Service working via Cluster-IP, but not via Ingress - kubernetes-ingress

Question

I need help figuring out what I'm doing wrong with my Ingress setup Minikube/ingress-nginx-controller. Kubectl version is 1.19. Minikube version is 1.13.1
I have 2 services: 1 that I created from an image I built myself in dotnetcore and another that I pulled from an example. The example is giving me no problems: I can reach it via http://myapp.com/web. The one I built can be reached directly via its Cluster IP (port 80) in the browser, but can't be reached from the browser using http://myapp.com/datasvc (404 error). Here's a snippet from my Ingress yaml:
- host: myapp.com
http:
paths:
- path: /web2 #works
pathType: Prefix
backend:
service:
name: web2
port:
number: 8080
- path: /datasvc
pathType: Prefix
backend:
service:
name: datasvc
port:
number: 80
And here's what my backends look like:
Rules:
Host Path Backends
---- ---- --------
myapp.com
/web2 web2:8080 172.17.0.7:8080)
/datasvc datasvc:80 172.17.0.8:80)
Services:
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
datasvc ClusterIP 10.100.7.119 <none> 80/TCP 11hde here
web2 NodePort 10.98.6.48 <none> 8080:31122/TCP 12h
CURL output from curl -H "HOST: myapp.com" localhost/web2 -v:
* Trying 127.0.0.1:80...
* TCP_NODELAY set
* Connected to localhost (127.0.0.1) port 80 (#0)
> GET /web2 HTTP/1.1
> Host: myapp.com
> User-Agent: curl/7.67.0
> Accept: */*
>
* Mark bundle as not supporting multiuse
< HTTP/1.1 200 OK
< Server: nginx/1.19.1
< Date: Sun, 04 Oct 2020 16:16:55 GMT
< Content-Type: text/plain; charset=utf-8
< Content-Length: 61
< Connection: keep-alive
<
Hello, world!
Version: 1.0.0
Hostname: web2-7d85fb54bf-f26p2
* Connection #0 to host localhost left intact
CURL output from curl -H "HOST: myapp.com" localhost/datasvc -v
* Trying 127.0.0.1:80...
* TCP_NODELAY set
* Connected to localhost (127.0.0.1) port 80 (#0)
> GET /datasvc HTTP/1.1
> Host: myapp.com
> User-Agent: curl/7.67.0
> Accept: */*
>
* Mark bundle as not supporting multiuse
< HTTP/1.1 404 Not Found
< Server: nginx/1.19.1
< Date: Sun, 04 Oct 2020 16:20:13 GMT
< Content-Length: 0
< Connection: keep-alive
<
* Connection #0 to host localhost left intact
The only difference I can see between the 2 is the fact that the example (web2) service uses type: NodePort and mine uses the default (type: ClusterIP). I tried doing the same with my service, but that made no difference.
I don't know what else to look at diagnostically or where to go from here. I've checked out many Medium posts but haven't come across anything that describes my situation. Please let me know if I should provide more information.

Answer 1

Ingress supports either NodePort service type or LoadBalancer. ClusterIP service will be available only inside minikube VM.
For minikube you should use NodePort service type configuration. To setup ingress on minikube you can follow official documentation.
To expose your deployment you can use kubectl:
kubectl expose deployment <deployment_name> --type=NodePort --port=<port_number>
To check if service is correctly working you can run minikube service list and curl URL that you have exposed.
If everything is working correctly you can setup your ingress and add IP Address and HOST to /etc/hosts.