Dr. Stefan Schimanski
Dr. Stefan Schimanski
When the DCOS package is uninstalled, the DCOS cli calls the /teaddown endpoint (http://mesos.apache.org/documentation/latest/scheduler-http-api/) which kills all executors of a framework. I think this leave https://github.com/mesosphere/kubernetes-mesos/issues/573 open to delete pods...
Our minion code restarts the executor/kubelet when it dies. But there are situation (see description) when the minion/executor/kubelet are not restarted.
I thought we had the RC=0 special case. But as we don't support recovering pods after restart, it wouldn't help.
Good idea. Can you create a Jira for Mesos for this? I guess we won't come up with another solution within k8sm for this. So a mesos-slave extension is probably...
Sound interesting. Can you ping them again and invite me as well?
On the other hand, we will not see anything on the horizon early enough I fear.
@jdef please take a look at our "Upgrade Orchestration Idea"
About your questions: - etcd config is not really pre-requisite. As long as the ExecutorId changes on those changes, the controller is able to update a cluster. Of course, as...