chore: Get resource from cache instead of APIServer #193
Conversation
Currently in ACK, at the beginning of the Reconciler for each event, we get the resource from the k8s APIServer. This is an unecessary call, as the latest object that triggered the reconciler is already stored in the informers cache. With this change, we will be retrieving the resource from the cache instead of the APIServer.
|
[APPROVALNOTIFIER] This PR is APPROVED This pull-request has been approved by: michaelhtm 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 |
|
/test iam-controller-test |
|
/test s3-controller-test |
|
/test sagemaker-controller-test |
|
@michaelhtm: 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. |
| @@ -197,15 +198,15 @@ func (r *fieldExportReconciler) getFieldExport( | |||
| req ctrlrt.Request, | |||
| ) (*ackv1alpha1.FieldExport, error) { | |||
| ro := &ackv1alpha1.FieldExport{} | |||
| // Here we use k8s APIReader to read the k8s object by making the | |||
| // Here we use k8s ackResourceCache to read the k8s object by making the | |||
| // direct call to k8s apiserver instead of using k8sClient. | |||
| // The reason is that k8sClient uses a cache and sometimes k8sClient can | |||
| // return stale copy of object. | |||
| // It is okay to make direct call to k8s apiserver because we are only | |||
| // making single read call for complete reconciler loop. | |||
| // See following issue for more details: | |||
| // https://github.com/aws-controllers-k8s/community/issues/894 | |||
There was a problem hiding this comment.
Is the problem described in this issue no longer relevant?
There was a problem hiding this comment.
It's not related to a problem..it's mainly just an optimization.
I'll expand more on it in a community PR
There was a problem hiding this comment.
From this comment it sounds like the use of direct server calls was to avoid duplicate create calls. Or are you referring to this change when you say optimization?
|
/hold |
ack-prow
bot
added
the
do-not-merge/hold
Description of changes:
Currently in ACK, at the beginning of the Reconciler for each event, we
get the resource from the k8s APIServer. This is an unecessary call, as
the latest object that triggered the reconciler is already stored in the
informers cache.
With this change, we will be retrieving the resource from the cache
instead of the APIServer.
By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.