Simple way to issue a command to restart the VPN.

[sboger]

I am looking for a simple way to trigger a vpn disconnect and reconnect. Not restart the container, just a vpn restart like when the healthcheck fails. I want to issue this a few times a week to randomly change the VPN endpoint.

For instance, I use this snippet to check the current location of the VPN:

docker exec -ti [container id] 'wget' '-qO-' 'https://ipinfo.io' | jq -r '.city + ", " + .country'

Is there a similar command to issue inside of the container to perform a restart? Or maybe a curl request to the HTTP control server?

[sboger]

Answering my own question. One can utilize a "bug" (per qdm12) that triggers auto-healing to rotate to a new vpn endpoint by issuing this to the control server:

docker exec -ti [gluetun container id] 'wget' '-qO-' '--method=PUT' '--body-data={"status":"stopped"}' 'http://127.0.0.1:8000/v1/openvpn/status'

[ezekieldas]

Hey @sboger --A few questions, if you a have a moment I'd really appreciate your input.

• Are you using a wireguard or openvpn config?
• What gluetun version are you using? gluetun:v3 or earlier?
• You mentioned a 'bug (per qdm12)' --would it be either of these:
  • Bug: OpenVPN no client-side authentication method is specified. #2217
  • Bug: Control Server endpoint /v1/openvpn/status on non-OpenVPN causes {"outcome":"crashed"} #2277

I'm wanting to collect some more detail on this and other updater issues.

Also, as noted in the issues I've mentioned, there is a deadly combination of using this to reissue an ipaddr / gateway if your cocktail involves: > gluetun:v3 + wireguard + /v1/openvpn/status

I just summarized my previous use, etc. in this recent comment:

#2277 (comment)

[Yuoter]

@sboger I also found such option, it works a bit faster:
GET request to URI /openvpn/actions/restart