Hi Rodrigo,
Thank you for your answers.
>At the minions? Why do you want to do that? Sorry, don't follow
I followed the coreos kuberentes installation. The minions are connected to
etcd as well as the master node,
here the flannel config for the nodes:
*/etc/flannel/options.env*
FLANNELD_IFACE=${ADVERTISE_IP}FLANNELD_ETCD_ENDPOINTS=*${ETCD_ENDPOINTS}*
Thats why I asked how the nodes should communicate with a three node etcd
cluster.
Cheers, Thomas
Am Dienstag, 5. Juli 2016 22:31:20 UTC+2 schrieb Rodrigo Campos:
>
> On Tue, Jul 05, 2016 at 01:07:54PM -0700, Thomas Privat wrote:
> > Hi All,
> >
> > Im running a Kubernetes Cluster with one master node and three minions
> and
> > Im wondering what will happen to the services and pods on the minions if
> > the master node dies.
> > Are the services and pods continue running in current state ?
>
> Yes
>
> > Is it a typical setup that the master node doesnt run services and pods
> Im
> > creating in the Kubernetes cluster (except of the kubernetes management
> > pods) ?
>
> I think so, kube-up does that on AWS and I think it generally does that on
> other
> providers. But not sure
>
> > If yes, is this the reason why "kubectl get nodes" shows the status
> > "SchedulingDisabled" for the master node ?
> >
> > What is the usual way to run a three node etcd cluster in kubernetes,
> > should we add a http load balancer in front of the etcd nodes to have
> one
> > ip-address for etcd ?
> > Or is it possible to configure the three ip addresse from the etcd nodes
> at
> > the kubernetes minions ?
>
> At the minions? Why do you want to do that? Sorry, don't follow
>
>
>
> Thanks a lot,
> Rodrigo
>
--
You received this message because you are subscribed to the Google Groups
"Containers at Google" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/google-containers.
For more options, visit https://groups.google.com/d/optout.
