Update resource-propagating.md #782
Conversation
Signed-off-by: LavredisG <[email protected]>
|
[APPROVALNOTIFIER] This PR is NOT APPROVED This pull-request has been approved by: The full list of commands accepted by this bot can be found here.
Needs approval from an approver in each of these files:
Approvers can indicate their approval by writing |
karmada-bot
added
the
size/XS
| @@ -409,7 +409,7 @@ For example, when using `SpreadByFieldRegion` to specify region groups, at the m | |||
|
|
|||
| It has two replicaSchedulingTypes which determine how the replicas are scheduled when Karmada propagates a resource: | |||
|
|
|||
| * `Duplicated`: duplicate the same replicas to each candidate member cluster from resources. | |||
| * `Duplicated`: duplicate the same replicas to each candidate member cluster from resources. This is the <b> default </b> `replicaSchedulingType`. | |||
There was a problem hiding this comment.
According to the comments of ReplicaSchedulingType, the default should be Divided:
https://github.com/karmada-io/karmada/blob/58c3765684d386473d77a904652bb5fab272e375/pkg/apis/policy/v1alpha1/propagation_types.go#L554
There was a problem hiding this comment.
Could it be that there is a mistake there? This is a local test I did:
Besides, if Divided is default, it would have to specify a default ReplicaDivisionPreference besides the Type, right?
There was a problem hiding this comment.
Oh, I'm surprised but thanks for letting me know. I will explore it further.
There was a problem hiding this comment.
I revisited the previous discussion and we made the consensus that if replicaScheduling == null, the default behavior is Duplicated, otherwise the default behavior is Divided.
Maybe we need to update the comments on API definition.
There was a problem hiding this comment.
Which Preference would be chosen in the case that default behavior is Divided?
There was a problem hiding this comment.
Yeah. Since your test above omit the replicaScheduling, it defaults to Duplicated, the behavior is expected. (We made a mistake on the API part, hopefully to fix it on next API version, like v1alpha2 or beta.)
There was a problem hiding this comment.
Understandable! However, in the case that we have for example replicaScheduling == test, then as you said since it is not null, it will default to Divided. But Divided requires a preference, either Aggregated or Weighted if I am not mistaken, so which one of these would be chosen in that case?
I came up with this while experimenting with custom schedulers. Correct me if this isn't the case, but if it is, I think it's helpful to note it on the docs.