It seems that the TaskManager pod could not resolve the JobManager address
"franz-01.default", which is constructed in "k8s-service-name.namespace".
I think you need to check whether the coreDNS is running normally in your
K8s cluster. You could start a busybox pod on the same node with
TaskManager
Pulling in Yang Wang who may shed some light on the matter.
You could also have a look at
http://apache-flink-user-mailing-list-archive.2336050.n4.nabble.com/Native-kubernetes-setup-failed-to-start-job-td39066.html;
while the issue was not actually resolved it may give some hints.
On 5/10/202
