Extremely slow response times when in GKE

System Information
  • v4.8.2:
  • Operating System:
  • postgres:
  • v16.17.0:
  • 8.15.0:
  • 1.22.19:

I have a combination of nextjs+expressJS+Strapi+Postgres. Until now it was working on a basic self managed k8s cluster on a openstack infrastructure. All fine but we decided to move to Google Kubernetes Environment.
When i deploy there the response times to strapi are extremely slow. I.e. from 50-200ms to up to 12 000!
I tried to connect my nextjs to the strapi hosted on old k8s, and it works fine. Tried all possible combinations and it turns out that only Strapi on GKE is having this issue.
It is not Nexjs or express or postgress as all of them work fine with Strapi on the old kubernetes. So the only failing element is Strapi on GKE. Both Strapis are exactly the same (same REPO).
It is not lack of resources on GKE. I scaled enormously for the test.
There are no failing logs from strapi, just extremly slow POST:graphql requests, which turn to timeouts.

Any ideas? Suggestions in what else direction to go. I can no believe that a payed k8s service in google makes life so harder.

Could it be that strapi and the frontend are in 2 diffrent datacenters or countrys?