Garvin Pang

We would love this feature also. We have increased service quotas and currently rely on a fork of the VPC RC to get around this

The issue is because of the optimize list watcher in custom controller. First let's establish that the lowest level list call is going to use the [listFunc](https://github.com/aws/amazon-vpc-resource-controller-k8s/blob/master/controllers/custom/custom_controller.go#L184) Then I am...

The issue is actually https://github.com/aws/amazon-vpc-resource-controller-k8s/blob/f61320b5401f0fe0c02c25e5bd3ad0066ef8394f/controllers/custom/custom_controller.go#L191 What happens during 410 is that client go has a fallback that performs a full list call against the API server but because of the...

Let me know if it would be better if we jump on a call to go over this

Ya we had it reoccur. One way to test is change the optimizedlistwatcher to sleep for 5+ minutes between pages (assuming your etcd compaction interval is default). This should trigger...

If this is a test cluster it might be hard to reproduce. One thing you might have to do is make writes happen on etcd. I would say you would...

For sure let me take a look

We can close this issue against https://github.com/aws/amazon-vpc-resource-controller-k8s/pull/452

I have added TTL cache. Sorry for the delay.

@sushrk