maybe-sybr

> > Workaround based on @zzzeek's answer from [sqlalchemy/sqlalchemy#5593](https://github.com/sqlalchemy/sqlalchemy/issues/5593): > > ```python > > def upgrade(): > > status_enum_postgres = postgresql.ENUM('pending', 'accepted', 'canceled', name='mytable_status_enum', create_type=False) > > status_enum = sa.Enum('pending',...

Actually, I just tried the `iptables` driver for kube-proxy (using the same simple change to `boot/kube-proxy.sh` you made in #194) and it's a bit more clear what is happening here....

Oh, and I probably should have checked the logs for the kube-proxy... ``` Mar 10 01:53:14 virt-test.lan kube-proxy.sh[58910]: E0310 01:53:14.321564 155 proxysocket.go:109] Dial failed: dial tcp 10.0.2.15:6443: connect: connection refused...

It's because we set the endpoint for `kubernetes` to `U7S_PARENT_IP` somewhere. From kube-proxy logs in userspace mode: ``` kube-proxy.sh[70336]: I0310 02:39:08.332225 152 roundrobin.go:235] LoadBalancerRR: Setting endpoints for default/kubernetes:https to [10.0.2.15:6443...

So I think this comes down to the same issue I raised back in October of last year (#193) where a machine with no configured host IP couldn't run usernetes...

If that's the case, I can make a PR and mark it as closing both this issue and #193 since there'll no longer be a problem with the apiserver not...

I've not been able to try with containerd yet. I'll try it when I have a bit of free time and get back to you.

@AkihiroSuda - appears to break under containerd as well. ``` $ kubectl describe pod Name: badness Namespace: default Priority: 0 Node: host/10.0.42.100 Start Time: Wed, 16 Sep 2020 10:06:53 +1000...

I'm using a hand modified `kube-apiserver.sh` to advertise 10.0.42.100 and bind 0.0.0.0 and things seem to be working fine for me so far. I'll report back if I notice any...

Would you be able to elaborate on how multi-node works with u7s at the moment, @AkihiroSuda ? I'm using it standalone and probably don't understand enough about the use cases...