Fix master scaleup #6793
Fix master scaleup #6793
Conversation
michaelgugino
requested review from
abutcher,
kwoodson,
mtnbikenc,
sdodson and
vrutkovs
openshift-ci-robot
added
the
size/M
|
This seems to work pretty well. Deployed 1 master + 1 node, then added 1 new master and 1 node (simultaneously) and ran openshift-master/scaleup.yml Output: PLAY RECAP ***********************************
master1 : ok=79 changed=4 unreachable=0 failed=0
node2 : ok=167 changed=57 unreachable=0 failed=0
master2 : ok=343 changed=132 unreachable=0 failed=0
node1 : ok=3 changed=0 unreachable=0 failed=0
localhost : ok=23 changed=0 unreachable=0 failed=0
INSTALLER STATUS ***********************
Initialization : Complete (0:00:16)
Load balancer Install : Complete (0:00:00)
Master Install : Complete (0:03:19)
Node Install : Complete (0:02:08) |
|
I did confirm with oc get pods and oc get nodes everything looks to be in order. |
michaelgugino
force-pushed
the
fix-master-scaleup
branch
from
23296db to
77aa3f4
Compare
Allow playbooks/openshift-master/scaleup.yml to call prerequisites.yml at the proper time. Related-to: openshift#6784
michaelgugino
force-pushed
the
fix-master-scaleup
branch
from
77aa3f4 to
682ddb8
Compare
| - fail: | ||
| # new_masters must be part of new_nodes as well; | ||
| msg: > | ||
| Each host in new_masters must also appear in new_nodes |
There was a problem hiding this comment.
Not sure why wouldn't scaleup disallow scaling up dedicated masters
There was a problem hiding this comment.
Our masters must be nodes for various reasons, mainly they need to be on the SDN but in 3.9 masters will actually be expected to run the console pods.
There was a problem hiding this comment.
Do we actually run the node scaleup playbook to complete the node specific tasks or do we need to document the need to run node scaleup too?
There was a problem hiding this comment.
@sdodson this openshift-master/private/scaleup.yml and openshift-node/private/scaleup.yml both call openshift-node/private/config.yml. There is no need to run openshift-node/scaleup in addition to this play if you're adding both masters and nodes. The nodes were already being configured at the same time as masters, according to our plays.
If one is adding nodes and 0 masters, then they can run openshift-node/scaleup.yml.
There was a problem hiding this comment.
in 3.9 masters will actually be expected to run the console pods.
Oh, I see, makes sense.
Do we actually run the node scaleup playbook to complete the node specific tasks
Yes, the node is being setup there
|
/lgtm |
|
/test all [submit-queue is verifying that this PR is safe to merge] |
|
Automatic merge from submit-queue. |
|
@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. DetailsInstructions 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. |
5 participants
This commit builds off #6784
Enables openshift-master/scaleup.yml to call prerequisites.yml as required. Also modifies some failure conditions around how scaleup plays are called to ensure inventories are better aligned with playbook behavior.