(I ended up hand-adding klog.ErrorS(err, ...) around all the error-sites to find out what happened)
Given I use this Helm-chart (with some modifications to fit our deployment-infrastructure), I'm guessing it's an upstream issue somewhere. But if it's in this code (I don't quite see how, tho), the Helm-chart or the upstream webhook code, I don't know.
After quite a bit of digging with kubectl log ..., it turns out the errors turn up in the cert-manager-webhook-container and not the one actually producing the errors. Feels extremely weird.
When I have an error in the way I configured the webhook (invalid credentials, wrong upstream URL, ...), no errors how up in my logs:
(I ended up hand-adding
klog.ErrorS(err, ...)
around all the error-sites to find out what happened)Given I use this Helm-chart (with some modifications to fit our deployment-infrastructure), I'm guessing it's an upstream issue somewhere. But if it's in this code (I don't quite see how, tho), the Helm-chart or the upstream webhook code, I don't know.