etcd migrate: instead of scaleup playbook etcd server should be start… #7556
Conversation
…ed back master doesn't need to be restarted and etcd URLs updated as new etcd nodes are being added. There is no need to run scaleup playbook, as etcd nodes are already added to the cluster. The migrate procedure now does the following: * checks if the etcd data needs to be migrated * makes etcd backup * stops etcd services on all nodes except the first one * migrates etcd data on first etcd node * clears data on other etcd nodes * updates etcd cluster configuration, updating ETCD_INITIAL_CLUSTER and ETCD_INITIAL_CLUSTER_STATE * starts the etcd service again one by one on etcd nodes Now the only copy of migrated data is on the first etcd cluster, which would replicate it to other nodes. After migration is done master configs are updated and services restarted if needed (cherry picked from commit 7e30cda)
openshift-ci-robot
added
the
size/M
|
We should remove the $ grep -rn "scaleup.yml"
openshift-etcd/private/migrate.yml:113:- import_playbook: scaleup.yml
openshift-etcd/scaleup.yml:4:- import_playbook: private/scaleup.yml
openshift-master/scaleup.yml:23:- import_playbook: private/scaleup.yml |
@ingvagabund That's because it's an 'entry point.' There should be no other playbooks importing that playbook, it's a playbook that users would run when they wish to add etcd hosts to the cluster. |
|
/retest |
|
@michaelgugino: The following tests failed, say
Full PR test history. Your PR dashboard. Please help us cut down on flakes by linking to an open issue when you hit one in your PR. Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository. I understand the commands that are listed here. |
|
Replaced by #7662? |
…ed back
master doesn't need to be restarted and etcd URLs updated as new etcd
nodes are being added. There is no need to run scaleup playbook, as etcd
nodes are already added to the cluster.
The migrate procedure now does the following:
ETCD_INITIAL_CLUSTER_STATE
Now the only copy of migrated data is on the first etcd cluster, which
would replicate it to other nodes.
After migration is done master configs are updated and services
restarted if needed
(cherry picked from commit 7e30cda)