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
[COST-1913] Keda implementation #471
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Early look, looks good - just one small change at this point
apiVersion: v1 | ||
kind: Namespace | ||
metadata: | ||
name: test-basic-app |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The name here should match the name of the test, so in this case test-autoscaler
With #481, you should be able to refer to |
Ideally we'd like to have some tests around this, just ensuring that it puts out the right resource, and secondly we need to run |
2e30155
to
b2bf73d
Compare
Gah, looks like there's some conflicts with #486. Also, I'm not sure if you GPG signed your last commit, but our PR check requires that. |
Ok, I'll resolve the conflicts and sign the next commit. |
27ce32a
to
dd065dd
Compare
I looked into the failing test ( For a Deployment, the default value for For the failing test, we create the ClowdApp, set the ClowdApp disabled to true, set the Deployment replicas to This worked previously because we were always setting the deployment replica count to the specified minReplicas in the clowdapp spec. For this test though, that value is not set. So this test was previously unsetting the This behavior is implicit and I think it "just happened to work". I think logically it makes sense to specify a minReplica count for this test. However, if this test is demonstrating the expected behavior of clowder, then I'll need to update the deployment logic to reset a |
I think we should add |
This PR is an idea for a KEDA implementation. This takes a different approach compared to #105. The CRD here adds an optional
autoScaler
field to eachdeployment
. This field matches the KEDA ScaledObjectSpec, except for theMinReplicaCount
(this value is left out of the spec in favor of the deploymentminReplicas
).An example Clowdapp. This definition will create a standard HPA based on CPU utilization for the
hello-app
deployment:TODO: