daemon.notice openvpn[19719]: /sbin/route del -net 128.0.0.0 netmask 128.0.0.0
Sep 13 07:14:42 bidibidi daemon.notice openvpn[19719]: Closing TUN/TAP interface
Sep 13 07:14:42 bidibidi daemon.notice openvpn[19719]: /sbin/ifconfig tun11 0.0.0.0
Sep 13 07:14:42 bidibidi daemon.notice openvpn[19719]: updown.sh tun11 1500 1558 x.x.x.x 255.255.255.0 init
Sep 13 07:14:48 bidibidi user.notice root: vpnrouting: clean-up

[forest]

Possible cause:

Pace 5268AC, DMZPlus Issue, drops lan every 10 minutes

I had this problem (or at least the same behavior) a couple weeks ago. Rebooting the Pace gateway and my router/vpn box made it go away (for now).

[jkuner]

Just opened a ticket with sonic support. I have FTTN and it looks like I was getting a lot of connect / disconnect messages even from the modem's perspective.

Code: Select all

INF	2016-09-15T21:47:40-07:00	sys	Node x.x.x.x is down
INF	2016-09-15T21:42:23-07:00	sys	Node x.x.x.x  is up
INF	2016-09-15T21:37:40-07:00	sys	Node x.x.x.x  is down
INF	2016-09-15T21:32:27-07:00	sys	Node x.x.x.x  is up
INF	2016-09-15T21:27:39-07:00	sys	Node x.x.x.x  is down

[nstjohn]

I'm debugging OpenVPN disconnects and stumbled across this thread. I too noticed the ATT router node up/down messages. What was the resolution with Sonic support?

[nstjohn]

I just read regarding the up/down messages. The node (local IP address) I'm debugging doesn't have down messages so I don't think the log shows any problems but I still have OpenVPN (on my Mac) dropping apparently after being idle for a while. This is most apparent first this in the morning when I can browse the internet until the OpenVPN connection is (automatically) reestablished.