google cloud endpoints api_method not found on gke - flask

Question

404 response Method: 1.api_endpoints_gcp_project_cloud_goog.Postoperation failed: NOT_FOUND on google cloud endpoints esp
I'm trying to deploy my API with google cloud endpoints with my backend over GKE. I'm getting this error on the Produced API logs, where shows:
Method: 1.api_endpoints_gcp_project_cloud_goog.Postoperation failed: NOT_FOUND
and i'm getting a 404 responde from the endpoint.
The backend container is answering correctly, but when i try to post http://[service-ip]/v1/postoperation i get the 404 error. I'm guessing it's related with the api_method name but i've already changed so it's the same in the openapi.yaml, the gke deployment and the app.py.
I deployed the API service succesfully with this openapi.yaml:
swagger: "2.0"
info:
description: "API rest"
title: "API example"
version: "1.0.0"
host: "api.endpoints.gcp-project.cloud.goog"
basePath: "/v1"
# [END swagger]
consumes:
- "application/json"
produces:
- "application/json"
schemes:
# Uncomment the next line if you configure SSL for this API.
#- "https"
- "http"
paths:
"/postoperation":
post:
description: "Post operation 1"
operationId: "postoperation"
produces:
- "application/json"
responses:
200:
description: "success"
schema:
$ref: "#/definitions/Model"
400:
description: "Error"
parameters:
- description: "Description"
in: body
name: payload
required: true
schema:
$ref: "#/definitions/Resource"
definitions:
Resource:
type: "object"
required:
- "text"
properties:
tipodni:
type: "string"
dni:
type: "string"
text:
type: "string"
Model:
type: "object"
properties:
tipodni:
type: "string"
dni:
type: "string"
text:
type: "string"
mundo:
type: "string"
cluster:
type: "string"
equipo:
type: "string"
complejidad:
type: "string"
Then i tried to configure the backend and esp with this deploy.yaml and lb-deploy.yaml
apiVersion: apps/v1 # for versions before 1.9.0 use apps/v1beta2
kind: Deployment
metadata:
name: api-deployment
namespace: development
spec:
strategy:
type: Recreate
selector:
matchLabels:
app: api1
replicas: 2 # tells deployment to run 2 pods matching the template
template:
metadata:
labels:
app: api1
spec:
volumes:
- name: google-cloud-key
secret:
secretName: secret-key
containers:
- name: api-container
image: gcr.io/gcp-project/docker-pqr:IMAGE_TAG_PLACEHOLDER
volumeMounts:
- name: google-cloud-key
mountPath: /var/secrets/google
ports:
- containerPort: 5000
- name: esp
image: gcr.io/endpoints-release/endpoints-runtime:1
args: [
"--http_port=8081",
"--backend=127.0.0.1:5000",
"--service=api.endpoints.gcp-project.cloud.goog",
"--rollout_strategy=managed"
]
ports:
- containerPort: 8081
kind: Service
metadata:
name: "api1-lb"
namespace: development
annotations:
cloud.google.com/load-balancer-type: "Internal"
spec:
type: LoadBalancer
# loadBalancerIP: "172.30.33.221"
selector:
app: api1
ports:
- protocol: TCP
port: 80
targetPort: 8081
my flask app that serves the api, is this app.py
app = Flask(__name__)
categorizador = Categorizador(model_properties.paths)
#app.route('/postoperation', methods=['POST'])
def postoperation():
text = request.get_json().get('text', '')
dni = request.get_json().get('dni', '')
tipo_dni = request.get_json().get('tipo_dni', '')
categoria,subcategoria = categorizador.categorizar(text)
content = {
'tipodni': tipo_dni,
'dni': dni,
'text': text,
'mundo': str(categoria),
'cluster': str(subcategoria),
'equipo': '',
'complejidad': ''
}
return jsonify(content)

Answer 1

Looks like you need to configure route in your flask app.
Try this:
#app.route('/v1/postoperation', methods=['POST'])

Answer 2

Some bits from kubectl expose -h
--port='' - The port that the service should serve on. Copied from the resource being exposed, if unspecified
--target-port='' - Name or number for the port on the container that the service should direct traffic to.
Optional.
While proxy directing your trafic to --backend=127.0.0.1:5000, use container name isntead --backend=api-container:5000.