kubernetes 1.3.5
when you set terminationGracePeriodSeconds: 999 for the pod and this pod terminates tool long (it has to reschedule the data), kubelet terminates pause container (after 2 minutes when --sync-frequency=1m0s) and pod became inaccessible. and as a result data rescheduling fails too.
pod.yaml:
apiVersion: v1
kind: Pod
metadata:
name: long-waiting-pod
spec:
containers:
- name: base
image: alpine:3.4
command:
- sleep
- '999999'
terminationGracePeriodSeconds: 999999kubectl create -f pod.yaml
run in another terminal and watch for eth0 network interface.
kubectl exec -ti long-waiting-pod -- watch -n1 ifconfig
delete the pod:
kubectl delete pod long-waiting-pod
and after 2 minutes you'll see:
Every 1s: ifconfig 2016-10-21 21:53:48
lo Link encap:Local Loopback
inet addr:127.0.0.1 Mask:255.0.0.0
inet6 addr: ::1%32587/128 Scope:Host
UP LOOPBACK RUNNING MTU:65536 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1
RX bytes:0 (0.0 B) TX bytes:0 (0.0 B)
kubernetes 1.3.5
when you set
terminationGracePeriodSeconds: 999for the pod and this pod terminates tool long (it has to reschedule the data), kubelet terminates pause container (after 2 minutes when--sync-frequency=1m0s) and pod became inaccessible. and as a result data rescheduling fails too.pod.yaml:
run in another terminal and watch for
eth0network interface.kubectl exec -ti long-waiting-pod -- watch -n1 ifconfigdelete the pod:
and after 2 minutes you'll see: