Restart behaviour of the VM daemon

containers / gvisor-tap-vsock

A new network stack based on gVisor

Apache License 2.0

250 stars 47 forks source link

Restart behaviour of the VM daemon #9

Open gbraad opened 3 years ago

gbraad commented 3 years ago

Both in the Docker instructions

docker run -d --name=gvisor-tap-vsock --privileged --net=host -it quay.io/crcont/gvisor-tap-vsock:latest

and in SNC

podman generate systemd --restart-policy=no gvisor-tap-vsock > /etc/systemd/system/gvisor-tap-vsock.service

there is no restart policy. What if the daemon crashed? On several occassions I have been unable to connect to the VM when a restart was performed with a running host daemon: crc daemon&; crc start; crc delete -f; crc start

ghost commented 2 years ago

I don't really understand how this issue is applicable to the gvisor-tap-vsock project persay.

Wouldn't the restart policy for the VM daemon fall onto the project it is being implemented in?