Closed alpe12 closed 2 years ago
The route add default is in the down process of openvpn after having stopped all others services, to avoid leaks, in order to re establish a lost connection to vpn servers. I've not fully tested this feature as previously, every time the connection was lost, all the container was stopped.
As you said, the error is just a symptom of an uncompleted vpn connection. When the connection is established with vpn server, openvpn calls the up.sh script wich drop default route. So later, re enabling that route should be ok.
#when vpn is up, delete default route going through eth0
route del default eth0
The setup process check dns resolution and perform a ping. Here the dig +short downloads.nordcdn.com
command returned a result not null, but performing a ping on that resolved address failed.
I'm afraid you have a dns issue or network issue.
Should be this issue still active please le me know.
route add default dev eth0
is failing with this error.This is the output of the route command inside the container.
If this is the way it's supposed to look like at this point of the script execution, make it ignore "route: ioctl 0x890b failed: File exists" errors and continue.Replacing
route add default dev eth0
withip route replace default dev eth0
might be a solution.Edit: It failing is possibly a symptom, not the cause of the issue. So using 'replace' is probably only a partial fix.
Thanks again.
PS: after restarting the service many times it eventually work. very weird since I always use
docker-compose down
to start clean. PPS: I'm still debugging to find out why this happens.Sometimes I also get: