This repository has been archived by the owner on Feb 9, 2022. It is now read-only.
WIP: providerid controller #17
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Contributor
rgolangh
commented
Jul 27, 2020
rgolangh
commented
- Increase leader election lease time
- Add controller to reconcile node's providerID
Add readiness/liveness probes
MOTIVATION This fix handles the network addresses reconciliation problems which leads to mishandling of status and phase reporting and lack of linkage, which causes the machine to apear in the wrong state. The main issue is around network reconciliation, and for that a major refactoring been made to better address this. See [1] below. RESULT While the VM is not up it still doesn't have any IP addresses and this should be reported as nil addresses. If the VM is UP it means we have to report error, and that will trigger a future reconcilliation. A change in the node object that, that happen when a VM goes down for example, triggers reconciliation and that should change the list of addresses and state accordingly. When the VM boots a reverse thing happens - the machine boots with no IP addresses, till the node object is detecting a change because kubelet starts responding. That triggers a reconcillation that will detect the addresses correctly. If not that should keep firing reconcillation till it does. MODIFICATION [1] structural changes: Instead of having pile of changes in one or two methods, every change to the status or spec has its own function. The high level break down of the any handling now is: - create or delete the vm. no-op on the vm for update - reconcile provider id - reconcile network addresses - reconcile annotations - reconcile providerStatus - update machine - update machine/status Signed-off-by: Roy Golan <[email protected]>
Bug 1817853: Reconcile network addresses according to VM status
Implement leader election for manager
The machine-api-controller components are refreshing their lease more than all other components combined. Bringing this to 90s each, will decrease etcd writes at idle.
BUG 1858400: [Performance] Lease refresh period for machine-api-controllers is too high, causes heavy writes to etcd at idle
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: rgolangh The full list of commands accepted by this bot can be found here. The pull request process is described here
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
|
@rgolangh: PR needs rebase. 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. |
rgolangh
force-pushed
the
providerid-controller
branch
3 times, most recently
from
079c7b6 to
5321dbf
Compare
This controller is a replacement for a cloud provider. Its purpose is to to set the node spec.providerID by querying ovirt/RHV api, where the node name is the VM name. This will make autoscalling work. Signed-off-by: Roy Golan <[email protected]>
rgolangh
force-pushed
the
providerid-controller
branch
from
5321dbf to
a865d06
Compare
|
@rgolangh: The following tests failed, say
Full PR test history. Your PR dashboard. 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. |
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.