Write per service health checks

Registered by Ryan Hallisey on 2016-05-14

Kubernetes gives the operator the ability to keep their stack healthy with built in health checks. The community needs to add health checks per service that will make sure services come to life properly.

http://kubernetes.io/docs/user-guide/liveness/
https://github.com/kubernetes/kubernetes/issues/21315

Blueprint information

Status:
Complete
Approver:
Steven Dake
Priority:
High
Drafter:
Ryan Hallisey
Direction:
Needs approval
Assignee:
None
Definition:
Superseded
Series goal:
None
Implementation:
Not started
Milestone target:
None
Completed by
Steven Dake on 2017-02-12

Related branches

Sprints

Whiteboard

We could have a readiness check poll etcd to determine if a bootstrapping has been completed and it's safe to proceed. When it comes time to upgrade, we could let a service know it needs to bootstrap and have the services wait until the bootstrap completes. --rhallisey

Some type of readiness probe is needed in the containers I suspect for entrypoint to work correctly. -sdake

(?)

Work Items

Work items:
unassigned)ceph: TODO
(unassigned)mariadb: TODO
(unassigned)rabbitmq: TODO
(unassigned)memcached: TODO
(unassigned)keystone: TODO
(unassigned)glance: DONE
(unassigned)neutron: TODO
(unassigned)nova: TODO
(unassigned)cinder: TODO
(unassigned)openvswitch db/vswitchd: TODO
(unassigned)horizon: TODO
(unassigned)heat: TODO

This blueprint contains Public information 
Everyone can see this information.