Charles-Edouard Brétéché

Closing, fixed in 1.8.0. Please reopen if not fixed.

I’m not sure an HPA will help, having more replicas will not distribute the load unfortunately.

Scaling will definitely help serving admission requests (the requests will be spread across all replicas). Memory is not necessarily related to the number of requests though and I’m afraid scaling...

Timeout only applies to resource webhooks.

@sandeshlmore for me it works as expected.

Closing, things seem to be working as expected. Please reopen if there's anything left to discuss.

How about putting this behind a feature flag so that we can test it ?

Do we want to keep this ? I thought it was not well suited for our use case. cc @JimBugwadia @chipzoller

@JimBugwadia if it's not going to be merged can you close it ?

> today implementation has plenty of flaws when running on heavy clusters. Ok so if I understand correctly, disk cache is not going to work in our case, but today...