c12s-kubespray/roles/kubernetes/master/handlers/main.yml
Bogdan Dobrelya 13f5217d88 Controll delayed startup/restart for kubelet/etcd
In order to mitigate sporadic data races in etcd
(publish error: etcdserver: request timed out"):
- Add etcd_start_delay and kubelet_start_delay (defaults to a 5 sec.)
- Increase default start sleep times to foo_start_delay from a 1 sec.
- Add restart sleeping as well.
- Add missing start sleep commands as appropriate.

Closes: https://github.com/kubespray/kargo/issues/342

Signed-off-by: Bogdan Dobrelya <bdobrelia@mirantis.com>
2016-10-20 15:25:32 +02:00

45 lines
1.1 KiB
YAML

---
- name: Master | restart kubelet
command: /bin/true
notify:
- Master | reload systemd
- Master | reload kubelet
- Master | wait for master static pods
- name: Master | wait for master static pods
command: /bin/true
notify:
- Master | wait for the apiserver to be running
- Master | wait for kube-scheduler
- Master | wait for kube-controller-manager
- name: Master | reload systemd
command: systemctl daemon-reload
when: ansible_service_mgr == "systemd"
- name: Master | reload kubelet
service:
name: kubelet
state: restarted
sleep: "{{ kubelet_start_delay|int }}"
- name: Master | wait for kube-scheduler
uri: url=http://localhost:10251/healthz
register: scheduler_result
until: scheduler_result.status == 200
retries: 15
delay: 5
- name: Master | wait for kube-controller-manager
uri: url=http://localhost:10252/healthz
register: controller_manager_result
until: controller_manager_result.status == 200
retries: 15
delay: 5
- name: Master | wait for the apiserver to be running
uri: url=http://localhost:8080/healthz
register: result
until: result.status == 200
retries: 10
delay: 6