Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[proposal] Get node bandwidth automatically for network qos #1883

Closed
zwzhang0107 opened this issue Jan 31, 2024 · 2 comments · Fixed by #1982
Closed

[proposal] Get node bandwidth automatically for network qos #1883

zwzhang0107 opened this issue Jan 31, 2024 · 2 comments · Fixed by #1982
Assignees
Labels
area/koord-manager area/koordlet good first issue Good for newcomers help wanted Extra attention is needed kind/proposal Create a report to help us improve

Comments

@zwzhang0107
Copy link
Contributor

What is your proposal:
In cloud environment, node bandwidth is limited according to the instance type by providers. Define in slo-controller-config is complex since they may various instance types in cluster.

We need to define a protocol on Node, such as annotation called node-bandwith. koordinator users can implements their own controller to patch the annotation on node according to the instance type.

koordinator also need to check whether the network qos config is valid for each node and alert if there is something unreasonable.

Why is this needed:

Is there a suggested solution, if so, please add it:

@zwzhang0107 zwzhang0107 added area/koordlet area/koord-manager kind/proposal Create a report to help us improve good first issue Good for newcomers help wanted Extra attention is needed labels Jan 31, 2024
@sjtufl
Copy link
Contributor

sjtufl commented Mar 28, 2024

/assign

@sjtufl
Copy link
Contributor

sjtufl commented Apr 2, 2024

Split into 2 parts:

  1. Introduce an extension API (through annotation) to specify the total bandwidth for a node. This parameter takes precedence over settings established in the corresponding NodeStrategy or ClusterStrategy for the given node. The annotation will influence the NodeSLO, which is subsequently consumed by the network QoS component.
  2. Implement validation of the total-node-bandwidth configuration within webhook plugins (both node & cm).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
area/koord-manager area/koordlet good first issue Good for newcomers help wanted Extra attention is needed kind/proposal Create a report to help us improve
Projects
None yet
Development

Successfully merging a pull request may close this issue.

2 participants