Installed via Helm Chart, Now What?

I’m sure that I could stand to read further into the documentation, but I’m now on Day 2 of trying to get a Kong proof-of-concept running. I am working the problem from a couple of different angles. I began by trying a very basic installation on my mac, using a local postgres instance, ran into problems with that pretty quick.

Then I switched to trying out the docker-compose version, had a lot more success with it. Some little weird bit still though. And we’re not going to run this in production using docker-compose, so I set that aside.

At the end of the day we are going to want to run Kong Enterprise in a Kubernetes cluster. Eventually in AWS but (luckily) we’re going to install it in Google GKE. And if I can have my choice, I will install it using Helm. Which I have done.

But now what? I cannot seem to get the Admin GUI working. The only thing I set is a big 'ol JSON document back from the admin service. Everything is coming up green in the GKE UI, looks like a successful installation. Still I’m at a loss as to what my next step is. I plan to install some demo API, say echo-server, and then rate-limit it with Kong.

If anyone has suggestions would be pleased to follow along. FYI I have disabled TLS trying to get a very very simple demo going and getting TLS working might slow that down. I’ve also enabled Ingress for both the admin service and the proxy.

From the 5 Minute Quickstart, this would seem to be my next step:

curl -i -X POST \
  --url \
  --data 'name=example-service' \
  --data 'url='

Which is giving me a 404 error. I see in my Google load-balancer that there are 3 host/path rules, one for /, another for /, and a rule for “unmatched”. Seems like the / would work for the /services path, but no luck.

Checking back in…much better now. Set both the admin & proxy to be type: LoadBalancer, I can now do things on the /service endpoint. Looking good!