You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Is your feature request related to a problem? Please describe.
Almost every time a value is modified on the CPA custom resource, the CPA Pod is recreated. Although in most cases this is not a problem, sometimes this affects scaling because the downscale stabilisation period is reset. For example, if a user modifies the CPA's min/max replicas, the CPA Pod will be recreated and most likely the replica count will be updated to the one computed on startup, rather than considering previous values (within the stabilisation window).
Moreover, I think that sending events or updating status and/or annotations of the CPA CR (as HPA does on its resources) would be nice for debugging/inspecting scaling using CPA.
Describe the solution you'd like
Ideally, CPA should do this (updating CPA CR status / annotations / sending events) in the background after the evaluation stage was ran. There is also the solution to manually call the kubernetes api from within the CPA Pod at the end of the evaluation. This gives the user greater flexibility, but it take some more work / is more error prone. In order to have at least this last solution, the CPA Operator should add some more permissions to the Role that it creates for the CPA Pod
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
Almost every time a value is modified on the CPA custom resource, the CPA Pod is recreated. Although in most cases this is not a problem, sometimes this affects scaling because the downscale stabilisation period is reset. For example, if a user modifies the CPA's min/max replicas, the CPA Pod will be recreated and most likely the replica count will be updated to the one computed on startup, rather than considering previous values (within the stabilisation window).
Moreover, I think that sending events or updating status and/or annotations of the CPA CR (as HPA does on its resources) would be nice for debugging/inspecting scaling using CPA.
Describe the solution you'd like
Ideally, CPA should do this (updating CPA CR status / annotations / sending events) in the background after the evaluation stage was ran. There is also the solution to manually call the kubernetes api from within the CPA Pod at the end of the evaluation. This gives the user greater flexibility, but it take some more work / is more error prone. In order to have at least this last solution, the CPA Operator should add some more permissions to the Role that it creates for the CPA Pod
The text was updated successfully, but these errors were encountered: